journal The Journal of The Foundation for Science and Technology fstVolume 22 Number 2 March 2018 www.foundation.org.uk

Editorial Sir David Cannadine: The role of the Academies in providing independent advice to Government An industrial strategy for the UK Lord Hennessy: Searching for a strategy that makes a difference Lord Heseltine: Establishing a strategy for the whole economy Lord Willetts: A tension at the heart of Government activity A business strategy for Scotland Professor Iain Gray: Translating research excellence into economic benefit Nora Senior: Plugging gaps in performance Dame Susan Rice: An ecosystem for business Paul Wheelhouse: Driving innovation Meeting air quality targets Dr Stephen Bryce: The energy emissions challenge Professor Frank Kelly: The health consequences of air pollution Diagnosing cancer earlier Sir Harpal Kumar: Early diagnosis has the potential to transform patient outcomes Dr Clare Turnbull: Using genetics to combat cancer The rise of machine learning Dr Mike Lynch: An opportunity or a threat to society? Dr Claire Craig: Giving society the confidence to embrace opportunities Amir Saffari: The potential to augment human efforts Dame Wendy Hall: The opportunities for the UK Comment Norman Lamb: The future of social care Obituary The Rt Hon Sir Brian Neill COUNCIL AND TRUSTEES

COUNCIL CHIEF EXECUTIVE Chair Dr Dougal Goodman OBE FREng The Earl of Selborne* GBE FRS Deputy Chairs The Baroness O’Neill of Bengarve* CH CBE FBA FRS FMedSci Dr Mike Lynch* OBE FRS FREng DL

President, The Royal Society Professor Polina Bayvel CBE FRS FREng Sir Venki Ramakrishnan PRS FMedSci Sir John Beddington CMG FRS FRSE HonFREng President, Royal Academy of Engineering Sir Leszek Borysiewicz FRS FRCP FMedSci FLSW Professor Dame Ann Dowling OM DBE FRS FREng The Lord Broers FRS FREng President, British Academy Sir Geoffrey Chipperfield KCB Sir David Cannadine PBA The Lord Haskel* President, The Academy of Medical Sciences The Lord Hennessy FBA Professor Sir Robert Lechler PMedSci Dr Julian Huppert President, The Royal Society of Edinburgh Sir David King ScD FRS HonFREng Professor Dame Jocelyn Bell Burnell DBE FRS FRSE FRAS FInstP The Lord Krebs FRS FMedSci President, The Learned Society of Wales The Rt Hon Norman Lamb MP Sir Emyr Jones Parry GCMG FInstP PLSW Sir Rob Margetts CBE FREng Chair, Arts & Humanities Research Council The Lord Mair CBE FRS FREng Sir Drummond Bone FRSE Stephen Metcalfe MP Chair, Biotechnology and Biological Sciences Research Council The Rt Hon the Baroness Neville-Jones DCMG Sir Gordon Duff FRCP FRCPE FMedSci FRSE Sir Paul Nurse FRS FMedSci HonFREng Chair, Economic and Social Research Council Chi Onwurah* MP Dr Alan Gillespie CBE The Lord Oxburgh KBE FRS HonFREng Chair, Engineering and Physical Sciences Research Council The Lord Rees of Ludlow OM FRS HonFREng HonFMedSci Dr Paul Golby CBE FREng Dr Peter Ringrose Chair, Medical Research Council The Baroness Sharp of Guildford Donald Brydon* CBE Sir Adrian Smith FRS Chair, Natural Environment Research Council Dr Graham Spittle CBE Sir Anthony Cleaver HonFREng The Lord Trees Chair, Science and Technology Facilities Council The Baroness Wilcox Sir Michael Sterling FREng The Rt Hon the Lord Willetts Chair, EngineeringUK Sir Peter Williams CBE FREng FRS Malcolm Brinded CBE FREng The Lord Willis of Knaresborough Chair, Innovate UK Phil Smith Honorary Treasurer Chair, Steering Board, UK Space Agency Tony Quigley* Professor David Southwood Honorary Treasurer Designate President, The Science Council John Neilson* Sir Keith Burnett CBE FRS Honorary Secretary Patrick McHugh* *Trustee

The Foundation for Science and Technology FST Journal was redesigned in 2015 by IOP Publishing’s 10 Carlton House Terrace Design Studio, under the art direction of Andrew Giaquinto. London SW1Y 5AH IOP Publishing provides publications through which leading-edge scientific research is distributed worldwide and is central to the Tel: 020 7321 2220 Institute of Physics, a not-for-profit society. Email: [email protected] FST Journal publishes summaries of all the talks given at its meetings. Full audio recordings are available on the website Editor Dr Dougal Goodman OBE FREng Production Editor Simon Napper Neither the Foundation nor the Editor is responsible for the opinions Layout Simon Clarke of the contributors to FST Journal. Sub-editor Judy McBride © 2018 The Foundation for Science and Technology ISSN 1475-1704 Charity Number: 00274727 Company Number: 01327814 CONTENTS

journal fstVolume 22 Number 2 March 2018 THE COUNCIL AND TRUSTEES OF THE FOUNDATION Inside front cover

journal The Journal of The Foundation for Science and Technology fstVolume 22 Number 2 March 2018 www.foundation.org.uk UPDATE

Editorial Sir David Cannadine: The role of the Academies in providing independent advice to Government #ThisisEngineering aims to enthuse a new generation • Decarbonising the transport sector • 2 An industrial strategy for the UK Lord Hennessy: Searching for a strategy that makes a difference Lord Heseltine: Establishing a strategy for the whole economy Lord Willetts: A tension at the heart of Government activity A business strategy for Scotland PM announces new Institute of Coding Sam Gyimah is new science minister Successful launch Professor Iain Gray: Translating research excellence into economic benefit • • Nora Senior: Plugging gaps in performance Dame Susan Rice: An ecosystem for business Paul Wheelhouse: Driving innovation Meeting air quality targets promises low-cost, high-payload spaceflight Government aims to make the UK the ‘most innovative Dr Stephen Bryce: The energy emissions challenge • Professor Frank Kelly: The health consequences of air pollution Diagnosing cancer earlier Sir Harpal Kumar: Early diagnosis has the potential to transform patient outcomes Dr Clare Turnbull: Using genetics to combat cancer nation’ • Latest research finds global ozone levels ‘not recovering’ The rise of machine learning Dr Mike Lynch: An opportunity or a threat to society? Dr Claire Craig: Giving society the confidence to embrace opportunities Amir Saffari: The potential to augment human efforts Dame Wendy Hall: The opportunities for the UK Comment Norman Lamb: The future of social care Obituary The Rt Hon Sir Brian Neill EDITORIAL The role of the Academies in providing independent advice to Government Sir David Cannadine 4 AN INDUSTRIAL STRATEGY FOR THE UK Searching for a strategy that makes a difference Lord Hennessy 7 Establishing a strategy for the whole economy Lord Heseltine 10 A tension at the heart of Government activity Lord Willetts 12 Science and innovation strategies since 1946 16 A BUSINESS STRATEGY FOR SCOTLAND Translating research excellence into economic benefit Professor Iain Gray 17 Plugging gaps in performance Nora Senior 19 An ecosystem for business Dame Susan Rice 20 Driving innovation Paul Wheelhouse MSP 22 MEETING AIR QUALITY TARGETS The energy emissions challenge Dr Stephen Bryce 24 The health consequences of air pollution Professor Frank Kelly 25 DIAGNOSING CANCER EARLIER Early diagnosis has the potential to transform patient outcomes Sir Harpal Kumar 28 Using genetics to combat cancer Dr Clare Turnbull 31 THE RISE OF MACHINE LEARNING An opportunity or a threat to society? Dr Mike Lynch 35 Giving society the confidence to embrace opportunities Dr Claire Craig 37 The potential to augment human efforts Amir Saffari 40 The opportunities for the UK Dame Wendy Hall 41 COMMENT The future of social care Norman Lamb MP 44 OBITUARY The Rt Hon Sir Brian Neill 46 EVENTS Foundation events held since 26 October 2015 47

fst journal www.foundation.org.uk March 2018, Volume 22(2) 1 UPDATE

#ThisisEngineering aims to enthuse a new generation #ThisIsEngineering is a new campaign is an annual demand for at least 124,000 by the Royal Academy of Engineering in engineers and technicians with core engi- collaboration with EngineeringUK and neering skills, and an additional 79,000 industry partners. It aims to give more roles that require engineering knowledge young people from all backgrounds and skills alongside other skill sets. the opportunity to explore a rewarding On 1 February, Business Secretary ROYAL ACADEMY OF ENGINEERING ACADEMY ROYAL engineering career across a wide range Greg Clark also announced, as part of of industries from film, to sport, gaming the Industrial Strategy, a Government and music. #ThisisEngineering is backed by a investment of £184 million for 41 UK It comes as the Government has des- consortium of major engineering compa- universities to help train the next gener- ignated 2018 as the Year of Engineering nies, and has been created in response to ation of world-class engineers and excep- in a bid to tackle the engineering skills significant demand for engineering talent tional scientists at British universities. gap and widen the pool of young people in the UK. Findings from a forthcoming www.yearofengineering.gov.uk who join the profession. Engineer­ingUK ­report show that there www.raeng.org.uk Sam Gyimah is new Decarbonising the transport sector science minister Within a week of taking up his new Last summer, the UK Plan for Tack- role as a Minister in the Department ling Roadside Nitrogen Dioxide Con- for Transport, Jo Johnson set out the centrations was published by Defra and Government’s ambition to phase out all the Department for Transport. It out- diesel-only trains by 2040. Referring to lines how councils with the worst levels the already-adopted target to end the of air pollution at busy road junctions sale of all new conventional petrol and and hotspots will be expected to take diesel cars and vans in the UK by 2040, robust action. While this is focussed on

he argued that the rail industry would delivering nitrogen dioxide (NO2) com- have to ‘clean up its act’ as well. pliance at the roadside, the Government

Policy Exchange (CC-BY-2.0) Exchange Policy In a speech to the Knowledge Quarter plans to issue a comprehensive Clean Air Sam Gyimah MP has been appointed conference held at the British Museum Strategy which will address other sourc- Mini­ster of State for Universities, Science, on 12 January, he noted that in absolute es of air pollution in 2018. Research and Innovation. As Minister terms, carbon emissions from rail had Local authorities will be able to bid for Higher Education, his responsibilities increased by 33% since 1990. And he for money from a new Clean Air Fund cover both the Department for Business, told his audience that this state of affairs to support improvements which will Energy and Industrial Strategy (BEIS) could not continue. In order to tackle the reduce the need for restrictions on and the Department for Education. He problem “with the urgency it deserves”, polluting vehicles. This could include took up the post on 9 January. He was he committed the Government to setting changing road layouts, removing traffic elected the Conservative MP for East tough new environmental performance lights and speed humps, or upgrading Surrey in 2010. goals for each rail franchise. bus fleets. Previously, he had held the posts of: Parliamentary Under Secretary of State at PM announces new Institute of Coding the Ministry of Justice from July 2016 to January 2018; Parliamentary Under Sec- The next generation of digital specialists The consortium is formed of busi- retary of State at the Department for Edu- could be created through a new Institute nesses including IBM, Cisco, BT and cation from May 2015 until July 2016; and of Coding, a consortium of more than Microsoft, small and medium-sized Parliamentary Secretary at the Cabinet 60 universities, businesses and industry enterprises (SMEs), 25 universities, and Office from July 2014 until March 2015. experts set to receive £20 million to professional bodies such as the British He spent five years working for Gold- tackle the UK’s digital skills gap. Computer Society and CREST. man Sachs and then went on to help Speaking at the World Economic The 25 universities involved, led by build and develop a number of small Forum 2018 in Davos, Prime Minister the University of Bath, range from sector businesses in the training, recruitment Theresa May spoke about how the Insti- leaders in business and computer science and internet sectors. tute of Coding, a part of the Govern- (UCL and Newcastle University), experts His predecessor as Science Minis- ment’s efforts to drive up digital skills in arts and design (University of the Arts) ter, Jo Johnson MP, has been appointed through the Industrial Strategy, will to specialists in widening participation Minister of State at the Department for equip people of all ages with the skills and outreach (Open University and Birk- Transport and Minister for London. they need. beck, University of London).

2 March 2018, Volume 22(2) fst journal www.foundation.org.uk UPDATE

Successful launch promises low-cost, high-payload spaceflight Space X, the American developer of rockets and spacecraft, successfully launched its Falcon Heavy rocket from Cape Canaveral, Florida, on 6 February. It is designed to carry payloads of up to 64 tonnes into low earth orbit. That is three times as much as the European Space Agency’s Ariane 5 and nearly twice as much as the new rocket’s nearest rival, HEAVY – FALCON MISSIONSPACEX DEMO (CC0) the Delta IV Heavy, according to Space X. Indeed, only the massive Saturn V rocket used on the Apollo moon programme could deliver more payload into orbit. The Falcon Heavy is also partly reus- able, with the two side boosters landing safely afterwards. The third, central booster was lost, something the com- pany later ascribed to its running out of reducing the cost of launches dramati- about the launch was the choice of pay- ignition fuel. cally. He believes that the cost of a Falcon load for this inaugural venture: Musk’s The founder of Space X, Elon Musk, Heavy launch would be less than $100 Tesla sports car with a space-suited believes that the new rocket with its million, compared with more than more dummy in the driver’s seat and a David increased payload and partial recyclabil- than $400 million for its main rival. Bowie soundtrack. ity could represent a ‘game changer’, by One of the other main talking points www.spacex.com/falcon-heavy Government aims to make the UK the ‘most innovative nation’ The Government launched its Industrial its Industrial Strategy Challenge Fund The White Paper also confirmed Strategy White Paper on 27 November to respond to some of the major global that the Government would be press- 2017, setting out its vision for how challenges and the opportunities faced ing ahead with a series of Sector Deals. Britain can build on its economic by the UK. This will include £170 mil- The first of these sectors to participate strengths, address its productivity lion to help transform the construction in these strategic and long-term part- performance, embrace technological sector helping to create places to live and nerships with Government are con- change and boost the earning power of work that are affordable, safer, healthier struction, life sciences, automotive and people across the UK. and which use less energy. AI. The partnerships will be backed by With the aim of making the UK the There will also be up to £210 million private sector co-investment. world’s most innovative nation by 2030, to improve early diagnosis of illness- • See also page 7 of this issue. the Government plans to invest £725 es and develop precision medicine for www.gov.uk/government/topical- million over the next three years through patients across the UK. events/the-uks-industrial-strategy Latest research finds global ozone levels ‘not recovering’ The ozone layer is recovering at the poles, of the ozone layer at more populated lati­ one is even used in the production of an but unexpected decreases in part of the tudes is not recovering. ozone-friendly replacement for CFCs. atmosphere may be preventing recovery The cause of this decline is not certain, Study co-author Professor Joanna at lower latitudes, between 60°N and 60°S although the authors suggest a couple of Haigh, Co-Director of the Grantham (London is at 51°N), according to new possibilities. One is that climate change is Institute for Climate Change and the research. altering the pattern of atmospheric circu- Environment at Imperial College Lon- Global ozone has been declining since lation, causing more ozone to be carried don, said: “The potential for harm in the 1970s owing to certain man-made away from the tropics. lower latitudes may actually be worse chemicals. Since these were banned, The other possibility is that very short- than at the poles. The decreases in ozone parts of the layer have been recovering, lived substances (VSLSs), which contain are less than we saw at the poles before the particularly at the poles. The new result, chlorine and bromine, could be destroy- Montreal Protocol was enacted, but UV published in the European Geosciences ing ozone in the lower stratosphere. radiation is more intense in these regions Union journal Atmospheric Chemistry VSLSs include chemicals used as solvents, and more people live there.” and Physics, finds that the bottom part paint strippers, and as degreasing agents: www.atmos-chem-phys.net/18/1379/2018

fst journal www.foundation.org.uk March 2018, Volume 22(2) 3 EDITORIAL The role of the Academies in providing independent advice to Government David Cannadine

mong other things, governments are in The Academies the business of serious decision making. One of the important functions of the British To do this in the absence of good evi- Academy and the other national academies is to denceA and advice is foolhardy at best, at worst it champion and strengthen our respective disci- can be disastrous. History is littered with exam- plines as well as to promote their public value and ples of where governments have proceeded with- the contribution they can make to policy making. out these, with significant adverse consequences. We did just this during the passage of the Higher Professor Sir David For example, in the 1990s there were major fail- Education and Research Bill through Parliament Cannadine PBA FRSL FSA ings in the way evidence was sought and used in in 2016-17. The Bill presented a critical opportu- FRHistS is an historian tackling the BSE crisis: by 2000 this had cost the nity to revisit the principles and structures which of modern British history government £3.7 billion and damaged public underpin the funding of the UK’s world leading from 1800 to 2000. He 1 succeeded Lord Stern as trust . Many other examples can be found in the research base. We engaged with parliamentarians President of the British devastating book co-authored by Anthony King in both Houses. We emphasised both the impor- Academy in July 2017. He is and Ivor Crewe, alarmingly and appropriately tance of the Haldane Principle in enabling aca- Dodge Professor of History entitled The Blunders of our Governments2. demic expertise in decision making about fund- at Princeton University, a Visiting Professor of We are living in an age of vast and fast-moving ing of research as well as the need to limit the History at Oxford University technological change. Our world is increasingly Secretary of State’s powers to change the remit or and Editor of the Oxford complex and across the next five years the Gov- name of the Research Councils without consulta- Dictionary of National ernment will be taking some momentous and tion. These points were addressed in amend- Biography. He has previously taught at the University of far-reaching decisions in relation to a number of ments tabled to the Bill, thereby helping ensure Cambridge and Columbia issues. These range from withdrawal from the that UK Research and Innovation (UKRI) will University, New York. He EU, responding to the forces of globalisation, cli- continue to recognise the value of research across was Director and Professor mate change and dealing with the disruption the disciplinary spread. of History at the Institute of Historical Research brought about by technological transformation. from 1998-2003. (Picture The stakes of poor decision making by Govern- A target courtesy Benedict Johnson) ment have rarely been higher. We are pleased that the Government recognises The good news is that there is a rich resource the importance of investing in this excellence and of high quality evidence available for politicians has committed to achieving a target of 2.4% of and civil servants to use, as the UK has a long GDP invested in UK R&D within 10 years, as well and distinguished track record of research excel- as a longer-term goal of 3% – something that the lence on which Government can draw. Discovery national academies have jointly been calling for4. and innovation helped the UK become an So, while there is no shortage of high-quality, ­economic powerhouse during the Enlightenment diverse and excellent research being carried out in and the Industrial Revolution. As the recently the UK across all academic disciplines, there is a published Industrial Strategy3 notes, the UK need to ensure that this excellence is used effec- is still recognised as a global leader in science tively by Government, and that is where the Acad- and research, top in measures of research excel- emies, including the British Academy, also have a lence and home to four of the top 10 universities key role to play. We at the British Academy, along- in the world. side our colleagues at the three other national academies, The Royal Society, The Royal Acade- my of Engineering and The Academy of Medical There is a rich resource of high quality evidence Sciences, play a vital and unique role in providing evidence and independent advice to Government. available, as the UK has a long and distinguished The British Academy is the UK’s independent track record of research excellence. national academy representing the humanities

4 March 2018, Volume 22(2) fst journal www.foundation.org.uk EDITORIAL

The British Academy has hosted roundtables on a number of pertinent issues, including industrial strategy. MIKE PEEL (WWW.MIKEPEEL.NET) / (CC-BY-SA-4.0) / (WWW.MIKEPEEL.NET) PEEL MIKE and social sciences. It is first and foremost a fel- ficult for other organisations to emulate in quite lowship of distinguished academics selected for the same manner. their excellent research. Through the breadth and Like the British Academy, all the national acad- depth of expertise of more than 1,000 Fellows emies can draw upon both the expertise of their across the UK and internationally, we offer a Fellows and the research they fund in order to pro- unique and valuable resource to Government as it vide independent advice to Government. Between seeks (or should be seeking) to understand many us we represent a hugely valuable resource for pol- of the great questions facing humanity today. iticians and civil servants alike. This is especially Our Fellows have spent their lives studying the case when we join together, as we often do, and everything from how individuals behave to how pool our collective wisdom and insights from a nations are governed, how economies operate, range of disciplines. A good example of this is the and how cultures evolve. Taken together, the sub- work that The British Academy has done over the jects our Fellows study put the whole of human past year with The Royal Society on the gover- experience in context. Whatever the challenge nance of data management and use. policy makers are grappling with, or whichever opportunity they want to open up, our Fellows Data can bring to bear independent perspectives, rele- The rate at which data is accumulating, and the vant insights and robust evidence. inferences that can be drawn from it, are both increasing at exponential rates and this is likely to Supporting research continue at a breath-taking pace. There are huge We do not only offer the wisdom and insight of benefits to be gained from a data-driven econo- our Fellows as a resource from which policy my, but to realise these benefits societies must makers can draw. For over a century we have navigate significant choices and dilemmas. In supported and celebrated the best in UK and this fast-moving landscape, governance chal- international research. We fund outstanding lenges need to be addressed in a timely manner if academics at key stages of their careers under- the system for oversight of data management and taking innovative research – all of which excel- use is to retain public trust. lence we can, and do, put at the disposal of Gov- This is exactly the sort of complex and ernment on a regular basis. multi-faceted issue to which the national acade- Our research touches every corner of the mies can make a significant contribution. By col- world, helping us to understand our society laborating across many of our disciplines, the today, where it came from and how it might change in the future. Because we seek, in all of There are huge benefits from a data-driven our activities, to safeguard scholarly interests and academic freedom we can provide truly indepen- economy, but to realise these benefits societies dent advice to Government in a way which is dif- must navigate significant choices and dilemmas. fst journal www.foundation.org.uk March 2018, Volume 22(2) 5 EDITORIAL

the UK leaving the EU. To date we have jointly published four briefings related to the land border, the Common Travel Area and the Good Friday Agreement. These briefings raise awareness on these topics and the questions that need consider- ation and/or response, as the UK negotiates its withdrawal from the EU and beyond that period. We hope that they may help build broader pol- icy and public reflection on the need to address such questions and the various ramifications that they might incur. The briefings provide an informed and impartial arena for the understand- ing of these complex matters, illustrating items that need policymakers’ consideration, often in ways that they may not have considered.

Independent advice These examples illustrate just some of the ways in which the academies assist in providing indepen- dent advice to Government as well as the breadth of issues on which we can and do offer advice. As

TOM MORRIS / (CC BY-SA 3.0) BY-SA (CC / MORRIS TOM I mentioned at the beginning of this article, we are Over the past year The Royal Society has worked with the British living in a time of huge change so we need to keep Academy on the governance of data management and use. pace with the vexed issues that are coming our way so rapidly and those generational issues that academies were able to pull together a working remain unsolved. But to do this successfully there group drawn from a wide range of academic sub- must be more respect for, and appreciation of, evi- jects, including law, computing and data science, dence-based learning, as well as for truth, reason anthropology and ethics. and ideas. There also needs to be a greater eager- By working in this multi-disciplinary way, we ness on the part of policy makers to engage with were able to set out some high-level principles for those who are the experts on subjects of such vital data governance in the 21st century and recom- contemporary concern. mended the setting up of a body to steward the We must approach these questions with rigour governance landscape. In essence, we made the and by drawing on all disciplines. The challenges intellectual case, which enabled the Government we are facing today are as much human and soci- to act. They have now done so through their com- etal as they are technical and scientific ones. The mitment to establish a new Centre for Data Ethics combination of the humanities and social scienc- and Innovation. es working together with science and technology The academies also have a unique power to can help us all to understand and shape our eco- bring policy makers, Fellows and other experts nomic and social future. It will take fresh perspec- together to provide focussed, evidence-based tives and many minds to respond to the complex advice on a range of current policy issues. Over and interconnected questions that face our soci- the past year, The British Academy has hosted ety and inform the debate about our future. The roundtables on a number of pertinent issues, British Academy, alongside our colleagues at the including industrial strategy. other academies stand ready, willing and able to provide essential and independent advice to Gov- International cooperation ernment. ☐ We work with colleagues from other countries’ national academies and have recently joined with 1. Building a stronger future, Royal Society, British the Royal Irish Academy to harness expertise Academy, Royal Academy of Engineering and the across our islands on some of the thorny and com- Academy of Medical Sciences, 2015 plicated issues that we now face in the context of 2. The Blunders of our Governments, Anthony King and Ivor Crewe, One World 2013 The academies have a unique power to bring policy 3. Industrial Strategy, BEIS, Nov 2017 4. https://www.britac.ac.uk/sites/default/files/ makers, Fellows and other experts together to Investing%20in%20UK%20R%26D%2022.01%20 provide focussed, evidence-based advice. %28web%29.pdf

6 March 2018, Volume 22(2) fst journal www.foundation.org.uk INDUSTRIAL STRATEGY There have been a number of science and innovation strategies since the end of the Second World War. The present Government has been developing its own. A meeting of the Foundation for Science and Technology on 18 October 2017 considered what light history could shed on different approaches to industrial strategy.

Searching for a strategy that makes a difference Peter Hennessy

eviews can be very revealing about areas of public policy. By my calculation, the SUMMARY current so-called ‘capability review of the R2015 strategic defence and security review’ is the • There have been more than 20 science and 13th defence review since the Second World War. innovation initiatives by Government since 1945 The current industrial strategy is the eighth since • The UK must now surpass the performance 1945. There has, by contrast, been only one com- levels achieved by previous strategies The Lord Hennessy FBA prehensive review of welfare and deprivation in • Successive governments have struggled to started his career as a the past 75 years: the Beveridge report of 1942. meet funding aspirations for research and journalist working on In a contribution to the debate about the latest innovation The Times, The Financial industrial strategy, the Foundation for Science • Transforming research insights into commercial Times and The Economist, success has been a continuing challenge unearthing the hidden wiring and Technology has undertaken an exercise in of the constitution and the ‘documentary cartography’, identifying at least 20 • The productivity gap between the UK and its power of the machinery science and innovation initiatives of various international competitors remains a serious of Government in Britain. kinds since 1945 (see page 16) – and that does not concern. Those themes remained at the heart of his research include those mounted by individual Govern- and teaching at Queen ment Departments. A quick glance shows that Mary when he moved from their frequency has picked up since the early fore, we can find ourselves pointing out that, least journalism to academia in 1990s. Suffice it to say that several generations of of all nations can Great Britain afford to neglect 1992. Since then, it has been his aim to write the our best political, scientific and official minds whatever benefits the scientists can confer upon history of the country, in his have pitted themselves against this particular her. If we are to maintain our position in the world own times, for academic and bundle of problems. and restore and improve our standard of living, we public audiences alike. Lord have no alternative but to strive for that scientific Hennessy is a member of the Council of the Foundation The challenge achievement without which our trade will wither, for Science and Technology. Given the uncertainty we are facing as a major our colonial empire will remain undeveloped and (Photo © Layton Thompson manufacturing, trading, thinking and innovating our lives and freedom will be at the mercy of a www.laytonthompson.com) nation, this time we must somehow surpass the potential aggressor.” Take out the colonial refer- performance levels of the past on several fronts. ence and that could still serve pretty well today. Happily, there may be an embryonic political con- sensus on several aspects. Abiding concerns The first of these strategies was the Barlow There are three crucial themes that run through Report on Scientific Man-power of May 19461. It our score of reviews. was quite a committee that produced it, members First, there is the question of funding. A including Edward Appleton, Patrick Blackett and thread that runs through the post-war period has Solly Zuckerman, with C.P. Snow as assessor. been the difficulty in meeting the stated aspira- The opening paragraph states: “We do not tions of successive governments to raise spend- think it is necessary to preface our report by stat- ing on science and innovation as a percentage of ing at length the case for developing our scientific GDP and, in addition, to encourage the private resources. Never before has the importance of sector to follow suit. science been more widely recognised or so many The second one is commercialisation. There hopes for future progress and welfare founded has been a continual struggle to take to market upon the scientist. By way of introduction there- the world-class ideas created in UK science and fst journal www.foundation.org.uk March 2018, Volume 22(2) 7 INDUSTRIAL STRATEGY

The productivity gaps between ourselves and our plumbing rather than the poetry. The substance chief competitors are a prime anxiety in all the of a policy rarely appeals to the imagination. I remember a wonderful line of R.H. Tawney: “Only industrial strategies since the Attlee Government. those institutions that are loved appeal to the imagination.” The same applies to White Papers. ­innovation through conversion into patents and Getting this right, though, has been and must thence to commercial opportunities. We have not remain a great national endeavour. Knowing just done well in comparison to many of our leading a little about how we got to where we are today international competitors – both longstanding can help. Mark Twain put it rather well when he and more recent ones. said: “History may not repeat itself but some- The third theme that runs through this period times it rhymes.” concerns skills, education and productivity. The productivity gaps between ourselves and our chief Significant milestones competitors ring out as a prime anxiety in all the Several strategies stand out as significant mile- industrial strategies since the surveys produced by stones over the past 50 years. The recommenda- the Attlee Government’s Central Economic Plan- tions of the Rothschild and Dainton reports5 ning Staff. The same drumbeat of anxiety booms were accepted by the Heath Government and out from the latest industrial strategy and the implemented in the 1972 Framework for Gov- OECD has highlighted the problem too. ernment Research and Development. The cus- Huge efforts have been put into skills training. tomer-contractor approach remains contentious The Robbins Report on Higher Education2 in today. The question of who should decide the 1963 stressed the need for more science and tech- priorities for Government-funded research is a nology courses. Today, scientific and technical continuing debate. education elements remain a prime concern. Dr Miles Parker, the former Defra Deputy Enhanced performance is ever more vital. Chief Scientific Adviser, reviewed the impact of In their recent examination of the latest indus- the Rothschild Report of 1971 on The Organi­ trial strategy, the House of Lords Select Commit- sation and Management of Government Research tee on Science and Technology3 stressed the and Development, as well as the role of scientists potential benefits of a vigorous industrial strategy in policy development6. Rothschild had recom- with science and innovation as ‘pervasive themes’. mended the appointment of Chief Scientists to act as the customer for research in Government The big picture Departments, rather than Chief Scientific While there has been a variety of reports on ­Advisers who have an explicit role in advising aspects of the science and technology landscape, Ministers. there has never been a Beveridge-style examina- Miles Parker concluded that: “Rothschild tion, embracing in one sweep the whole breadth never really addressed the question of the func- of these issues and isolating the concerted and tion of scientists in Government policy-making simultaneous action required to break through Departments. His mind was wholly focussed on the crust of under-performance. The closest is research, not only because of the question put to Lord Heseltine’s 2012 report4 on growth entitled him and the Central Policy Review Staff (Ted No Stone Unturned. Heath’s creation in the Cabinet Office) but also The key to Beveridge was its focus on the slay- because of his previous Shell experience as Direc- ing of what he called the ‘five giants on the road to tor of Research. reconstruction’ – Ignorance, Idleness, Squalor, “Industry actually is very forward-looking and Disease, Want. Beveridge’s argument was that willing to wait for research results to flow in, but unless all of those five giants were smitten simul- policy makers in general cannot afford to wait for taneously, the crust of poverty would not yield. A the slow products of research to emerge. Politi- Beveridge for science, technology and innovation cians need the best available advice now, to take could take a similar, over-arching approach. action on the policy problems of today.” Beveridge used a particularly buoyant, perhaps The 1993 Realising our Potential: a Strategy for almost poetic, language. All too often Whitehall Science, Engineering and Technology7 document reviews concentrate, understandably, on the refocussed attention on cross-cutting funding of research to improve competitiveness and the There has been a variety of reports on aspects of quality of life. A spin-off from the review was the establishment by John Major of the Office the science and technology landscape, but there of ­Science and Technology in the Cabinet Office has never been a Beveridge-style examination. and the Technology Foresight process which

8 March 2018, Volume 22(2) fst journal www.foundation.org.uk Government Private and Research Non-Pro t Councils Organisations INDUSTRIAL STRATEGY

­successfully brought together end-users to artic- ulate what they wanted. In 2000, the seminal House of Lords Select 25 Committee inquiry on Science25 and Society8, chaired by Lord Jenkin, promoted the idea of a 20 two-way dialogue between scientists20 and the public, which of course is still a political theme. The 2006 Next Steps report on the science and 15 innovation framework set out15 the arguments for companies to stimulate investment in innovation £billion 10 and recommended the establishment10 of the Tech- nology Strategy Board which has since become Innovate UK. 5 As Minister for Science in the5 Blair Govern- ment, Lord Sainsbury stressed the importance of 0 funding for science and innovation0 during the Business Higher Government Private financial crisis with his Race to the Top report9. Education and Research Non-Profit The recommendations were accepted and the Councils Organisations Innovation Nation strategy with it. Source: Office for National Statistics Lord Willetts, as Science Minister in the ­Coalition, successfully defended research and Composition of UK gross domestic expenditure on research and innovation funding at a time when all Depart- development by performing sector, 2015 (£billion) mental budgets were being cut by 20-30%. Paul Nurse’s review set out recommendations for ­radical reform of the Research Councils which 1. Barlow Report on Scientific Man-power, 1946. were put into law in the Higher Education and www.educationengland.org.uk/documents/ Research Act of 201710. On 1 April 2018, UK barlow1946/barlow1946.html Research and Innovation (UKRI) will be up and 2. The Robbins Report on Higher Education, 1963. running, bringing together the seven Research www.educationengland.org.uk/documents/ Councils, Innovate UK and Research England. robbins/robbins1963.html However, discussion of how Departments are 3. House of Lords Select Committee on Science and managing their own research programmes Technology, 2017. www.parliament.uk/documents/ should not be forgotten. lords-committees/science-technology/Industrial- strategy/2017-05-02-Industrial-strategy-ltr-to- A vision for the future BEIS-Secretary-of-state.pdf The past 70 years have been a story of high inten- 4. Lord Heseltine, No Stone Unturned: tions, great expectations and sustained attempts in pursuit of growth, 2014. to find the Holy Grail of a science and innovation www.gov.uk/government/publications/no-stone- strategy, a solution that would somehow bring unturned-in-pursuit-of-growth about a shining synthesis and close the ever-stub- 5. Rothschild and Dainton reports. HMSO. (1972) born productivity gap, a solution that would Framework for Government Research and transmit waves of stimulus into the research and Development. Cmnd. 5046, HMSO: London development community and the private sector. 6. Dr Miles Parker, 2016. http://www.csap.cam. That was always a tall order and remains so ac.uk/news/article-rothschild-report-1971-and- today. Money is tight and every ounce of value purpose-government-fund has to be squeezed out of the science budget 7. Realising our Potential, 1993. www.gov.uk/ and particularly the extra £4.7 billion over five government/publications/realising-our-potential- years announced in the 2016 Autumn Statement. a-strategy-for-science-engineering-and-technology Government, as Nye Bevan once said of social- 8. Science and Society, 2000. ism, is all about the language of priorities and https://publications.parliament.uk/pa/ setting them is always stretching. ld199900/ldselect/ldsctech/38/3801.htm So, we must build on what has already been 9. Lord Sainsbury, Race to the Top, 2007. achieved but reach for a new way of working www.rsc.org/images/sainsbury_review051007_ together, striving for a new national trajectory tcm18-103118.pdf and indeed a great national endeavour in pursuit 10. Higher Education and Research Act, 2017. of a more vibrant economy, rising productivity www.legislation.gov.uk/ukpga/2017/29/ and a sustained betterment of our society. ☐ contents/enacted fst journal www.foundation.org.uk March 2018, Volume 22(2) 9 INDUSTRIAL STRATEGY Establishing a strategy for the whole economy Michael Heseltine

ne of the last initiatives I introduced in my company before I became a Minister was to SUMMARY bring in a regime to approve every expen- Oditure over 50p. Three weeks and an election cam- • The current global market place is not one of free paign later and I was a Minister in the Department of trade and open competition Transport. I was welcomed with a request to autho- • Companies need a stable home market in order rise £6 million investment in the electrification of a to successfully compete internationally The Rt Hon the Lord Heseltine branch line in Essex. When I asked to see the man- • We need some standard metrics to show just CH was elected an MP in agers and ask a few questions, I received the reply: how competitive British business is in the world 1966. He served on the “Oh, no, Parliamentary Secretary, if you start query- • Government strategy should focus on wealth front bench as a Minister ing the details of every £6 million we put in front of creation across the whole economy, not just with briefs for transport, traditional manufacturing sectors industry, environment, and you, we will never get anything done in this depart- defence. Under John Major, ment.” I had joined the public sector! Within three • Excellence needs to be driven through all levels he returned to the Cabinet years I had been responsible for the rescue of Rolls- of our economy if we are to succeed against our as Secretary of State for Royce and for persuading my European colleagues international competitors. the Environment, President of the Board of Trade, and to create the European Space Agency. I had master- between 1995 and 1997 he minded the sales campaign for Concorde. was First Secretary of State In 1983, US General Abrahamson came to see me possible to source one’s needs from overseas. But and Deputy Prime Minister. to tell me about a £29 billion research and develop- what happens to the industrial base when it does ment programme for which he was responsible. He not have a reliable home market, allowing it to sell was going to build an impenetrable defence system at marginal cost to the rest of the world? against missiles – it was to be called Star Wars. He The situation is complex. I know all the argu- came to see me because he had £100 million ear- ments about picking winners. But put simply, peo- marked for Heriot-Watt University in Edinburgh. ple by and large want to succeed and the responsibil- He explained that in this particular field of ity of Government is to enable them to do just that. research the University was a world pace-setter and we would become a partner in the pro- A really serious attempt gramme. He was expecting me to say: “Well that’s The fascinating thing about the latest proposals amazingly generous.” However, I knew that all for an industrial strategy is that they constitute a that commissioned research would be trans- really serious attempt to do something more com- ferred, at our taxpayers’ expense, back to the prehensive, in my view, than any other Govern- American industrial base. ment has tried. They also have one absolutely If anyone comes to me and says “just have open essential ingredient – the Prime Minister is in competition and free trade: that will really solve charge and without that single ingredient, noth- your problems,” I think back to Rolls-Royce. They ing works. So, she deserves credit for putting it on tried to compete with an American defence proj- the agenda and for being the chairman of the ect called the CF6 engine financed by the US tax- committee that is going to make it work. payer. Rolls-Royce had no defence project with One interesting aspect is the enormous frank- which to pay for the development of the RB211 ness in the way the initial Green Paper commits to engine – and so they went bust. large numbers of reviews. Clearly there is no need Now I am in favour of free trade. I am in favour for a review when the answer is already known. So of competition and of reducing barriers. Yet, in every promised review is a clear indication of what my experience, that has very little to do with the we do not know. Why not? Because Whitehall has world in which our companies operate. Time and not taken the trouble to find out. again, when the Ministry of Defence wanted to One of my criticisms of the proposals concerns acquire some sophisticated, high-tech weaponry, the attempt to focus the debate away from the total there would be a working American model avail- picture to those areas that suit the Government. I able immediately off-the-shelf. So it is perfectly want to widen the debate, particularly on the issue

10 March 2018, Volume 22(2) fst journal www.foundation.org.uk INDUSTRIAL STRATEGY

Bodies like the Port of London Authority could help to generate wealth for the country

of competitiveness. Underlying everything else is policies with associated agencies doing as they are the need to know just how good you are and, just told. So Government consists of 20 or 25 function- as important, how good that evaluation is in a al monopolies – and monopolies are dangerous. competitive global market place. Any worthwhile I believe that every Government Department company regularly carries out a SWOT analysis should have its own responsibility for wealth cre- (strengths, weaknesses, opportunities, threats) in ation. The word ‘industrial’ should be abandoned order to determine whether it is ahead of the com- because traditional manufacturing industry only petition, whether it is behind, what is coming up, accounts for 10% of the economy. The emphasis what the challenges are. should be on a ‘whole economy’ approach: many of Yet this does not happen in Government. There the most exciting growth areas lie outside the indus- is no reservoir of worldwide comparisons. Get into trial manufacturing base. Not only that but we must the detail, though, and the scale of the problem go beyond the traditional Departmental monopo- becomes clear. We all know the statistic – if Ameri- lies. Will the Department responsible for education can and German workers stopped work on Thursday talk to educational publishers (who report to evening, we in Britain would have to continue to the another Department) about the huge opportunities end of Friday to produce as much. in high-technology, individually-tailored learning? It certainly does not happen now. Competitive position Unless Departments are forced to explain their Knowing our competitive position is absolutely contribution to wealth creation to Ministers other fundamental. There should be an institutional than their own, little will change. process to do just that. At the heart of the com- The Royal Commission on Local Government petitiveness committee the Prime Minister has in England, 1966-1969 (more generally known as set up, there has to be a means by which, when the Redcliffe-Maud report after its chairman) Department A says: “We are winning out here looked at the devolutionary situation in this coun- Prime Minister,” a still, small voice can interject: try. There were at the time 1,400 local authorities “Prime Minister, I am afraid it is not quite as rosy – perfectly understandable because when they as you are being told.” were designed the only way to get anywhere was by That will create tension which people will not walking or on horse, so you needed a lot of local like because it disrupts and it challenges – but it is authorities. By the time of Redcliffe-Maud, the essential. In the private sector it happens every day. motor car and the railways had arrived. Another problem is that this country is The report proposed about 60 unitary author- over-centralised: I cannot think of a single ities and in East Anglia the four were Norfolk, Suf- advanced economy that is so functionally cen- folk, Peterborough and Cambridge. Today, tralised in its capital city. In the vast majority of though, there are five counties, the combined I cannot think of a cases, Government Departments carry out their authority of Cambridge/Peterborough, 15 district single advanced functions within a narrow definition of their pur- authorities as well as three independent study economy that is so pose. Someone does housing, someone does ser- groups (one looking at the corridor from Cam- functionally vices, someone does transport, someone does law bridge to Norfolk, one at the corridor from Cam- centralised in its and order. In large measure, they also execute the bridge to Stanstead and onto London and one at capital city. fst journal www.foundation.org.uk March 2018, Volume 22(2) 11 INDUSTRIAL STRATEGY

the corridor from Cambridge to Milton Keynes of our young people are being educated in schools and Oxford). Is it possible to design anything which the Department for Education itself knows more likely to delay development around one of are not good enough and yet this is tolerated. Why the world’s great universities? is it tolerated? The inevitable consequence will be The great thing about quasi-autonomous unemployment queues in the ever more competi- non-governmental organisations (quangos) is that tive and technological world of tomorrow. they are autonomous and are potentially huge wealth There is always a tension here. I know just how creators. Think of the Port of London Authority – good so much of this country is. However, while what could be more capable of improving the trade there is a great amount of excellence, it does not per- on the Thames? Some of the environmental agencies meate to those areas where it is needed if the average have huge ability to influence tourism – although that is to be raised. This patchiness reveals itself in the is not currently within their remit. problem of productivity. It is also the cause of not Today, though, there is a multi-layered local exporting enough, of selling too many of our compa- authority environment, with quangos and func- nies too cheaply and too quickly. We do not follow tional Departments just interested in their own nar- through in the depth that competitor nations seem to row roles. It is not the way to run a ship. do. That must change – and that is our challenge. ☐ In the end nothing matters as much as the skill, ability and energy of people. So where is the UK in • Lord Heseltine has written a response to the the world education league? Depending upon Government’s Green Paper. More details at: which statistics are chosen, the ranking comes in www.themagazineshop.com/bookazines/ at 21st, 29th or something of that order. Some 20% industrial-strategy-by-michael-heseltine A tension at the heart of Government activity David Willetts

hen Britain was lobbying America to enter the First World War, the realisa- SUMMARY tion dawned that all America’s lead- ingW universities were staffed by people who had • Many of the research funding arrangements that studied in Germany. Nine of the 25 professors we recognise today were originally devised heading departments at Harvard had been to during the First World War Germany. There was no comparable opportunity • There has been continuing tensions between Ministerial aspirations and pre-existing science The Rt Hon the Lord Willetts for them to study in the UK. served as Minister for The doctorate, a 19th century German inno- and research frameworks Universities and Science vation, was therefore established in Britain in • The current dual-funding system obscures a gap from 2010 to 2014. He is 1917 so that there was something to offer Ameri- in funding for institutions that are not universities the Executive Chair of the Resolution Foundation and a can students. At the same time, rather ingenious- • The UK is unusual in that its universities are not member of the Council of the ly, the universities argued that if they were to in the public sector Foundation for Science and deliver such services they needed funding: so the • Industrial strategies should be disruptive and Technology. He served as Higher Education Funding Council for England not merely reward incumbents. the Member of Parliament for Havant from 1992 to 2015. (HEFCE) came into being in 1918. Prior to his appointment The Department for Science and Industrial as a Minister, he worked at Research, the progenitor of the Research There followed 50 years of close links between HM Treasury and the No 10 ­Councils, was instigated in 1918 following the research and industrial policy because it was real- Policy Unit. Lord Willetts is a visiting Professor at King’s Haldane report. Universities UK had its origins ised that to be a serious military and imperial College London, Governor in the meetings of the Committee of Vice power in the 20th century required an organised of the Ditchley Foundation, ­Chancellors and Principals which began in 1918. research effort. Chair of the British Science In all, a flurry of institutional creativity, trying But confidence in that model waned. In 1964, Association and a member of the Council of the Institute for to catch up with what Germany and the USA the Wilson Government had been elected around Fiscal Studies. were doing. the principle of the ‘white heat’ of technological

12 March 2018, Volume 22(2) fst journal www.foundation.org.uk INDUSTRIAL STRATEGY revolution, but by 1970 Tony Benn thought the country was spending too much on R&D of insuf- ficient quality. Both main parties became wary of 140 the massive R&D-based, industrial effort. Many of the research associations created after the war were closed down. Over the past 20 years, though, there has been a revival, with industrial strategies 120 in some form or other. Yet these strategies have consistently been frustrated by the tension between what Ministers were trying to achieve and the existing science 100 and research framework which was based on: • an ambitious and successful strategy of maintaining research excellence across a whole range of disciplines. Britain has an 80 exceptionally wide range of areas where we Japan Canada UK Germany France Italy G7exUK US score very high citation rates, way above the average achieved by other similar-sized Source: Office for National Statistics economies; • rewarding research excellence, even at the compensation as they did not have endowments 2016 UK productivity expense of objectives like geographical like the Oxbridge colleges. per hour compared distribution or indeed industrial relevance. Although there are two flows of funding, there with other G7 So Nobel Prizes, academic citations and the is an omission: there is no reliable flow of funding nations (UK=100) prestige of the universities became the focus; for institutions outside universities. So it can be • a university-based research system. Looking tough for The Welding Institute, for example, (the at the distribution of publicly-funded R&D last survivor from the post-war creation of in advanced western countries shows Britain research associations) because it does not slot as one of the outliers, with an unusually high neatly into this regime. The same goes for the proportion based in universities. National Institute of Agricultural Botany. In fact, those that do survive tend to have prior and inde- Another peculiar feature of our education sys- pendent charitable status. One of the reasons I tem – one I regard as a fundamental weakness – is was keen to create the Catapult institutes was that students choose to focus on a narrow range of because we needed to sustain and fund research subjects (normally opting for either arts or sci- establishments linking public and private sectors ences) at age 16. They then go to university to outside universities. study these subjects. The universities in turn are So, behind the dual funding model, there is a staffed to teach those subjects and research gap in funding. focusses on them too. In essence, university research is heavily influenced by the subject pref- The Rothschild doctrine erences of 16-year olds. Lord Rothschild was crucial in shaping our cur- Our system is also very decentralised, with rent research model. He argued in his report that very high levels of institutional autonomy. While the science budget should be reserved for blue that has many strengths, it is built on three crucial skies ‘pure’ research. However, applied research doctrines which make it very hard to reconcile should be funded on a customer-contractor science strategy with the ambitions of industrial model and there was an obvious place to do so – strategy. These are: the way in which dual fund- central Government Departments. So, some of ing works; the Rothschild doctrine; and the par- the funding for the Research Councils was ticular position of universities. ­re-directed to Departments who would be the Dual funding appears to have a satisfying customers contracting to purchase research. completeness. Essentially, there is Research The Departments proved to be appalling cus- Council funding, which is project-based and not todians of this responsibility. Whenever there dependent on universities (though the people was budgetary pressure they cut funding, so the capable of doing that research are usually found in universities). Then there is HEFCE funding to Some of the funding for the Research Councils universities through the Research Excellence Framework (REF), whose roots go back to the was re-directed to Departments. They proved to argument that the Red Brick universities needed be appalling custodians of this responsibility. fst journal www.foundation.org.uk March 2018, Volume 22(2) 13 INDUSTRIAL STRATEGY

The ambition of successive governments – to nationalised industries, I cannot recall any that create an industrial strategy that included a focus considered what would happen to their R&D capabilities. That was, in fact, where a lot of the on R&D – has been very hard to deliver in practice. energy and water research had previously taken place. Yet, the financial framework chosen for science budget languished and then diminished these privatised bodies was price competition. over time. It is fascinating to see Margaret Thatch- Years later as Science Minister, I would talk to pri- er’s role in this. As Education and Science Minis- vatised utilities about R&D and they would ter, she supported the approach. However, when respond that the regulators were opposed to what she came back into office as Prime Minister, with was regarded as ‘gold plating’ or ‘non-essential’ Rothschild as Head of the Central Policy Review expenditure: price-regulation was the focus. Staff (CPRS), she said that the obvious customer So, the UK has a very different model of for applied research was the business sector. That research from almost any other advanced western again left an enormous gap in the system because country. That is why the ambition of successive business was not willing to do this. It is not that governments in the past 20 or 30 years – to create British businesses are unusually risk averse, rather an industrial strategy that included a focus on that in other countries public funding for research R&D – has been very hard to deliver in practice. goes closer to market. Concerns around ‘picking winners’ hang over So, in Britain we ended up with the ‘Valley of this debate, but in reality all Governments make Death’, the result of a particular set of institutional these judgements. The purists say Government arrangements not repeated in other countries. To should only have horizontal policies that operate counter this, a further set of intermediate institu- across the economy as a whole. But if you extend tions or funding arrangements were needed: Lord the Jubilee Line into Canary Wharf to enable the Sainsbury’s ingenious solution was the creation of City of London to expand, why not build a railway what became Innovate UK. I added to that with line linking the east and west coasts in the north of the Catapults. England to boost industry there? But one goes ahead and the other does not. It is impossible to Universities and research institutes have a pure horizontal strategy. Universities are private sector bodies which gives In the Coalition Government, an industrial them an advantage over research institutes in the strategy emerged out of four different approaches. public sector. It is one reason why such a high The Business Secretary Vince Cable focussed on proportion of R&D in Britain is done in universi- industrial sectors like aerospace and automotive ties. Internationally, research institutes have his- which were self-organised, ones where it was pos- torically been in the public sector and most coun- sible for a Minister to have a conversation with tries have a proud network of them. Germany has ‘the industry’. the Max Planck and Fraunhofer Institutes, while Second, you can focus on places – and of course the USA has the Argonne and Lawrence Liver- Michael Heseltine has been an eloquent advocate more National Laboratories, just to quote a few of that. It has a powerful logic: if investment is examples. The UK has never been successful in based on a cost-benefit analysis that is itself found- sustaining anything similar. Such institutions ed on where people are already living and where would here be subject to public sector rules on there is an identified need, then that essentially recruitment and on how they should function, so provides a responsive strategy. It is also possible to the private sector in general and universities in take a strategic view on what could happen particular has looked a better bet for research. through bold investments in particular places, as Departments themselves were uninterested in Greg Clark has been doing with his city deals and such mechanisms. George Osborne has been attempting with his British agricultural productivity has barely agenda for the North. grown in the past 10-20 years, precisely because of There is an excitement in tackling big challeng- Defra’s failure to support institutions intermedi- es and this constitutes a third approach. David ate between pure research bodies and farmers. Cameron was excited by the challenge of anti­ The former Department of Energy has hardly microbial resistance and by the challenge of been more successful. The National Nuclear Lab- sequencing the genome of 100,000 NHS patients. oratory (NNL) shrank until it survived just on Under his administration, Number 10’s agenda funding for individual projects. So, as the new was to bring strategies to life by setting challenges. civil nuclear programme gets underway, there is And then, fourth, my focus was on technolo- no core capability in this country. gies. One of the first tests I applied when During all the discussions on privatising approached by people with a set of particular

14 March 2018, Volume 22(2) fst journal www.foundation.org.uk INDUSTRIAL STRATEGY technologies was whether I could remember the hard to make progress in other areas. list: too many were random, there needed to be One criticism of previous industrial strategies some conceptual framework. What we tried to do is that they have rewarded incumbents. My view with the Eight Great Technologies was to start is that these technologies are inherently disrup- with digital, IT-based technologies. That covered tive. They should also be general purpose tech- e-infrastructure, autonomous systems and robot- nologies, so it is not possible to predict which ics. It meant satellites as increasingly important business sector will take them up. collectors and transmitters of data. Next was to Today’s industrial strategy has the backing of look at the life sciences: Britain is the home of the Prime Minister. UK Research and Innovation genetic sequencing, of fantastic advances in life (UKRI) is a body that could step up to the plate to sciences. By extraordinary coincidence, biologi- resolve some of the obstacles that have stopped us cal information now comes in digital form, as hitherto linking R&D strategy and industrial genetic code. Particular applications include syn- strategy coherently. thetic biology and agri-science. I believe we will now see a sustained focus on Then, there are the areas like energy storage industrial strategy in order to address the chal- and advanced materials without which it is very lenges facing the UK economy. ☐ The debate he new UK Research and Innovation the market. Patient capital is also in short supply, In the debate (UKRI) should use its new structures to with no real equivalent of the German KfW govern- following think strategically about how to get best ment-owned development bank. One role of Gov- the formal Tvalue from the additional £4.7 billion of funding. ernment is to bear risk that no commercial entity presentations, the It should also look at encouraging researchers to can take, and the UK is not strategic in this area. invited audience work more closely with industry to boost growth. Without such a proactive strategy, then letting raised a number of Years of pressure on public spending have meant innovative companies be sold abroad with the guar- issues including: that overall funding for research, as a percentage of antee of a continuing research focus in the UK may the role of UKRI GDP, has remained static at 0.7% between 1997 and be the best available option to keep value at home. in achieving 2010, largely due to declining Departmental spend- Polarisation between political parties is not best value on ing. Those same pressures, this time on applied conducive to maximising the value of our investment; technology spending, mean that restoring the research effort, though there is some underlying the function funding of Innovate UK should be a priority. consensus on what needs to be done. While the of disruptive Disruptive technologies, including machine German system is not the Holy Grail, it avoids technologies; learning, autonomous vehicles, block chain and short-term and confusing changes in the research the need for smart contracts, as well as synthetic biology and innovation landscape by modernising within strategic, patient (where the costs of gene synthesis are falling rap- familiar existing structures. capital; and the idly) need to be identified and supported. So an place of political industrial strategy must be quick to respond and Measuring success consensus. be flexible. For example, the car industry’s value Measuring university success in innovation should model is shifting rapidly towards companies pro- not be measured by start-ups alone: there should viding transport as a service rather than one be more focus on scale-ups and the best business where vehicles are owned by the customer. models for the future. The neglect of technical In terms of place, Cambridge now has a education has to end. research ecosystem, building connections There is an important and continuing role for between research parks, innovative graduates, foreign direct investment, but it would be naïve to and entrepreneurs. Academics need these net- think some overseas governments would not works to build good ideas into companies. want to move resulting advanced technologies Smart, flexible regulation has a role in driving back to their own base. innovation. Research excellence can help the UK Within Government, Ministers should avoid set future standards in areas such as embryology, short-term decision-making, and civil servants or 5G telecoms (if Ofcom prioritises density of should build more expertise in key issues by coverage over competition). remaining in post longer. There is a case for making The UK is short of R&D-intensive companies all public bodies report annually on how they have able to buy innovative research and scale it up for supported innovation and wealth creation. ☐ fst journal www.foundation.org.uk March 2018, Volume 22(2) 15 INDUSTRIAL STRATEGY

SCIENCE AND INNOVATION STRATEGIES SINCE 1946

As a contribution to the debate on the new Industrial Strategy set out in the Government’s Green Paper at the beginning of 2017, the Foundation for Science and Technology prepared a cartography covering initiatives to promote science, research, innovation and technology since the Second World War. Looking back 25 years, there has been a multitude of science and innovation strategies as well as supporting reviews and guidance. Each was set in the context of the time, e.g. in 2017 the themes are Brexit and continued financial pressures. Financial pressures were a feature of the 2008-2014 strategies. In Lord Sainsbury’s review in 2007 the theme was globalisation. The strategies since 1993 have contained many notable similarities on the contribution of science and innovation to the UK’s prosperity, funding (government and private funds), setting priorities, skills and education. Recent strategies have noted that the UK has ring-fenced and prioritised funding for science and innovation at a time when other public spending has been reduced. However, other nations have been increasing their spending on these areas during the same period. The full text of Science and Innovation Strategies since 1946: a cartography can be downloaded from the Foundation website at: www.foundation.org.uk/Events/pdf/20171018_Summary.pdf. It can be found at the end of the summary of the meeting.

UK science and innovation strategies – a timeline 2017 Industrial Strategy: building a Britain fit for the future, White Paper, November 2017 Building our Industrial Strategy, Green Paper, January 2014 Our plan for growth: science and innovation (Cm 8980) 2013 Government response to the Heseltine Review 2012 Heseltine Review - No Stone Unturned in pursuit of Growth 2012 DECC Science and innovation strategy 2011 BIS Innovation and Research Strategy for Growth 2008 Innovation Nation: Department of Innovation, Universities & Skills (White Paper) 2008 Implementing “The Race to the Top” Lord Sainsbury’s Review of Government’s Science and Innovation Policies 2007 Lord Sainsbury Review of the UK Science and Innovation system 2006 Science and innovation investment framework 2004-2014: next steps 2005, 2006, 2007 The ten-year science & innovation investment framework annual reports 2004 Science and Innovation investment framework, 2004-2014 2003 21st Century Skills: Realising Our Potential 2003 Innovation Report – Competing in the global economy: the innovation challenge, DTI 2002 Investing in Innovation: A strategy for science, engineering and technology, DTI, HMT, DES 2001 Opportunity for all in a world of change, A White Paper on Enterprise, Skills and Innovation, Cm 5052 2000 Excellence and Opportunity: a science and innovation policy for the 21st century 1998 Our Competitive Future – Building a Knowledge Driven Economy, White Paper 1996 Information Society Initiative (ISI) 1996 UK Government: Green Paper, “Government Direct”: A Prospectus for the Electronic Delivery of Government Services 1993 Realising our Potential: A strategy for science, engineering and technology (White Paper) 1972 A Framework for Government Research and Development, Cm 5046 (contains the Rothschild report and Dainton report – see below) 1971 The Organisation and Management of Government Research and Development (the ‘Rothschild Report’) 1971 The future of the Research Council System (the ‘Dainton Report’), Cm 4814 1968 Report by the Committee on Manpower Resources for Science and Technology on the Flow into Employment of Scientists, Engineers, and Technologists, Cm 3760 1965 Science and Technology Act 1963 Committee of Enquiry into the Organisation of Civil Science, Cm 2171 (The Trend Report) 1946 Barlow Report of 1946 on Scientific Man-Power

16 March 2018, Volume 22(2) fst journal www.foundation.org.uk SCOTTISH BUSINESS STRATEGY How should the funding of research and innovation by Government and private enterprise be channelled in support of businesses in Scotland? That question was debated in a joint meeting of the Foundation and the Royal Society of Edinburgh on 6 November 2017.

Translating research excellence into economic benefit Iain Gray

• fostering a culture of innovation, research SUMMARY and development; • promoting growth and creating opportunity • Despite its reputation for excellent science, through a fair and inclusive jobs market and Scotland’s expenditure on research and regional cohesion; development lags significantly behind that of promoting Scotland on the international other countries • stage to boost our trade and investment and • Current economic strategy focusses on investing extend our influence and networks. Professor Iain Gray CBE in people and infrastructure, fostering FREng FRSE is Vice- innovation, promoting growth and opportunities To support this strategy, a number of initia- President for Business of The for all, and boosting international trade tives have been launched. The Scottish Funding Royal Society of Edinburgh and Director of Aerospace • Enterprise funding has been increased and Council has funded the development of innova- at Cranfield University. He innovation centres have been developed to bring tion centres that will seek to bring Scottish spent his early career as universities and businesses closer together ­universities and businesses closer together. an engineer working for • Growth needs to be inclusive, covering all Funding through Scottish Enterprise Agencies Airbus rising to become geographical areas and harnessing digital the Managing Director has been increased. and General Manager for technology The RSE Enterprise Fellowship scheme, sup- Airbus in the UK. In 2007 • The future of the Scottish economy should be ported by Scottish Enterprise, three of the UK he became the first Chief viewed in both a UK and an international context. Research Councils and Quantic, has supported Executive of the Technology Strategy Board later many talented scientists and engineers in devel- renamed Innovate UK. oping business and entrepreneurial skills. cotland has an international reputation for The RSE has contributed to the UK Govern- excellent science and research. Five of its uni- ment’s consultation on Industrial Strategy1 and has versities are in the top 200 in world university engaged with the Scottish Government on the rankings.S In the 2014 Research Excellence Frame- future of the economic and skills agencies2. In addi- work, 77% of Scotland’s university research was tion, the Royal Society of Edinburgh’s Business deemed ‘world leading or internationally excellent’. Innovation Forum has agreed to commission stud- Despite this, expenditure on research and ies on three areas that are critical for economic development is significantly below that of the UK growth: commercialisation, productivity and skills. and even further below the OECD average, at a little over half that of the USA and Germany. How Fostering inclusive growth can we better translate our research excellence We must ensure that all our citizens have access to into economic benefit? training and knowledge throughout their lives. Inclusive growth also means seeking to involve all Current initiatives geographical areas of Scotland. Currently the Scotland’s economic strategy focusses on the two highest concentrations of economic and research mutually supportive goals of increasing compet- activity are around our major cities. itiveness and tackling inequality. The Scottish One of the ways we can seek to address this is Government has based its future growth plan on by harnessing the opportunities presented by four priorities: ­digital communications. Another is by providing • investing in people and infrastructure in a greater opportunities for people in more rural sustainable way; parts of Scotland to learn and train within their fst journal www.foundation.org.uk March 2018, Volume 22(2) 17 SCOTTISH BUSINESS STRATEGY

to the Migration Advisory Committee4. At the same time, Scotland should seek to extend its trade beyond Europe, particularly to the fast-growing economies of Asia, and to the USA. We have some large global companies investing in Scotland, such as Toshiba Medical, who were attracted by Scotland’s science base and incentives from Scottish Enterprise. We also have some promising emerging companies, for exam- ple Clyde Space, which is Europe’s largest micro- satellite manufacturer, based in Glasgow. We need to develop policies and an economic envi- ronment that will nurture such companies and enable them to grow into large-scale enterprises.

Improving future strategy The Innovation Centre programme was launched in 2012 to support To help improve Scotland’s business strategy fur- transformational collaboration between universities and businesses. ther, the RSE will be devoting considerable resources to finding the answers to some key own communities. The University of the High- questions, including: lands & Islands is a prime example of this. • Why is business investment so low? The skills necessary to support inclusive • Why is business research and development growth are not simply about scientific or techni- investment so persistently low? cal knowledge. It is also critical that we seek to • What are the obstacles to commercialisation? incorporate business and entrepreneurial think- • What is our place in partnership with others? ing into all levels of education. We cannot afford to be parochial – we must We will use the expertise of our Fellowship, consider the future of the Scottish economy in a together with that of other economic and business UK and global context. The UK Government is experts, as well as support from our staff. We have close to publishing a new Industrial Strategy been discussing this proposed work with the (published in November 20173). It is important country’s leading business organisations. I hope that this strategy reflects the needs of Scotland, and expect that many of them will contribute and equally important that the Scottish Govern- their knowledge and help us engage with their ment leverages the full opportunities provided by membership. We will also welcome evi- the UK strategy. The newly established UK dence-based contributions from individuals and Research and Innovation (UKRI) has a budget of organisations. more than £6 billion a year. Scottish universities and businesses need to ensure that they engage Shining a light effectively with UKRI. Scotland has to get its economic policy right if it is to proper. To do this, we must start by shining a A good Brexit light on why our previous actions have not Internationally, in my view the highest priority worked. In collaboration with partners from uni- should be to negotiate a positive outcome to the versities and business, as well as the UK and Scot- Brexit negotiations with our current partners tish Governments, we aim to develop a strong and within the European Union. By a positive out- productive strategy for business, and hence the come I mean one that maintains our relationship economy, across the country. Scotland’s research with the European research community and base is among the best in world – whatever we do, enables researchers throughout the European we must build on this. ☐ Economic Area to continue to work collabora- tively and move between countries. 1. www.rse.org.uk/wp-content/uploads/2017/04/ The RSE has produced advice on the econo­mic AP17_09.pdf impacts of Brexit, including a recent submission 2. www.rse.org.uk/advice-papers/phase-two- review-enterprise-skills-agencies We cannot afford to be parochial – we must 3. www.gov.uk/government/topical-events/the-uks- industrial-strategy consider the future of the Scottish economy in a 4. www.rse.org.uk/advice-papers/rse-response-to- UK and global context. mac-inquiry-into-eea-workers-in-uk-labour-market

18 March 2018, Volume 22(2) fst journal www.foundation.org.uk SCOTTISH BUSINESS STRATEGY Plugging gaps in performance Nora Senior

and the Scottish Funding Council) and represen- SUMMARY tatives of business, economics and local authori- ties. This membership allows business strategy to • Identifying and addressing performance gaps is be considered from a broad range of viewpoints, vital to bring Scotland into the top quartile of and it has an important role in finding ways to OECD countries plug gaps in Scotland’s business performance. • Workforce training, including management Successful competitors in other OECD coun- training, needs to be improved across Scottish tries are aggressively and affordably up-skilling Nora Senior CBE is Chair business their workforces, with a focus on the competences of the Enterprise and Skills • Digital technology needs to be used much more needed to advance themselves and their compa- Strategic Board as well as widely by businesses nies. I believe training budgets should become a Chair, UK Regions, for Weber Shandwick. She has been in • Businesses need to be supported to do more key focus for Scottish businesses. Much of what public relations for most of research and development our delivery agencies in this area are doing at the her career, working first with • The Strategic Board for Enterprise and Skills is moment could be taken on by businesses them- Saatchi & Saatchi and then well placed to create new partnerships between selves. Part of any new enterprise and skills strat- forming her own start-up company – the PR Centre. businesses, educators and Government egy has to be a call to arms to businesses, encour- agencies. aging them to engage much more fully in the training dialogue, alongside the research and innovation as well as technology dialogues. he Scottish Government aims to move The quality of management can be a key dif- the country into the top quartile of ferentiator in economic performance. Poor OECD countries for productivity, equal- management, poor managerial practices and Tity, wellbeing and sustainability. Business strate- poor attitudes have a direct impact on productiv- gy must put in place the elements needed to ity. There is evidence that the right type of achieve that goal. upskilling and re-training leads to significant To develop a meaningful strategy, it is vital to improvements in organisations, particularly in understand where the gaps in performance lie. productivity. All the research shows that by Scotland’s level of productivity is near the bottom improving standards of management, even by a of the second quartile and needs to improve by at small margin, substantial gains in efficiencies least 27%. Capital investment needs to rise by £10 can be made. Too many Scottish companies do billion, an increase of 95%. The country needs at not regard management training, re-training or least 6,000 more exporting companies. up-skilling as priorities in their workforces and so they do not invest enough in this area. Innovation imperative Scotland has to become three times more innova- Digital technologies tive, tripling our investment in research and Scottish businesses tend to perform poorly in development (an increase of £1.9 billion). Cur- terms of the take-up and use of digital business rently, in companies with 10 or more employees, tools. Although 87% of businesses use websites only 56% of those employees are actively involved and email, only 6% employ tools such as resource in innovation. planning, which are used by 43% of businesses in Finally, we need to reduce the proportion of Sweden. For example, only 19% of Scottish busi- the population with no or low qualifications, cur- nesses use customer relationship management rently 12% of those aged 16-24 (and 25% in the (CRM) tools, and even fewer, just 8%, use digital most disadvantaged parts of the country). The tools to manage their supply chains. people of Scotland face difficult challenges, but A key part of our strategy is to look at how to the economy has strengths in many areas, includ- embed technology into our businesses and devel- ing life sciences and financial services. op skills in the use of data and analytics. Invest- The Strategic Board for Enterprise and Skills ment in the use of digital technology is crucial. It encompasses five agencies (Scottish Enterprise, is worth noting that the UK will need to fill Skill Development Scotland, Highlands and 750,000 new digital jobs by 2020. In Scotland we Islands Enterprise, South of Scotland Enterprise will need people trained to fill 11,000 jobs that fst journal www.foundation.org.uk March 2018, Volume 22(2) 19 SCOTTISH BUSINESS STRATEGY

It will be crucial to will be created in the fintech (financial technolo- lems is the lack of data or real insight into Scot- further embed gy) industry alone. land’s asset base. By analysing such information, innovative practices, We need to increase the awareness of sources we could see where our competitive advantages generate efficiencies of innovation support in Scotland. Research and might lie. We also lack insight into the business or and create new development has a well-documented positive investment priorities of companies and universi- markets. This will not impact on growth and productivity, but R&D ties. These insights are needed in order to priori- happen through remains a high-risk activity for firms. It entails tise investments in research and development. investment alone. significant investment, it has uncertain outcomes It will be crucial to further embed innovative and it creates knowledge spill-over which goes to practices, generate efficiencies and create new other firms and organisations. markets. This will not happen through invest- ment alone. It will also depend on having people Essential incentives with the right skills. Since firms are not able to capture all the benefits We are leaving behind the days when employ- of their investment for themselves, they perform ers were passive consumers of the education and less R&D than would be socially optimal. So Gov- skills system. The way forward must include an ernment incentives to stimulate increased activi- increase in the partnerships between educators, ty, particularly among businesses, are justified businesses and official agencies. My aim, as Chair and need to be explored. of the Strategic Board, will be to move towards a The Strategic Board for Enterprise and Skills model in which our agencies work together for the will be looking at innovation support for different benefit of the user, rather than as a collection of business sectors, including cross-cutting oppor- parts or a selection of silos. Their activities should tunities to offer access to academic expertise and be combined into a coherent system that is easy for facilities across Scotland. One of the key prob- users to navigate. ☐ An ecosystem for business Susan Rice

here are a number of strands to a success- ful business strategy, and the key to devel- SUMMARY oping a strong and effective strategy is to Tlink these together into a mutually supportive • Business operates most successfully in a system. For example, innovation and investment mutually supportive ecosystem hubs could be aligned. Digital energy and trans- • Ongoing education is a key component of a port strategy could be positioned to support each successful business strategy Dame Susan Rice is a banker other. We need to think of our businesses as oper- • We need a diverse workforce with opportunities by profession. She was Chief ating within an ecosystem, and determine what for international collaboration Executive and then Chairman we can do to support this. • New regulatory or consumer pressures create of Lloyds TSB Scotland new opportunities for innovation, such as and a Managing Director of carbon-neutral solutions Lloyds Banking Group. She Education now pursues a career as a Although a high value is rightly placed on having • Large and small companies can be mutually non-executive director, and a skilled workforce, we also need a well-educated supportive, and both should be encouraged. trustee committee member. She is the Chair of Scottish workforce. This goes beyond literacy and numer- Water, of Scotland’s first acy, and extends to communication, prob- Fiscal Commission and lem-solving and questioning. We may teach busi- sense – one that includes people from different has chaired the Edinburgh ness or entrepreneurship but a successful worker, backgrounds, demographics, countries and International Book Festival. whether in a large organisation, self-employment career routes. They can have a strong positive or as a serial entrepreneur, is one who has been impact because they see things in the workplace educated to think and analyse. Education should through a different lens – keeping each other alert be continuous, beginning at school age, continu- and imaginative, and potentially more creative. A ing through further and higher institutions and business strategy that keeps the doors open for on into the workplace, throughout the span of an people to come and go is essential. individual’s working life. We need to recognise the importance of the We also need a diverse workforce in the widest international element and avoid separating

20 March 2018, Volume 22(2) fst journal www.foundation.org.uk SCOTTISH BUSINESS STRATEGY investment into inward or outward flows. Collab- We need to support the start-ups, the gazelles and oration in research, investment, test sites and markets should be encouraged, with all flowing the unicorns, knowing that when they interact with easily across borders. the big boys – the elephants – there will be success. Companies, whatever their size, have no choice but to respond to growing regulatory and consum- er pressures. The impact of climate change on might or might not be put. We turn to smaller business is fast becoming apparent through companies to get this work done, which helps extreme environmental events on the one hand them to become more innovative and thus more and ever more stringent regulation and investment marketable and successful. constraints on the other. Businesses have to exam- Having large companies close at hand enables ine their own models, their markets and their place smaller companies to gain early customers who within supply chains more critically now. are making new demands that require innova- tion and change. So it is important that a busi- Opportunities abound ness strategy keeps the doors open for larger As the economy is progressively decoupled from businesses as well as smaller ones. An environ- a carbon-intensive model, some businesses are ment that is attractive to larger businesses, as well finding new opportunities. That is especially true as sustaining small ones, would be the best out- of the small and medium sized businesses (SMEs) come because this approach would support the that are such a critical part of our economy here in whole market. Scotland. Scotland has a wealth of universities for a For example, Edinburgh-based Vegware devel- country of its size. Some are research universities ops, manufactures and distributes a range of com- and a number of them regularly produce new pletely compostable food packaging. Although spin-off businesses and start-ups. Many of these they are a local business, they work with some of new businesses grow from science and technolo- the largest contract caterers and distributors, Gov- gy. In these spheres, scientists need to be able to ernment offices, NHS trusts, tourist attractions, collaborate without borders or other artificial food retailers and events organisers. These clients constraints. So a business strategy should enable all demand zero-waste solutions. Vegware has or enhance international research links among grown nine-fold over the past five years and is the institutions and with industry. third fastest-growing firm in Scotland. Another example is John Lawrie Ltd, an Leading role exporter of mainly tubular metal for recycling Edinburgh University, for example, plays a lead- and reuse. Their business model is based on the ing role in the field of informatics and technology. circular economy – re-use what you have again It gave birth to Dolly the sheep and the research and again. It has brought them great success in behind the Higgs Boson. Not only do universities recent years, leading to 90 jobs here and around create spin-off businesses, but the quality of their the world, as well as trade with North America, research also attracts business and this can be sig- Europe, the Middle East and China. Not bad for nificant. Skyscanner was attracted to Scotland a scrap metal merchant that started out in Aber- because of Edinburgh University’s strengths. deen in the 1930s! FanDuel of Fantasy Sports Company grew out of Businesses like Vegware and John Lawrie will an Edinburgh spin-off. thrive if transport and communications within Business unicorns are start-ups worth a bil- the country work well and if there are agencies lion pounds. Business gazelles are fast-growing and organisations to provide support, such as companies that start with a million pounds and Entrepreneurial Scotland and the Scottish Inno- increase their revenue at least 20% annually for vation Centres. The Edinburgh Centre for Car- four years. That means they have doubled their bon Innovation, for example, delivers support revenue over four years and typically have had a and funds to help good low-carbon businesses get disproportionately positive impact on job cre- off the ground. It also hosts the Climate Knowl- ation. More jobs may lead to a growing popula- edge and Innovation Programme. tion, which is one of the characteristics of a thriv- Large companies also innovate. Scottish ing national economy. Water is using robots to travel through under- We need to support the start-ups, the gazelles ground pipes and detect potential stresses and and the unicorns, knowing that when they inter- leaks: this avoids the need to dig up all the pipes. act with the big boys – the elephants – there will Drones are being used to investigate above be success. As for the elephants themselves, we ground in order to find out where pipework should never discourage them! ☐ fst journal www.foundation.org.uk March 2018, Volume 22(2) 21 SCOTTISH BUSINESS STRATEGY Driving innovation Paul Wheelhouse

he fundamentals of the Scottish economy are strong – employment is above its SUMMARY pre-recession peak and international Texports increased by 41% between 2007 and 2015. • The Scottish economy is strengthening, In the first half of 2017 the number of people in although growth is slower than desired employment rose to 2.6 million. However, eco- • Businesses need practical help to realise their nomic growth is still slower than we would like. growth potential and an additional £45 million has been made available for R&D Paul Wheelhouse MSP Our major trading partners are the rest of the is Minister for Business, UK. We have been vocal in telling Westminster • Strengthening links with universities and Innovation and Energy in the that we and our agencies, and indeed the other colleges through innovation centres will enable Scottish Government. He devolved nations, need to be engaged fully in the Scotland to better align research with the needs has been in post since May of the economy 2016. He is a professional development and delivery of the UK Industrial economist and, since 1992, Strategy. Yet Government engagement is only • Improving digital connectivity and skills is vital to had specialised in Higher part of the answer. We need to encourage and speed up the innovation cycle and Further Education facilitate businesses to drive us forward. • Public sector collaboration with technology markets, as well as policy start-ups provides a route for entrepreneurs and evaluation, economic Small and medium-sized businesses with high appraisal and impact growth potential need practical support that businesses to win public sector contracts. assessment of capital includes access to finance. To this end, the Scottish projects. He has served on the Scottish Parliament’s Government announced an additional £45 million Finance Committee and in research and development grants over the next ing our world-leading research and making sure the Standards, Procedures three years. This uplift in support will enable more that wherever possible it is aligned with the needs and Public Appointments ambitious businesses to invest in research and of our economy, we can drive major innovation, Committee. development. The aim is to double business investment and growth in the economy. expenditure on R&D to £1.75 billion by 2025. We This does not just apply to universities. Col- expect the Government investment of £45 million leges also play a vital role in supporting innovation, to be topped up with business adding a further engaging with large numbers of businesses on a £222 million. That gives an idea of the scale of the weekly basis. We are looking to support this work leverage anticipated from this investment. through the College Innovation Fund pilot. There are still many SMEs across the economy which are Business catalyst not yet engaged. The Government has committed The Government is also investing in delivering a up to £120 million to support a network of eight national manufacturing institute for Scotland to innovation centres over a six-year period. facilitate innovation and to catalyse our businesses. This will not only contribute to business growth Digital connectivity and skills but will also help to boost Scotland’s overall inno- Scotland’s digital economy has made great strides. vation performance. There are many small and Digital technology allows us to speed up cycles of medium-sized businesses (SMEs) that lack access innovation and to link rural and urban econo- to the special equipment and necessary skills need- mies. We are on track to have 95% of Scottish ed to develop prototypes, so this will play a very premises able to benefit from fibre-broadband important role in meeting these requirements. thanks to continued funding and support from We need to strengthen the links between our the Scottish Government and hard work by local business and academic communities. Scottish uni- authorities. But we are not stopping there: our versities are the envy of the world – not only do ambition is for 100% coverage by 2020. they provide excellent teaching to the innovators Digital connectivity is vital, but to fully capital- and entrepreneurs of tomorrow, they also produce ise on it we must have digital skills as well. These some of the best research. Many institutions have skills are the key to opportunities and success in knowledge and experience that are internationally the global economy. We are therefore committed recognised. For example, a recent report on global to expanding our pool of digital skills and capabil- innovation identified the University of Dundee as ities. We are building upon foundations already the world’s most influential scientific research laid, such as the STEM programme and the institution in pharmaceuticals. By better harness- CodeClan digital skills academy, both of which are

22 March 2018, Volume 22(2) fst journal www.foundation.org.uk SCOTTISH BUSINESS STRATEGY helping to accelerate Scotland’s progress in build- ing a high-performing digital economy. FURTHER INFORMATION We are also aiming to embed digital literacy from early years onwards by expanding the num- Royal Society of Edinburgh ber of coding clubs in schools and boosting teach- www.royalsoced.org.uk er numbers in computing. We are working with industry to tackle skills shortages and to increase Strategic Board for Enterprise and Skills the number of pathways into the workplace. beta.gov.scot/groups/enterprise-and-skills-strategic-board It is imperative to address the gender gap that continues to blight the digital sector. The propor- tion of women in digital technology roles in our It will also provide a route for entrepreneurs, country is only 18%. Clearly that is far below what start-ups, SMEs and others to take forward digi- it should be, and limits both our talent pool and tal transformation in the public sector. CivTech our ability to understand the needs of our users. will offer a rapid and secure pathway for busi- nesses to win public sector contracts. Through Ground-breaking programme collaboration with technology start-ups as well Government has to practise what it preaches. I as the brightest and best in the private sector, the am therefore very pleased with the CivTech public sector will benefit from improved service pilot that was launched last September. CivTech delivery and outcomes. We have doubled the is a ground-breaking programme that is har- funding to £1.2 million to allow this programme nessing new technologies to drive innovation to be scaled up. Working together, we can create into the public sector. It is bringing together a more prosperous, innovative and inclusive private sector innovation, public sector organi- society. We want a Scotland where people have sations and citizens to develop more efficient the opportunity to flourish, no matter where they products and services. were born or who they are. ☐ The debate he STEM subjects (science, technology, and public services, together with a good quality In the debate engineering and mathematics) are of life generally, are among the factors that influ- following the important to any business strategy, as is ence graduate retention. presentations, a Tthe need to attract more women into them. How- There are excellent examples of innovation in number of topics ever, the arts and humanities should not be Scotland but not all areas have benefitted. Efforts were raised, ignored: some STEM graduates move into careers to embed the digital economy could help to including: the in other work areas. improve the career prospects of young people. importance Universities can make a fundamental contri- of STEM; the bution to commercialisation by turning out SME challenge skills agenda; good-quality graduates. There is, though, a more There is a particular challenge regarding small encouraging general need for a skilled and well-educated and medium-sized enterprises, since they graduates workforce at all levels, and in this context tackling account for the vast majority of Scottish business- to remain in current under-achievement in Scottish education es. In these, innovation usually needs to be incre- Scotland; is a priority. mental and takes time. Tax credits for research and digital The Government could give greater direction to and development to encourage SMEs to work technologies. universities on recognising the needs of business. with universities should be considered. The nur- Although strong links between universities and turing of medium-sized companies is especially business are important, there are other approaches important since some of them have real potential that are needed to create better links. For example, to grow and contribute to the Scottish economy. parents could be made more aware of business Scotland’s relatively poor business productiv- career opportunities at the stage when they are ity needs attention. Greater investment by busi- advising children on subject choices. nesses in technology is important. Equally, there Encouraging graduates to remain in Scotland is a need for management in some sectors to com- after graduation is important which may have municate more effectively and work more closely implications for the Scottish Government’s taxa- with the workforce in introducing technological- tion policy. The availability of high-quality jobs ly-driven change. fst journal www.foundation.org.uk March 2018, Volume 22(2) 23 AIR QUALITY What needs to be done to meet urban air quality targets and what are the consequences if the targets are not met? These issues were debated at a meeting of the Foundation, held at The Royal Society in London, on 26 April 2017.

The energy emissions challenge Steve Bryce

here are about 7.4 billion people in the world today, and this number is expected to SUMMARY rise to 10 billion by 2050. About 54% of us Tlive in cities, but this is expected to increase to 70% • The world’s energy needs are expected to by that date. The amount of energy we will need is increase by 67% by 2060 expected to increase by 67% by 2060. This is partly • Urban air quality needs to be addressed as part due to an expected increase in the number of vehi- of a broader perspective that includes energy Dr Stephen Bryce is Vice- cles, from 800 million today to 2 billion by then. accessibility, affordability and sustainability President, Fuels Technology, Renewable energies will become more of the • There is no ‘one size fits all’ solution – different Shell Projects and ‘norm’, although there will still be a significant role cities will need different solutions Technology. He is an expert in future and specialist fuels for hydrocarbons. Currently more than 50% of the • Changes in lifestyle, such as more car sharing and manages the research world’s energy is supplied by oil and gas, with 29% and reduced car ownership – as well as the and development that from coal. It is perhaps surprising that 30% of the advent of new technologies – have the potential to reduce transport energy use supports Shell’s technical world’s domestic and commercial energy is still partnerships. In 2008, he became President of Shell supplied by traditional biomass, including the • New energy sources will require partnerships Cansolv, an innovative, burning of wood, charcoal and animal dung. between Government and industry, for example technology-centred Urban air quality needs to be addressed as to provide hydrogen refuelling stations or ensure that charging points for electric cars are company that offers high- part of a broader view of sustainability from an efficiency capture solutions standardised. for the removal of sulphur environmental perspective. While that includes dioxide and carbon dioxide CO2, it is important not to focus just on one from gas streams in various aspect to the exclusion of others. We also need to industrial applications. look at air quality from the perspectives of energy New technology, new opportunities accessibility, affordability and sustainability. We In addition, new technology is opening up new should not optimise one priority at the expense opportunities. Autonomous vehicles available on of the others. To reach the best overall solution call might reduce the need for individual car own- we need to look at these three aspects together. ership in the future. Heavy goods vehicles can Transport-related energy consumption varies travel close together in platoons – lines of 10 – considerably across the cities of the world. For coupled electronically and benefiting from more example, the US city of Atlanta is geographically efficient aerodynamics. New developments such spread out and has a high level of car ownership. as these and others will have an impact on urban It uses more energy for transport than New York, air quality. which is less spread out, has a lower level of car We use the term ‘mosaic of potential solutions’ ownership and a more comprehensive public to describe how energy demand can be met at the transport system. same time as improving air quality. Different cit- There are changes in models of car ownership, ies will need different solutions. Our mosaic still however, and in the numbers of people who want includes fossil fuels, but increasingly will rely on to own a car. Where I live in Hamburg, car own- alternatives including renewable fuels, solar ener- ership is becoming less popular and car sharing is gy, batteries and hydrogen. Their role will vary increasingly prevalent. The cost of car ownership depending on local circumstances. What is good is also rising. There are now limits on the entry of for Manila will not be good for Manchester and vehicles into inner cities. In some cities such as what is good for London may not be good for Los Singapore, the cost of obtaining a licence is very Angeles. So we need to be very mindful of those high, and in others such as Beijing it is very diffi- local circumstances. cult. These policies are all, in one way or another, The issues must be looked at from a broad per- driven by the need to reduce air pollution. spective; we must not solve one problem by creat-

24 March 2018, Volume 22(2) fst journal www.foundation.org.uk AIR QUALITY ing another – for example, creating a problem with A joined up pathway is urgently required for electric electricity demand by switching to electric vehicles. The pathways from energy sources through to charging for cars. The industry needs to work with final products, whether they be fuels, electrons or governments to quickly establish a standard. something else, are very complex. Creating suc- cessful pathways will require partnerships between Another example is hydrogen refuelling. Ger- governments at local, regional and national levels, many is developing a number of hydrogen refuel- businesses and society as a whole. This includes ling sites. Working in partnership with the gov- making sure that end users are informed about ernment, vehicle equipment manufacturers and energy issues and can take these into account when energy providers, this programme will provide deciding how they want to live. 400 hydrogen refuelling sites across Germany by One area where a joined up pathway is urgently 2023. This has been made possible only because of required concerns electric charging for cars. At the these working partnerships. It is essential to have moment, there are at least three different connec- all the pieces of the puzzle in place when trying to tion types, which clearly limits uptake of electric develop a new energy source. If one piece of the cars. The industry needs to work with govern- puzzle is missing it will not be sustainable. ments to quickly establish a standard for electric We believe that improvements in urban air charging. For this reason we are working together quality are certainly attainable. However, they will with companies such as Daimler in the Charging depend on partnerships between governments and Interface Initiative to promote the standardised industry, and a broad-based approach that takes Combined Charging System. account of energy affordability and availability. ☐ The health consequences of air pollution Frank Kelly

update is in progress and is scheduled for publica- SUMMARY tion in 2019. The WHO guidelines cover key pollutants

• WHO issues targets for air quality in Europe, such as NO2 and PM2.5 (fine particulates less than including target levels of PM2.5 and NO2 2.5 micrometres in diameter, which can enter the 3 • In many UK cities including London these lungs). The 2005 target for NO2 is 40mg/m . For targets are regularly exceeded 3 PM2.5 it is 10mg/m . In London the level of PM2.5 3 • There is diverse and extensive evidence from is around 14-15mg/m – about 50% higher than Frank Kelly is Professor of studies in the laboratory and in the field that air the WHO target. Environmental Health at pollution damages health Breathing this quality of air over a long period King’s College London. He is • Much of the problem is caused by modern of time leads to health problems for many people. Director of the Environmental lifestyles Research Group, Director of In 2010 the Department of Health Committee on the NIHR Health Protection • We can improve the quality of our air, and our the Medical Effects of Air Pollution reported that Research Unit on the Health health, by adopting measures to tackle PM Impact of Environmental 2.5 poor air quality, measured as levels of PM2.5, was Hazards and Deputy and NO2, in much the same way the 1956 Clean associated with the loss of 340,000 life-years Air Act successfully tackled black smoke and Director of the MRC-PHE across the population – equal to 6 months for sulphur dioxide. Centre for Environment & every person in the UK, or equivalent to 29,000 Health. Professor Kelly leads premature deaths per year. substantial research activity But of course death is the extreme outcome. It spanning all aspects of air pollution research, from argets for air quality in Europe are based sits at the top of a pyramid of health problems toxicology to science policy. on guidelines issued by the World Health (see Figure 1, page 26). We know that on high air Organisation (WHO), derived from pollution days there are more hospital admis- Treviews of the health literature available at the sions, A&E attendances and GP visits to treat time. The guidelines were first issued in 1987, respiratory and cardiovascular problems. Epi- and were revised in 1997 and 2005. The next sodes of high air pollution occur regularly; during fst journal www.foundation.org.uk March 2018, Volume 22(2) 25 AIR QUALITY Impact of air pollution on health

losing up to about 9% of their lung volume. Death The implications for these children are life- Hospital Admissions long, because lung growth continues only until the age of 18. From about age 30 we start to lose A&E Attendances lung volume, which is why we find it increasingly Visits to General Practitioners difficult to carry out exercise such as running up stairs as we get older. The tragedy for these chil- Increased Medication dren is that they might reach age 18 without full Symptoms lung development, and will carry that burden for the rest of their lives. So it is crucial to better Impaired Lung Function understand this issue and stop this generation of Sub-clinical effects children being affected in this way. It has been known for a long time that breath- ing air of poor quality affects our lungs. But what we now appreciate is that the health effects go Figure 1. The the first four months of 2017 we had an episode beyond the lung. There are also now strong links impact of air lasting a few days in each month. People with with cardiovascular disease. In the report Every pollution on health health conditions suffer more consequences Breath We Take1, chaired by Professor Steven during these episodes. Holgate, it was shown that air pollution affects Poor-quality air does not just affect people nearly every tissue in the body at every stage of with existing health conditions. Laboratory life, including the unborn child in the womb. experiments in healthy volunteers in Sweden have Studies have shown that mothers in areas with shown that two hours of exposure to diesel fumes high air pollution have smaller babies. Effects on – roughly equivalent to the levels found on Lon- young children, in particular lower airway respi- don’s Oxford Street – produced an inflammatory ratory tract infections and, in some instances, response in the lungs, indicating that the lungs effects on lung growth, have been shown in some were struggling to combat the fumes. polluted environments. We also have evidence from studies conducted As we move into adulthood we may start to in the real world. A study about 10 years ago on develop chronic conditions such as diabetes, and people with asthma looked at their lung function later neurodegenerative conditions such as when they were walking in Hyde Park and com- ­Alzheimer’s or Parkinson’s disease. In the past, pared it with their lung function when walking on we have viewed these as inevitable consequences the much more polluted Oxford Street. Lung of ageing. Yet, increasing it seems that air function was measured using FEV1, (forced expi- ­pollution is contributing to the development of ratory volume), which is the volume of air that can these conditions. The evidence base linking be forcefully exhaled in one second. poor air quality and poor health is now diverse and extensive, consisting of thousands of studies Impact on children done across the world. The study volunteers walked for two hours and their FEV1 was measured for 24 hours afterwards. Seeking solutions After the Hyde Park walk their FEV1 had We all contribute to the problem in one way or decreased by about 2%, but after walking on another – driving our cars, jumping into taxis, Oxford Street for the same amount of time, their lighting our woodstoves at the weekend. Although FEV1 had dropped by about 6%. The reduction in we do not think about it, these actions are all lung function lasted for up to 24 hours afterwards. ­adding to the problem. However, we know it can This would happen to most people with asthma be solved, in the same way that the problems every time they experience high levels of pollution. of black smoke and sulphur dioxide were solved To find out what impact London’s air pollution in the 1950s. The concentrations of both pollut- Air pollution affects is having on children’s health, we did a six-year ants fell dramatically after the 1956 Clean Air Act nearly every tissue in study called the Exhale Study. This looked at the was enacted. We now need to find similar the body at every effects of exposure to NO2 on about 2,000 school ­solutions to deal with the modern air pollutants stage of life, including children in east London. Most of these children PM2.5 and NO2. ☐ in the womb. Studies live in areas or streets that exceed the WHO target 3 1. have shown that for NO2 of 40mg/m . We found that children with Royal College of Physicians: Every breath we take: mothers in areas with an annual exposure of 35mg/m3 had about 5% less the lifelong impact of air pollution www.rcplondon. high air pollution lung volume than expected. The children who ac.uk/projects/outputs/every-breath-we-take- have smaller babies. were exposed to a higher level – 55mg/m3 – were lifelong-impact-air-pollution

26 March 2018, Volume 22(2) fst journal www.foundation.org.uk AIR QUALITY The debate t is vital to widen public awareness of the scale We need to ensure we give enough attention The discussion of the air pollution problem, the nature of its to reducing sources of pollution outside large that followed adverse impacts, and the part everyone plays urban areas, where the lack of public transport the formal Iin creating the problem and could play in solving leads to greater use of private transport and presentations it. Politicians need to be persuaded that action to financial constraints preclude measures to considered a deal with air quality should be given higher prior- increase the use of bicycles. wide range of ity. For that to happen there needs to be much Taxi drivers in London, who are heavily issues, including greater popular pressure (stimulated by profes- exposed to air pollution, are well aware of the public awareness, sional public relations activity and by the media) need to clean up London’s air. It is therefore political will, as well as firm political leadership of the kind regrettable that the minicab lobby has succeeded international being shown by the Mayor of London. in persuading the authorities to allow continued experiences and use of diesel-powered vehicles until 2023. ☐ forecasting. Comparative regulation It is useful to look at the different regulatory poli- cies and practices of other countries such as Japan (which, unlike the UK, has not incentivised the use of diesel power), Hamburg, Los Angeles and Madrid, as well as the Netherlands which has many similarities to the UK but a better air quali- ty record. We should do more to help cities in the developing world, such as Calcutta and Kathman- du, to avoid the mistakes made elsewhere. One cause of increased traffic in London and other urban centres could be the growth of jour- neys delivering online purchases. Possible solu- tions to this could be consolidation of drop-off points or the use of new technology to optimise delivery systems. Although some progress has been made

toward the development of nationally available SHUTTERSTOCK.COM / JGOLBY forecasts of air quality, more needs to be done. Taxi drivers in London are heavily exposed to air pollution

FURTHER INFORMATION

City of London Air Quality Strategy 2015-2020 Monitoring the implementation of London Plan energy policies www.cityoflondon.gov.uk/business/environmental-health/ in 2015 environmental-protection/air-quality/Documents/city-of-london- www.london.gov.uk/sites/default/files/2015_monitoring_ air-quality-strategy-2015.pdf report_-_final_nov_2016.pdf

Clearing the Air: The Mayor’s Air Quality Strategy Royal College of Physicians: Every breath we take: the lifelong www.london.gov.uk/sites/default/files/air_quality_strategy_ impact of air pollution v3.pdf www.rcplondon.ac.uk/projects/outputs/every-breath-we-take- lifelong-impact-air-pollution Hansard Report on Urgent Question on Air Quality Strategy hansard.parliament.uk/commons/2017-04-24/ Studies on the Impacts of Air Pollution on Human Health, the debates/4E0AFC5E-6B08-4A88-AE5E-AACC3F1C47D9/ Environment, and Global Economies AirQualityStrategy www.niehs.nih.gov/health/topics/agents/air-pollution

IARC: Outdoor air pollution a leading environmental cause of World Bank Report: The Cost of Air Pollution: Strengthening the cancer deaths Economic Case for Action www.iarc.fr/en/media-centre/iarcnews/pdf/pr221_E.pdf www.openknowledge.worldbank.org/handle/10986/25013

fst journal www.foundation.org.uk March 2018, Volume 22(2) 27 CANCER DIAGNOSTICS Can cancer be diagnosed earlier and if so what are the implications for patients and the health system? The challenge was debated at a meeting of the Foundation held at The Royal Society on 11 July 2017. Early diagnosis has the potential to transform patient outcomes Harpal Kumar

or most types of cancer, early diagnosis can greatly improve outcomes. For example, if SUMMARY bowel cancer can be detected early, the sur- Fvival rate is higher than 90%. The rate is much lower • Early diagnosis is vitally important if patient in the case of late detection, in fact less than 10%. outcomes are to be improved The story is similar for lung cancer – outcomes • Early diagnosis and treatment is much less are much better with early detection. Indeed, this expensive than late-stage intervention Sir Harpal Kumar is Chief is the case for most types of cancer: the earlier it is • The UK lags behind other countries in cancer Executive of Cancer diagnosed, the more we can improve outcomes in treatment resources Research UK. He worked terms of survival. • Early diagnosis technologies have their own for four years in the London On average, half of all cancers are diagnosed at challenges, including over-diagnosis office of McKinsey and Co, specialising in healthcare. stages 1 or 2 and the remainder at stages 3 or 4 • Shifting the focus to early detection will require a He then became CEO of although the picture varies for different types (see re-assessment of funding for different elements The Papworth Trust. He table, page 29). With breast cancer, somewhere of the health system. left Papworth in 1997 to between 80-90% of cases are diagnosed at an early become founding CEO of Nexan Group, a venture- stage: the UK is in line with the best countries capital-backed medical around the world on this. On the other hand, less For those who do take part in bowel screening, device company, creating than half of bowel cancers are diagnosed at an early there has been a significant impact on mortality – and marketing novel cardio- stage – which is behind many other countries – so 25% lower in those who have taken part in screen- respiratory monitoring technologies to reduce there is an evident opportunity for improvement, ing compared with those who have not. While premature mortality in even with no new technology development. there are undoubtedly a whole range of reasons patients with congestive On average, if a colon cancer is diagnosed at why screening uptake is not as high as it could be, heart failure. stage 1, the cost of the entire treatment journey the need to improve this picture is clear: it would averages just over £3,000. If, on the other hand, it make a significant difference to survival and mor- is diagnosed late that rises to more than £12,000. tality rates. So there is a real imperative, both in terms of out- Yet the challenge is not solely about screening. comes and in terms of NHS resource utilisation, to The proportion of cancer patients who are diag- diagnose earlier. nosed early compared with those that are diag- There are a number of ways to do this. First, nosed late varies across the country. In some areas, though, it is important to recognise that diagnosis as little as 44% are diagnosed late, whereas in other itself is complex with a whole range of inter-related parts it can be as high as 55%. That variation (11% factors that determine how quickly or late a patient or so) represents an enormous number of patients is diagnosed. There are different ways in which – about 40,000 a year. patients interact with their GPs or the health sys- The past several years have seen a steady tem, for example. In fact, there are a whole range increase in delays caused by the lack of diagnostic of system issues and a further set of challenges capacity. Figure 1 (page 29) compares radiology about the quality or otherwise of the tests available. resources globally – both the level of equipment and the number of trained radiologists. The UK Screening scores poorly in national comparisons of the There are three national screening programmes resources available for diagnosis. That of course in the UK, but bowel screening continues to have then translates into patients being diagnosed later. much lower take-up than breast or cervical The 2015 Cancer Strategy1 highlights some of ­programmes. Sadly, there has been a decline in the system issues and proposes a target of four ­cervical screening over the last several years. weeks between a patient presenting with symp-

28 March 2018, Volume 22(2) fst journal www.foundation.org.uk CANCER DIAGNOSTICS

If a patient presents Stages in cancer development with relatively non- Stage 1: the cancer is small and within the original organ. specific symptoms, they often go round Stage 2: the tumour is larger, but has not started to spread into the surrounding tissues. With and round the system some types of disease, cancer cells may have spread into lymph nodes close to the tumour. with several presentations to their Stage 3: the cancer is larger. It may have spread into surrounding tissues with cancer cells in GP and different nearby lymph nodes. diagnostic sciences. Stage 4: the cancer has spread from where it started to another body organ (also called secondary This is a waste of or metastatic cancer). resources as well as a waste of time and toms to actually getting a definitive yes/no diagno- are diagnosed. There are some very interesting often it results in a sis. There are a whole series of elements to exam- possibilities in terms of examining existing data later diagnosis. ine in the way the diagnostic pathway works. For sets as well the use of machine learning and artifi- example, if a patient presents with relatively cial intelligence to search for patterns that might non-specific symptoms – abdominal pain, head- enable earlier detection of cancers. aches, persistent coughs – they often go round and round the system with several presentations to False positives and negatives their GP and different diagnostic sciences. This is So those are all opportunities, but what are the a waste of resources as well as a waste of time and implications? One is the possibility, as indeed often it results in a later diagnosis. A number of with any test, of false positives and false negatives. organisations are examining the possibility of There has been a very large study carried out in changing some of those pathways, particularly the USA, looking at low-dose helical CT scanning where the symptoms might not be clear or obvi- for earlier detection of lung cancers. That study ous. Multi-Disciplinary Centres (MDCs) such as showed a 20% reduction in mortality. What is not those established in Denmark provide examples of generally talked about is the false positive rate in this approach. A number of these changes can be that study was 96%. made with no new technology and no new tests. As a result of that particular technology being Other possibilities include developing a better used in this way, large numbers of people under- understanding of those groups with the highest went quite significant interventions – in many risk of cancer and then monitoring and/or inter- cases bronchoscopy, but in some cases more vening more frequently. extensive investigations. Now those tests have One of the most exciting things about the pres- significant levels of morbidity associated with ent situation is the myriad possibilities of new them. So the headline numbers of a 20% reduction technologies, with both pre-clinical and clinical in mortality have to be considered in context. applications. These may facilitate a significant There is, in fact, a great deal of activity under- shift in the stage at which a whole series of patients way to improve the false positive rate through the

COMPUTED TOMOGRAPHY SCANNERS MAGNETIC RESONANCE IMAGING (MRI)

Per 1,000,000 inhabitants 2013 ● TOTAL Per 1,000,000 inhabitants 2013 ● TOTAL 40

60 35

50 30

40 25

20 30 15 20 10 10 5

0 0

ISRAEL CHILE SPAIN LATVIAKOREA CHILE MEXICO RUSSIA TURKEY ISRAELRUSSIA LATVIA SPAIN KOREA FRANCECANADA POLAND IRELANDESTONIA AUSTRIA POLAND CANADAFRANCE TURKEY AUSTRIAGREECE HUNGARY SLOVENIA LITHUANIA DENMARKAUSTRALIA SLOVENIA ESTONIA IRELAND HUNGARY LITHUANIA AUSTRALIA NETHERLANDS NEW ZEALAND UNITED STATES NEW ZEALANDNETHERLANDS CZECH REPUBLIC UNITED STATES UNITED KINGDOM SLOVAK REPUBLIC CZECH REPUBLIC UNITED KINGDOMSLOVAK REPUBLIC

Fewer trained radiologists Per 1,000,000 inhabitants 2013 Figure 1. 47 81 108 121 Comparisons of cancer detection UNITED KINGDOM GERMANY SWEDEN DENMARK resources fst journal www.foundation.org.uk March 2018, Volume 22(2) 29 CANCER DIAGNOSTICS

DEVELOPMENTS IN CANCER DIAGNOSTICS

Billy Boyle, Chief Executive Officer of advantage of being non-invasive, but the screening. There is also the important Owlstone Medical, outlined some of the tests can be more effective as well. question of resource efficiencies – if the developments in cancer diagnostics. His One in two people will get cancer. There number of CT scanners and radiologists company has developed what is essentially is still a very uneven picture in terms of is limited, how can the best use be made a breathalyser for disease: every time a outcomes for patients. In some cases, of these? person exhales, there are thousands of treatments work and in other cases The nature of the test really matters. chemicals on their breath. Some of these screening works. In addition, some of In colorectal cancer screening using a are biomarkers – of conditions from cancer the cancers are extremely deadly but the faecal-based test, up to 50% of people do to infectious disease. Chemical sensor incidence rate is relatively low. not turn up for the appointment and so risk technology can be programmed to detect Even relatively modest improvements in developing late-stage disease. As well as specific markers. the rate of early detection can have a really the choice of technologies themselves, While a person’s genome indicates significant impact on patient survival as there is the broader picture of acceptability the theoretical risk of disease it is the well as reducing the cost of treatments. to patients. downstream metabolism that provides the It is not a question of throwing out In his talk, Boyle agreed with the best real-time snapshot of actual disease existing strategies but of using technologies conclusion of the other speakers that the state. In regard to cancer, this has been in conjunction with each other. The breath field of early detection offers significant known for about 75 years. Cancer alters test, for instance, offers an opportunity benefits to society, both in terms of patient metabolism, even at the very early stage. to reduce the number of false positives outcomes and in the efficient use of scarce Using breath markers has the obvious when used in combination with existing health-service resources and funding.

addition of other tests that may be used as a triage progress, but as yet this is not clear. into a low-dose CT screening programme. There are significant implications for the health The contrasting possibility of false negatives is system. With new technologies coming along and in potentially very serious. If a technology has a rela- the light of the challenges to diagnostic capacity, more tively low sensitivity as indeed several do, cancers funds will have to be invested to get a higher propor- will be missed. If the patient goes away thinking tion of people diagnosed at an earlier stage. The cor- there is nothing wrong, it might prevent them from ollary is that it should be possible to shift investment coming back if their symptoms become worse. This from other aspects – so less resources and less money is another factor we need to take very, very serious- may need to be spent on treating late-stage disease. ly and why sensitivity is really important. What role will primary care play in the future? Some would argue for a bigger role and some for Saving lives smaller. If symptoms that are known to be red A third critically important area concerns over-di- flags for cancer can be identified earlier, it might agnosis, i.e. the possibility of detecting quite be possible to put in place a mechanism by which slow-growing cancers which might not cause a patients go straight to test, rather than via GP problem for the individual in their lifetime. The referral which introduces delay. Why waste a GP’s best current example is the breast screening pro- time if someone has had persistent rectal bleeding gramme. There is no doubt that this saves lives, for a number of weeks? between 1300-1500 a year in the UK. A recent study On the other hand, some of the technologies estimated that for every woman’s life saved through that are in current (or future) use in secondary and the programme around three are diagnosed with a tertiary care might be carried out in a community cancer that would not cause the woman a problem setting. A very large programme of work is now in her lifetime. Almost always, these women have underway to see if this is possible. So, the jury is out If the symptoms that treatment – sometimes quite extensive treatment. on future activity in the primary care setting. are known to be red We need to get much better at distinguishing The key message in all this is the real potential flags for cancer can cancers which are lethal and need treating from for transforming outcomes and getting much high- be identified earlier, those that are not lethal. At present we do not have er levels of survival than we have seen previously, as it might be possible the ability to do that. With some of the newer tech- well as better utilisation of resources. It is so much to put in place a nologies looking at circulating biomarkers, it can less expensive to treat a patient with early-stage can- mechanism by which be difficult to determine where the cancer arose in cer than to treat one with late-stage cancer. ☐ patients go straight the first place. Of course, these technologies are in to test, rather than an early stage of development and it is quite possi- 1. Cancer Strategy Implementation Plan. via GP referral ble that they will offer greater opportunities as they www.england.nhs.uk/cancer/strategy

30 March 2018, Volume 22(2) fst journal www.foundation.org.uk CANCER DIAGNOSTICS Using genetics to combat cancer Clare Turnbull

increasingly futile. Overall, between tumour evo- SUMMARY lution and cataclysmic molecular events, devel- oping measures that meaningfully impact on • Stratification using genetic risk can improve cancer control in the advanced setting is a real targeting resources on early detection and challenge. Therefore, it remains the case that ‘cur- prevention ing’ cancer still relies on timely physical removal • The majority of cancers occur in the minority of of the tumour. In the vast majority of tumour individuals who are at elevated genetic risk types this is achieved through surgery. Effective Dr Clare Turnbull is a • Identification of individuals who carry mutations surgery to cure the cancer is contingent upon get- professor in translational in high-penetrance cancer susceptibility genes is ting to that tumour when it is small and localised genomics at Queen Mary University London and a a priority; the impact of screening/preventative – and before the catastrophic molecular event has interventions is highest in these individuals senior researcher at the caused ‘bolting’ of the tumour. Institute of Cancer Research, • Individuals at high and moderate cancer risk can Considering this at population level, a ‘left- London. Her main research also be identified using risk modelling that wards shift’ of the average time of cancer diag­ focus has been on the combines genetic and non‑genetic factors. genetic (inherited) basis of nosis will also shift the stage at which cancers are testicular cancer and she has operated on, with commensurate improvement also published extensively in the proportion of tumours effectively cured on genetic susceptibility he majority of patients who present with by surgery. to breast, ovarian and childhood cancers. advanced cancer die from their cancer. In Alongside her research, she addition, they incur very high treatment Detection and prevention is the Clinical Lead for Cancer Tcosts for investigations, drugs and prolonged National screening programmes for breast, col- Genomics for the Genomics in-patient stays. There are targeted drugs that orectal and cervical cancer are already in place: England ‘100,000 Genomes Project’. Having trained as a have been developed to treat cancer which are behavioural research, public education and Clinical Geneticist, she sees typically used in the advanced setting. These evolved participant tools are vital to better under- patients with familial cancer drugs do not cure patients but may extend their stand and improve uptake. predisposition at Guys and St life. However, this may only be for a matter of a Coupled very closely with, but even better Thomas’ NHS Trust, London. Also an honorary consultant few months. As these drugs have very high devel- than, early detection is prevention. Preventative in Public Health Medicine, opment costs (in excess of $1 billion to bring a approaches may comprise modification of she works with Public Health drug to market), ‘per patient’ costs often come to ­lifestyle exposures (e.g. to sunlight or tobacco), England on the linkage of more than £10,000 per treatment round. chemo-prevention (e.g. aspirin to prevent molecular (genetic) data to National Cancer Registration ­colorectal cancer) or surgical prevention such as datasets. Tumours ­risk-reducing bilateral salpingo-oopherectomy Tumours are not homogeneous, they are clonal. (removal of the ovaries and fallopian tubes to They are made up of sub-clones of cells which ­prevent ovarian cancer). have different molecular characteristics. Thus a Expansion of vaccine programmes is another targetable mutation which the expensive drug is area of prevention that warrants attention. HPV designed to treat may only be present in a sub-set vaccinations save lives, but they are only admin- of the cancer cells. The disease itself is ruthlessly istered to girls. Yet this can prevent, in addition Darwinian – treating it with a drug, creating a to cervical cancer, other HPV-driven cancers selective pressure, will drive development of such as anal cancer or head-and-neck cancers. resistance mutations. The resistant clone has a So urgent evaluation of vaccination of boys is selective advantage which means that the clone also indicated. expands, predominates and the tumour contin- Then there is chemo-prevention – drugs that ues to progress. Hence the limited period of effec- cost pennies but which reduce the incidence of tiveness of targeted drugs in the advanced setting. cancers. Aspirin, statins, metformin – these are Mathematical modelling of cancers suggests that for many tumour types there may be a cata- strophic molecular event at some point in cancer Cancer itself is ruthlessly Darwinian – treating it development and after that there is aggressive, unbridled replication of the tumour with the nor- with a drug, creating a selective pressure, will drive mal mechanisms of cell defence becoming development of resistance mutations. fst journal www.foundation.org.uk March 2018, Volume 22(2) 31 CANCER DIAGNOSTICS

tro-intestinal tract and of the gynaecological 100% tract. However, he foolishly did not put his name 90% to the condition and in 1966 Professor Henry 80% Lynch documented a couple more such families 70% MLH1 while giving his name to the now eponymous MSH2 60% ‘Lynch Syndrome’. MSH6 50% PMS2 In the 1990s, four genes underpinning Lynch 40% Syndrome were identified (MLH1, MSH2, MSH6 30% and PMS2) and the biological mechanism caus- Likely incidence of cancer Likely incidence 20% ing this cancer susceptibility was identified as 10% defects in base excision repair pathways. Studies 0% of individuals and families carrying mutations in 20 30 40 50 60 70 Age these genes indicated a lifetime risk of 70-80% of developing cancer (and some of these cancers Figure 1. Lynch not being used systematically and that is another develop young). Syndrome: cancer missed opportunity. Figure 1 shows a curve accelerating from age susceptibility due Education is also needed about public health 30: by 40 there is a substantial accrued incidence to mutations in interventions and lifestyle change – earlier action of cancer. Figure 2 shows a typical Lynch Syn- base excision repair on obesity, smoking and other public health indices drome pedigree. There are two cases of colorectal pathways. will have substantial impact on cancer incidence. cancer – they died young. There is a brain tumour, an ovarian cancer, another colorectal cancer and Targeting an endometrial cancer. When considering screening and prevention, it is Until recently the patient who had come to my logical to target the populations most likely to genetics clinic might have been a woman on the have the cancer in question. The two filters nor- lowest line of the tree. She would be concerned mally used are age and gender: so the three cur- that her sister had recently been diagnosed with rent screening programmes (breast, cervical and bowel cancer at a young age and that there was a colorectal) use these basic demographic factors to dreadful family history of cancer. identify the people most likely to have prevalent I would contact the sister’s clinician and advise cancers. Could we use genetics and non-genetic referral of the sister to genetics. The genetics ser- factors to refine and better target screening and vice would proceed to identify the cause of genet- preventive measures to sub-populations who are ic mutation in her. They would write back to me, at a prior highest risk? 12/I would07/201 test7 my patient for it and institute mea- Genetic susceptibility to cancer has been rec- sures to address her high risk of cancer. ognised since Roman times when there was doc- Yet, all those other people had to die of cancer umentation on familial clustering of breast cancer. before we were able to ascertain the family risk. Figure 2. Dr Aldred Scott Warthin wrote a paper in 1913 Supposing genetic testing had been available to Lynch Syndrome: detailing the family of his seamstress in which he those earlier generations. The first to suffer from a typical pedigree. observed a marked excess of cancers of the gas- colorectal cancer could have been tested for muta- tion. That might in turn have changed the course of the history of this family: everyone would have their genetic test as they reached early adulthood and they would know if they were at high risk of cancer or if they were ‘off the hook’. Preventative surgery could have allowed them to live long, full, Colorectal Endometrial cancer 37 cancer 55 yrs cancer-free lives. My patient could come along yrs (d. 38) (d.56) knowing that she carries the causative mutation and have IVF embryo selection to make sure her child does not carry the mutation. Colorectal cancer Brain tumour 39 yrs Ovarian Cancer 50 yrs Historically, though, it was not feasible to 35, 45 yrs (d.46) (d.41) (d.54) deliver genetic testing routinely in clinical cancer care: ‘screening’ of genes for mutations was ardu- ous, low throughput, slow and relatively insensi- Colorectal tive. However, in the last decade there has been a cancer 40 yrs renaissance in technology. ‘Next generation sequencing’ has been a game-changer in identify- ing mutations in patients. Sequencing is now

32 March 2018, Volume 22(2) fst journal www.foundation.org.uk

Colorectal Endometrial cancer 37 cancer 55 yrs yrs (d. 38) (d.56)

Colorectal cancer Brain tumour 39 yrs Ovarian Cancer 50 yrs 35, 45 yrs (d.46) (d.41) (d.54)

Colorectal cancer 40 yrs

6 CANCER DIAGNOSTICS cheap with high throughput; it is very quick and very sensitive. It offers the possibility of expand- ing the populations and patients that are offered genetic testing. Knowing about Lynch Syndrome mutations is really useful for early detection and prevention of cancer. Those at risk can be given regular colo- noscopy from age 25, picking up these cancers while they are still polyps. They can then have preventative surgery before they develop cancer. Aspirin is really effective in these patients: it reduces the incidence of bowel cancer very sub- stantially (there is a dosing study underway being led by Professor Sir John Burn). There is also interesting work taking place on immuno-modulating drugs which appear prom- ising in terms of chemo-prevention and of reduc- ing cancer in Lynch Syndrome patients. This is the population in which we may see the first successful 3.0) BY (CC Bainscou non-virus-related cancer vaccine, based on the tional risk of developing the condition. All Next generation immune-genetics of how these tumours arise. women carry some of them, but how many sequencing offers depends upon the individual. ‘Polygenic risk pro- the possibility of Breast/ovarian cancer filing’, summing the totality of common variants expanding the In a similar way, if we can ascertain the mutation carried by a specific woman, can be used to esti- populations and causing hereditary breast or ovarian cancers early mate her genetic risk of breast cancer, thus identi- patients that are on, we change the natural history of the cancer pat- fying an additional subset of the population at offered genetic tern in the family. Once again genetic testing gives high risk. testing. us this knowledge, allowing us to catch cancer much earlier or even intervene before the cancer Non-genetic factors happens. Rather than dying of ovarian cancer, Non-genetic factors can also be added in to these women can have preventative surgery. Rath- improve the accuracy of the breast cancer risk pre- er than developing fatal breast cancer, a woman diction. These factors include family history, might have screening with MRI mammography to body mass index, age at first period, number of afford early detection or else prophylactic mastec- children, breast feeding and so forth. By including tomy to prevent the disease developing. the common variants and the breast density on Another option is chemo-prophylaxis: five first mammogram, a risk profile can be derived. years of tamoxifen reduces the incidence of breast Using this model (with 94 breast cancer suscepti- cancer in women by 30%. There is also promising bility variants included in the analysis), the top work around RANK ligand inhibitors for che- third in terms of risk contains two-thirds of the mo-prophylaxis and these are under trial in the cases of breast cancer. So by using this ‘non-genet- United States. ic and polygenic risk profiling’ model, breast If the patient does develop cancer, knowledge of screening could be focussed more closely. the underlying genetic cause means the cancer can The average risk of breast cancer in the total be managed accordingly, with selection of chemo- population of women being screened between therapeutic agents based on its underlying molec- ages 47 and 70 is 2.5% per ten years. This figure ular characteristics, so-called ‘precision oncology’. increases with age. However, the two bottom There is a spectrum of genetic factors that quintiles of risk (using our ‘non-genetic and poly- underpin each tumour type varying from genic risk profiling’ model) never even hit that ‘high-penetrance’ mutations in genes such as average. So, rather than screen on the basis of age BRCA1 to common genetic variants which each alone, this tripartite model (of non-genetic fac- confer a tiny additional increment of risk. For tors, combined with the 94 risk variants and breast nearly all tumour types studied, common suscep- tibility variants of this have been identified. For any given tumour type it is likely that hundreds, if If the patient does develop cancer, knowledge not thousands, of these variants exist. In breast cancer, there are over 170 common susceptibility of the underlying genetic cause means it can be variants published, each of which confers a addi- managed using precision oncology. fst journal www.foundation.org.uk March 2018, Volume 22(2) 33 CANCER DIAGNOSTICS

density) could more accurately target women at target our resources for screening, early detection genuine elevated risk of breast cancer. It is esti- and prevention. Colorectal cancer is an exemplar mated that such an approach could reduce the cancer in which screening is highly effective: number of women being screened by 31% while approaches to better target screening based on lowering detection by just 2%. The resources prior risk would be strongly impactful in improv- saved could be redirected to offer more screening ing outcomes. Furthermore, there is massive to those at really high risk. potential gain in cancer outcomes and survival Combining these two approaches, (i) identify- from improvements in population-level uptake of ing individuals carrying high risk mutations and screening in combination with more systematic (ii) applying ‘non-genetic and polygenic risk pro- administration of aspirin, a drug of proven che- filing’ risk modelling will enable us to identify mo-preventative efficacy for this (and potentially individuals at elevated risk of cancer and better other) tumour types. ☐ The debate A range of equencing tumour material is a complex championed for some time) would provide rapid points were put process and the number of mutations that access to diagnostic assessment for individuals forward by the can currently be targeted by drugs is limited. with non-specific symptoms. This would reduce invited audience However,S as research on the whole genome devel- multiple, separate and consecutive testing for such after the formal ops there is clear potential for progress. There are individuals and support earlier diagnosis. Five presentations, promising signs that targeted therapy can work at pilot centres have been established and the results including: gene a molecular level. In addition, new technologies should known in the coming months. therapies; combining effective monitoring with combina- There is no link between preventive surgery to reducing the time tions of therapies to treat re-occurrence offer cure inherited cancer in an individual and subse- for diagnosis; promise for the future. quent inheritability. Reproductive interventions resource on the other hand could guarantee that high risk constraints; and Private sector role mutations were not passed on; but IVF is expen- access to capital. Does the private sector have a role to play in sive and carries other risks. risk-profiling for health? There may be scope Radiology and radiography are crucial to can- for collaboration with companies involved in cer diagnosis and treatment, but current services gathering genetic information at the population are over-stretched and under-resourced. Work- level, but this is very unlikely to be directly helpful force and technology issues have to be addressed. to an individual. Artificial Intelligence (AI) and machine learning The introduction of ‘one-stop-shops’ on the have real potential for interpreting data, in genetic Danish model (which Cancer Research UK has screening and to support radiology (Singapore is already using AI effectively). Access to capital for innovations in this field is FURTHER INFORMATION recognised as a limiting factor – and not just in the UK. However, the NHS, arguably, sets a Diagnosing Cancer Earlier: Evidence for a National Awareness and Early ­higher bar for wide-scale adoption than, for Diagnosis Initiative. example, the healthcare system in the USA, www.cancerresearchuk.org/prod_consump/groups/cr_common/@nre/@hea/ because of a requirement for value across the documents/generalcontent/ cr_044142.pdf whole chain – from screening through testing to treatment. However, momentum on prioritising Annual Report of the Chief Medical Officer 2016: Generation Genome. research and funding for early detection is www.gov.uk/government/publications/chief-medical-officer-annual-report- unquestionably building. The Chief Medical 2016-generation-genome Officer has, for example, prioritised the issue in her 2016 annual report. Cancer Research UK. www.cancerresearchuk.org More collaboration is needed to develop bigger data sets to support risk stratification as well as bet- Genomics England. www.genomicsengland.co.uk ter targeting. These developments could reduce the burden on the health system and improve clin- Owlstone Medical. www.owlstonemedical.com ical effectiveness. Longitudinal research studies would be particularly important. ☐

34 March 2018, Volume 22(2) fst journal www.foundation.org.uk MACHINE LEARNING Can we control the way machine learning is steering the course of our everyday lives, ensuring that we maximise the benefits while minimising potential risks? The issue was discussed at a meeting of the Foundation held at The Royal Society on 14 November 2017.

An opportunity or a threat to society? Mike Lynch

have very little idea – you will just know French. SUMMARY The situation is similar with machine learning. Rather than trying to programme a computer, • Machine learning will have a profound impact on modern machine learning is based on learning by society example. Before modern machine learning, many • Machine learning is based on learning from problems were not solvable by machines. A simple examples example might be to ask a computer what is in an • The fundamentals of machine learning can be arbitrary photograph: no amount of programming Dr Mike Lynch OBE FRS quite counter-intuitive will get very far with that, whereas machine learn- FREng DL is founder of • The power of machine learning comes from the ing systems will give a pretty good assessment. Invoke Capital and Deputy technology’s ability to adapt to the variation we Problems where performance was getting better Chair of the Foundation for see in the world Science and Technology. little by little, such as voice recognition, have recent- Invoke Capital invests in • We need to absorb the implications of living in a ly seen significant leaps in a very short amount of fundamental technologies. probabilistic world. time. A recent set of papers claimed speech recog- Dr Lynch is a member of the nition by machines is more accurate than humans. Prime Minister’s Council on Science and Technology and If true, that is a very significant milestone. is a Business Ambassador. he easiest way to think about the promise Other areas which have seen advances are He is best known for founding of machine learning is to compare it with computer vision and machine translation. Autonomy in 1996, the UK’s a fairy-tale genie. It can grant wonderful largest software company, a member of the FTSE 100 Twishes but you have to be very, very careful what Societal impact with a market capitalisation you ask for: it will deliver precisely that, which In the early days, this subject was only discussed of $11 billion. He was its may not be what you intended. by people working in this area. Now, it seems you CEO for 15 years until it was Machine learning will have a profound impact. cannot open a newspaper without seeing some- acquired by Hewlett Packard in 2011. The industrial revolution replaced muscle, while thing on it. the IT revolution which started at the end of the The difficulty is that much of the commentary 1960s replaced simple repetitive mental tasks like does not understand the fundamentals of machine calculating the payroll ledger. Today there is a new learning which are actually quite counter-intui- revolution which is replacing more advanced cog- tive. The danger is that ill-informed speculation nitive tasks, including many that people under- could lead to regulation which might not work. take at the moment. Machines do not get bored, There are three main themes in much of the they can often deal with more information, they coverage. The first, coming even from quite high- can now out-perform humans in some key areas ly-educated people, concerns the fear of machines in terms of accuracy. taking over and killing humans. Those working in I started working on machine learning in the the field know that there is very little to fear at the late 1980s. The easiest way to understand it is to moment. Putting aside the point that we may have compare it with learning French. There are two far more to fear from evil people with machines, methods: either sit in a classroom as a student with rather than evil machines themselves, it actually a grammar book, or go and live in Paris as an infant. shows a misunderstanding of where the subject is The latter tends to give better results. You will learn at this point in time. the realities and the idioms, as well as slang. When Recent advances have been in what has been you are five and someone asks why you have con- called ‘narrow AI’. This is the ability to resolve a structed a sentence in a particular way, you will narrowly-defined problem like recognising what fst journal www.foundation.org.uk March 2018, Volume 22(2) 35 MACHINE LEARNING

A machine can probably be as accurate as a fox in The skill sets required will be quite specific. determining whether there is a rabbit in your Obviously there will be some people with the mathematical processing skills that go into mod- garden. But try and build a fox with current AI and ern machine learning, but getting the best out of it would get run over by the night bus the algorithm may involve coupling technical abilities with more human skills. The biggest challenge concerns exception-pro- is in a photograph. There is, on the other hand, a cessing. There is a wonderful road along the whole class of problems which are ‘broad AI’: to Amalfi Coast in Italy where buses regularly meet give just a single, simple example, a machine can head-on. It seems physically impossible for them probably be as accurate as a fox in determining to pass, but by a process involving a barrage of whether there is a rabbit in your garden. But try medieval Italian, they somehow do! There is no and build a fox with current AI and it would get way a machine-learning algorithm is going to run over by the night bus because it would not manage that, so the arrival of autonomous vehi- be able to deal with the generalities that a real cles in that part of Italy is like to lead to gridlock! animal encounters. Problem-solving is not the issue: exception-pro- It is easy to assume from the progress of narrow cessing is crucial in these situations. AI that something similar will happen on broad AI. Yet there is no scientific basis for this. Data I believe that there will be a significant impact Machine learning needs data. Whoever gets the on jobs, but perhaps more important will be the data first produces the best system. They can requirement for skills and adaptability within the then get more data and probably end up with a workforce. Some of the knowledge-based tasks in monopoly. This is a question about strategic data society will be automated, but wisdom-based rather than open data. tasks are less likely to be affected. If we want the UK to be strong in this field, how Then there is the economic aspect. These tech- are we using our data? If we are providing data, nologies are going to affect pretty much every area what do we get in return? Does the NHS get back of commerce and they are going to have signifi- products at competitive prices because its data cant effects on existing industries. was used to create them? Mr Putin claims that these technologies will Reality and hype create domination in the future, so in that sense I often read about people who over-clean their data could be strategic in a true geo-political data, believing that everything must be cleaned sense. Cyber attacks are starting to be based on before it goes into a machine-learning algorithm. machine learning, which is making them a lot The power of these algorithms actually comes more difficult to counter. from their ability to deal with the variation in the In terms of economics, there will probably be real world, so the last thing you want to do is take over-investment in this area – and therefore some out that variation – that is the real benefit. disappointment. A skills-base will have to devel- One thing that will happen might be termed a op and there will be big effects on other industries. ‘hype curve’. Already, almost every start-up ask- When machine learning techniques are applied, ing for funding claims to be doing machine learn- imperfect pricing structures in markets are quick- ing. Very few are actually doing so which means ly exposed and so margins in many industries will that a lot of money will be wasted. Investors do come under pressure. not understand the difference between a demon- stration model and a useable system. Societal challenges Demonstrating something in the lab with Machine learning has great promise for some of machine learning is pretty much useless. Doing it the great challenges facing society. For example, at the side of the road in Peckham, dealing with all it may assist early-stage dementia patients to the real-world variation is what is important. remain at home for longer, adjusting for some of Again we are not seeing people that really under- the variations in cognitive level. Another appli- stand those differences. cation might be better optimisation of traffic flows, maximising the use of existing transport infrastructure, rather than commissioning a new Some of the knowledge-based tasks in society piece of motorway. The airline industry functions because planes will be automated, but wisdom-based tasks are can be insured, even though one of them might less likely to be affected. go down over central London (and that risk anal-

36 March 2018, Volume 22(2) fst journal www.foundation.org.uk MACHINE LEARNING ysis is an interesting one given there are no prece- information is probably against the law. How can dents). There is plenty of evidence about the risk society make sure that decisions are being made profile of human car drivers, but none for auton- on the right basis yet take into account cultural and omous vehicles even if it is believed they are legal acceptability? intrinsically safer. There are issues about derived We need to understand that we live in a proba- works: if my data is used in an algorithm, do I own bilistic world. Really good drivers crash cars – and the algorithm? so will really good autonomous vehicle algo- People may get very excited about CCTV sur- rithms. Is the algorithm at fault? veillance, but if no one is looking at the cameras it We will certainly go through a period of hype. does not really matter. With machine learning, There will be many machine learning systems that though, all the cameras will be monitored and the will not work. There will be lots of promises that data analysed, so many of our current concerns are not delivered, but underneath all of that there will become much worse. is a real revolution going on. I think we will see If an algorithm calculates that Ruritanians are society’s time-honoured approach of under-esti- more likely to default on their mortgages, is that mating how long the change will take, while also acceptable or fair? At present, using that kind of underestimating how deep the effect will be. ☐ Giving society the confidence to embrace the opportunities Claire Craig

adverse effects that result in a loss of public confi- SUMMARY dence and slow progress. The Royal Society published a report on • Machine learning and AI are about much more machine learning in 20171. Led by Peter Donnelly, than robots replacing humans, presenting a Professor of Statistical Science at Oxford, there massive opportunity to promote human were 15 scientists from academia and business on flourishing the working group. Around 500 practitioners • The challenge is to deliver benefits quickly, safely were involved in workshops. The project looked Dr Claire Craig CBE is Director and sustainably at the effects of machine learning in sectors such of Science Policy at The Royal • Governance should not focus on technology, but as pharmaceuticals, law and manufacturing. Society. Previously, she led the Government Office for on its applications, and should be based on Extensive public engagement and public sectoral governance Science which supports ­dialogue were undertaken. Around 15,000 the Government’s Chief • Existing frameworks governing the use of data ­people participated in a range of ways, with Pro- Scientific Adviser (GCSA). will not work in this new and developing fessor Brian Cox and an expert panel attracting a She originally joined the Civil landscape Service to run Foresight, capacity audience at the Royal Festival Hall. a programme of science- • A new stewardship body, with oversight across The report’s conclusions included five areas based strategic futures the whole range of applications, is needed. for action: projects, covering topics 1. Data – the creation of a suitable data environ- from flood risk to cognitive enhancement. Her career ment, through incentives and funding, to includes periods at McKinsey achine learning offers massive oppor- enable not just open or accessible data, but & Co and the Confederation tunities to promote human flourishing, machine-readable data. This will require data of British Industry. in fields ranging from health and home curation and the training of specialists with the Mefficiency to the conduct of science itself. The UK necessary skills, as well as open standards for is well placed to lead in the development of this data. technology. It has an excellent science base in both 2. Business – embedding machine learning academia and business. It also has a long history of in industrial strategy. This will include US successful governance of new technologies, as well ­‘DARPA-style’ funding, as well as support and as a more recent history of promoting open data. advice for businesses via Growth Hubs, for The challenge today is to deliver benefits example. quickly, safely and sustainably, without any 3. Skills – ranging across: achieving digital liter- fst journal www.foundation.org.uk March 2018, Volume 22(2) 37 MACHINE LEARNING

Figure 1. Perception of value/risk in machine learning applications. Self-driving Source: Ipsos MORI. cars Machines in Social risk Social the home

Personalised learning

Marketing/financial products Predictive pricing Art Health diagnoses

Social value Source: Ipsos Mori

acy in the primary school curriculum; embed- They also carried out structured public dia- ding the importance of machine learning with- logue. The results showed that people had broadly in data science, mathematics and other sub- four typical initial reactions, ranging from ‘I can jects through to age 18; and specific support for relate to this technology’, through suspicion, to a advanced courses such as Masters. The Royal denial that it could be made to work. Interestingly, Society is starting further work on the barriers we repeated the work with young people who have and enablers to keep the UK at the leading edge grown up immersed in digital technologies. They of research. had much more rapid and intuitive understanding of what machine learning might do and mean, but These first three areas are addressed in the AI their interests and concerns were fundamentally Review (see page 41), but there are two others: the same as the previous participants. 4. Engagement with society As the dialogue deepened, and people thought 5. The next wave of research through potential applications, their aspirations and concerns varied by context. Figure 1 shows a Engaging with society schematic representation of a discussion. Note that The 2014 annual report of the UK Government’s the axes are not quantitative, the point is simply that Chief Scientific Adviser2 focussed on innovation in terms of risk and benefit the different applica- and risk. The concerns citizens have about tions are seen very differently. machine learning are typical of previous emerg- This public dialogue alongside the expert work ing technologies such as stem cells or nanotech- led the report to two conclusions. The first is that nologies. existing frameworks governing the use of data will People want to understand a technology and the not be sufficient in the future. The second is that it purpose behind its development. They want to would be wrong to govern this as a technology per know how the benefits and risks will be distributed, se. There are many reasons for this, one being the who pays, what the alternatives are, who they can impracticality of trying, in effect, to regulate maths. trust and so on. As a technology is used and But perhaps more importantly, in many (or most) applied, these questions become more context-spe- contexts, machine learning is generally uncontro- cific. It is no longer the technology itself, but its versial and does not need new governance. The benefits and context-specific applications on which judgements The benefits and risks of different applications risks of different are made. vary widely: a major health or financial decision is applications vary Ipsos Mori carried out a quantitative survey different from a customer recommendation. Issues widely: a major which showed, among other things, that only 9% of around safety and proper testing in transport appli- health or financial respondents had heard the term ‘machine learning’, cations are likely to be better handled by existing decision is different although most were familiar with applications that bodies in that sector. Similarly, questions about the from a customer use it, such as predictive text or consumer recom- validation of machine learning in medical applica- recommendation. mendation systems. tions should be overseen by existing medical regu-

38 March 2018, Volume 22(2) fst journal www.foundation.org.uk MACHINE LEARNING latory bodies, while applications in personal deliver a clear set of functions – but with the specif- There are areas of finance would fall within the purview of the finan- ic form left deliberately open. Its functions would research that will help cial regulators. include ensuring that emerging gaps in governance ensure machine To take one specific example, views on the use are identified and addressed, and issues are antici- learning can promote of machine learning in diagnostic systems in GPs’ pated wherever possible. It would conduct inclu- human flourishing. surgeries were influenced by a range of assump- sive dialogue and expert investigation into new tions, including how such systems would actually questions and issues, attempt to anticipate future be put to use, how they would be tested for safety consequences of today’s decisions, and spread and accuracy, which patient groups would have learning from one sector to another. access to them if they were beneficial, and what the The Nuffield Foundation has since announced alternatives or consequences might be. People that it will work with others to create a new Con- might, for instance, be happy to have a machine vention addressing the longer term ethical and learning diagnosis where that was shown in some governance implications of new uses of data. way to be better than a human one, but they might still want a human ‘in the loop’ to discuss major The next wave of research decisions affecting them personally. There are areas of research that will help ensure machine learning can promote human flourish- Data management and use ing. They include urgent questions of verification The Royal Society had already begun work on the and validation, as well as security and robustness. wider governance issues of proliferating data and In addition, work is needed on ways to create its uses, including – but not solely – machine ­systems that can be better interrogated by users, learning. In 2017 it published, jointly with the so that they can used more effectively in public or British Academy, a report on governance in the private decision-making, and result in better 21st century3. Led by Ottoline Leyser and human-computer interfaces. Machine learning Genevra Richardson, the project brought togeth- is, in a real sense, creating its own research agenda er historians, philosophers and other scholars about the things we need to know in order for it to with cross-over membership to the Machine fulfil its potential safely and rapidly. Learning group. The impact of machine learning will vary This report argued that, in order to have a prin- depending upon the sector. Timing will also be cipled and connected set of discussions about the important, especially for SMEs looking to exploit future of governance, while at the same time accept- opportunities. ing that individual governance decisions need to be Machine learning, and AI more generally, is taken in context, two new elements were needed. likely to have significant implications for employ- First, very high level principles for governance of ment, both threats and opportunities. The profes- data use, and second, a stewardship function oper- sions – such as the law, medicine and accountancy ating across the governance landscape as a whole. – will have to address these challenges to their Because existing words come with 20th century career structures, and to their attitudes about con- connotations, the report recommends the introduc- tinuing professional development (CPD). Experi- tion of an overarching principle ‘to promote human ences in the transport sector illustrate issues that flourishing’ (the term is deliberately not defined, nor have broader relevance: questions of accountabili- is it capable of being measured). The purpose is, at ty, liability, and the ‘public licence to act’. least in part, to force people to consider what flour- A fundamental issue is the need to get beyond ishing might mean in different contexts. New uses thinking only in simple terms of robots replacing of data may create new trade-offs between individ- humans. It is much more complex than that: ual and collective benefits and risks. machine learning and AI have the potential to A governance system has a range of critical change business models, create new opportunities, functions, from anticipation to enforcement. and result in structures that augment or collaborate While many of these are already being carried out, with humans. ☐ there is no-one with responsibility for oversight of the entire landscape. Thus, the system risks losing 1. https://royalsociety.org/~/media/policy/ transferable learning, falling behind the ways that projects/machine-learning/publications/machine- the technology itself is interconnecting. For exam- learning-report.pdf ple, definitions of health data may not be meaning- 2. www.gov.uk/government/uploads/system/ ful when health facts can be deduced from retail or uploads/attachment_data/file/381905/14- transport information. 1190a-innovation-managing-risk-report.pdf The report therefore recommended the intro- 3. https://royalsociety.org/~/media/policy/ duction of a new stewardship body to oversee and projects/data-governance/data-management- fst journal www.foundation.org.uk March 2018, Volume 22(2) 39 MACHINE LEARNING The potential to augment human efforts Amir Saffari

he Artificial Intelligence journey began in the 1950s. It was thought to be a very easy SUMMARY technology to develop and, famously, some professorsT thought the challenge would be solved • Rediscovery of neural networks at the beginning within three months – well, it has been a very long of this decade has accelerated the progress of AI three months! Then came the 70s and 80s, what • Speech, computer vision, and natural language was known as the ‘AI winter’, when funding was cut processing are widely used applications today Amir Saffari is Director of and development stagnated. But things picked up • Increasingly, machine learning is being used to Applied AI at BenevolentAI. again in the 90s and on into the new century. accelerate scientific research in areas like drug He has been working in the About seven years ago, researchers rediscov- discovery field of Artificial Intelligence ered neural networks. These were rebranded as • AI has a large role to play in education, making it for more than 15 years, more interactive and personalised researching and developing ‘deep learning’ and led to massive improvements in machine learning (ML) what machines were capable of doing. At the same • In the near future, AI has enormous potential for theory, applications and time, large data sets and high-powered graphics augmenting human activity. industrial products. He was processing units (GPUs) started to be used. Most part of Sony’s Computer Vision R&D team working of the current public and media interest has result- on ML technology for ed from that renewed focus on neural networks. people over-estimate the capabilities. Then sud- Augmented Reality and In 2007, Image Net was developed to classify denly a change happens after which people actu- Virtual Reality applications. images into about 1,000 categories. By 2010, the ally underestimate the capability. The public’s He joined BenevolentAI in early 2017 where he heads best systems had an error rate of about 25% which image of AI – given the fuzzy language – probably the AI team, focussing on was not good enough for any serious applications. comes from science fiction, with killer robots and creating machine learning By comparison, humans have a 5% error rate on bad things happening to people. Current media technologies to accelerate this type of task. Even for humans, the number of coverage does not help. scientific discovery. different categories makes such detailed classifi- People typically fall into one of two miscon- cation a very challenging task. ceptions about AI. Either they think that AI is Fast-forward to 2017 and neural networks can here to take over the world and we are doomed, or achieve a 2.5% error on these datasets. That makes else AI is all hype, all smoke and mirrors. The it possible to build applications, solid applications. truth lies somewhere in between. Some people Combine these developments with others in com- also think that progress in AI will lead to more puter vision systems and the possibility of automation, more unemployment and our self-driving cars looks much more realisable. replacement by soulless machines. This, too, is Other applications include speech recognition, not true. In terms of full automation, we are still taking a sound wave and transcribing the lan- far away from being replaced by robots. However, guage, or using machine translation from one lan- the technology may soon allow humans to work guage to another – in fact, any mix-and-match of collaboratively with software – with AI – and help data modalities. In image captioning, it is possible us to become better at what we do. to take an image and ask the system to describe the content. There is good progress being made on AI augmentation the accuracy of such tasks. This in turn opens up BenevolentAI is an artificial intelligence compa- new applications in industry and elsewhere. ny based in London. It specialises in using AI for With all the successes inevitably comes the advancing scientific discovery. For the past three hype. It often happens with new technologies that years the focus has been on drug discovery. The rationale is this: there is one research paper ­published every 30 seconds, so a scientist would The public’s image of AI – given the fuzzy language only be able to digest a fraction of the amount of knowledge available. This is exactly where – probably comes from science fiction, with killer machine learning can help because it can read all robots and bad stuff happening to people. this material. Machine learning can turn that data

40 March 2018, Volume 22(2) fst journal www.foundation.org.uk MACHINE LEARNING into knowledge that our scientists can use: we machines working together, in this case to com- have discovery scientists in the company who can pose music. The majority of people would like to take this knowledge and accelerate their discov- make music if they could, so by lowering the bar- ery. Potentially, this can help find treatments for rier to entry, more people can be creative – and very difficult diseases more quickly than would that is a very good thing. otherwise be possible. AI also has a role in education – and specifical- One area of machine learning that is progress- ly personalised education, making it interactive ing rapidly is that of generative models. These and fun to learn. One example is Sonic Pi, from the techniques are now capable of creating images University of Cambridge Computer Laboratory. that look very authentic. Content creators and The software, created by Sam Aaron, is very inter- artists can interact with these systems to become active and he uses it to teach very small children better or faster. This extends to other areas such as programming. Because it is interactive and fun, generative design – where machine learning and people learn programming much faster than they generative algorithms produce better designs. could otherwise. I have been involved in a project over the past AI has enormous potential in terms of aug- year applying machine learning to music. This menting us and helping us to become better. There has led to the creation of a model that can com- is a great deal published about AI but some of the pose polyphonic music. We linked it to Twitter: good news stories often get overlooked. People tweeting at it generates a small music audio clip. often are not aware of the way in which machine This is another way of exploring humans and learning could be used for the good of society. ☐ The opportunities for the UK Wendy Hall

sia is also pursuing this technology. And it is not SUMMARY just the larger powers that are showing interest: Singapore, a small nation, is developing an ambi- • Realising the opportunities arising from AI is a tious AI strategy as are many others. priority for governments around the world The scope of our review was closely-drawn: it • Access to data is a critical enabler for small was to look at the opportunities for the UK. That businesses and researchers in this area included job creation and economic growth. It • Higher Education courses to help people acquire was envisaged that this would lead the way to one Professor Dame Wendy Hall AI skills will help to extend the technology across of the Sector Deals set within the Industrial Strat- DBE FRS FREng is Regius industry egy. That is very exciting because of course there Professor of Computer • It is important to build confidence across society is no easily-defined AI sector. It is quite ambitious Science at the University about AI of Southampton. She of the Government to consider it in those terms. was Dean of the Faculty • The UK computing industry must make AI is an emerging technology, not a traditional, of Physical Science and sustained efforts to increase diversity in what established industry sector like automotive, aero- Engineering from 2010 remains an overwhelmingly male environment. space or pharmaceuticals. to 2014. One of the first computer scientists In the four months we had to complete the to undertake serious review, it was not possible to cover everything. research in multimedia co-chaired with Jerome Presenti, the Chief However, others were also working on this sub- and hypermedia, she has Executive of BenevolentAI, a review on ject. The Royal Society and the British Academy been at its forefront ever 1 since. The influence of her Growing the AI industry in the UK for the were developing their Data Governance report. I work has been significant in IGovernment. We took some inspiration from do not think we could have produced our review many areas including digital what was happening in other countries. In the without knowing that these other efforts were libraries, the development USA, the Obama administration produced two underway: our scope was very firmly on job cre- of the Semantic Web, and the emerging research very good reports on AI, addressing threats, job ation and economic growth. There are huge discipline of Web Science. losses and opportunities. Canada has set up a issues about ethics and accountability that we did She was co-chair of the UK strategy for AI and there are programmes in not have time to consider but they are well cov- Government’s AI Review Europe. China has said it is aiming to be the lead- ered by the academies. which was published in October 2017. ing country in AI in the next decade, something The vital thing is to create a level playing field. they stand a very good chance of achieving. Rus- For small companies to grow they have must have fst journal www.foundation.org.uk March 2018, Volume 22(2) 41 MACHINE LEARNING

need people with skills. They need undergraduates FURTHER INFORMATION studying this subject and they need PhD students with machine learning skills, for example. AlphaGo - Silver D et al, Nature 550, 354–359 (19 October 2017) The report puts forward the idea of indus- doi:10.1038/nature24270 try-funded Masters courses, which take students www.nature.com/articles/nature24270 with a maths, computer science or statistics back- ground and turn them into machine learning pro- BEIS Press Release referencing funding for AI grammers. A 15-month course is proposed, the www.gov.uk/government/news/record-boost-to-rd-and-new-transport-fund- normal 12-month university course and then a to-help-build-economy-fit-for-the-future 3-month internship at the company funding the studentship. Hall and Pesenti Report to the Government on AI More generally, there is a need to up-skill the www.gov.uk/government/publications/growing-the-artificial-intelligence- whole of industry in terms of AI (and the Govern- industry-in-the-uk ment service too). That does not mean everyone has to be a machine learning programmer, but a Machine Learning information from The Royal Society website ‘conversion Masters’ could train people with all https://royalsociety.org/topics-policy/projects/machine-learning/videos- sorts of different disciplinary backgrounds to and-background-information become cognisant with AI or even to be able to take AI to their own industry. Machine Learning Education The economy needs a lot more PhD students if www.coursera.org/learn/machine-learning it is to have the skilled researchers needed for AI to grow in this country. Online courses and CPD Nesta on the fourth industrial revolution have an important part to play here, if whole sec- www.nesta.org.uk/blog/how-can-fourth-industrial-revolution-be-made-good? tions of society are not to lose out as we move to greater automation. The Royal Society project on machine learning We also talked about the development of fel- https://royalsociety.org/topics-policy/projects/machine-learning lowships to be offered through the Alan Turing Institute to attract the best researchers into the The Royal Society and British Academy Report on data governance UK and to retain the ones we already have here. https://royalsociety.org/~/media/policy/projects/data-governance/data- Of course, every developed country in the world management-governance.pdf is trying to do that, so there is major competition. At the moment, a graduate leaving university White House Report on AI with machine learning skills can almost name their https://obamawhitehouse.archives.gov/blog/2016/10/12/administrations- price, I have been told. It is not clear how long that report-future-artificial-intelligence will last, but there is a great scarcity of people with the right skills at present. Of course, it is not possi- MATLAB ble just to train hundreds more PhD students: they https://uk.mathworks.com/campaigns/products/trials/targeted/mal. need supervisors in the universities and these peo- html?s_eid=ppc_51241087401&q=matlab%20machine%20learning ple too are in danger of being poached. It is very hard to keep academics who are paid a university salary when they have the skills to work in industry. access to the data they need to create the algo- To maximise AI research in the UK, there are rithms and test that they work. plans to turn the Alan Turing Institute into the We developed the idea of data trusts. These are national institute for artificial intelligence as well places to get legal advice on how to talk to big as for data science. It ought to be easier for indus- companies, local councils, universities and try to license IP from universities. Small compa- research labs about access to data, as well as the nies and research labs should have easier access to ways in which this data can be used. Then, at the improved computer capacity, either to cloud ser- end, who owns what? Who owns the rights to vices or to high performance computer services. what deliverables? The report makes a number of recommenda- This is really complicated. There is no one tions to support the uptake of AI in the UK. It answer that fits everything. So the report address- recommends the setting up of an AI Council to es this challenge and it proposes the setting up of take proposals forward and to look at account- a data trust support organisation. ability issues. UK AI skills could be exported A large section of the report is concerned with through the Department for International Trade. skills. One message that was conveyed very clearly The report also proposes developing practical by the companies in the AI industry was that they guidance for UK companies who want to use AI.

42 March 2018, Volume 22(2) fst journal www.foundation.org.uk MACHINE LEARNING

We hope the Government will be an early adopter, It proposes coordinated, on-going action which both in terms of encouraging AI skills and also of has to start in schools, reaching girls in particular developing challenge funds that have AI as their and encouraging them to go into fields that will core purpose. lead to a career in AI – and it has to be recognised There are major ethical issues surrounding AI. that this will take many years to achieve. It is important that society has trust in what the If action is not taken, the AI industry will be industry is doing. Companies are in the business of one where all the algorithms are written by one making money, but at the same time they must be half of society and no matter how much you try, accountable for the algorithms they develop. Some those algorithms will be biased. ‘Bias in and bias aspects of the technology will need to be Govern- out’ is how it is with algorithms and the datasets ment-regulated and some will need to be self-reg- used to train them. ulated. Personally, I believe companies should state This has to be addressed. One way to do so is on their websites the principles under which they to require the involvement of interdisciplinary are using data and developing their algorithms. teams in the development of the algorithms from They could be asked to publish this informa- the very beginning; something similar to health tion as part of their annual reports. It is a matter service practice. This helps ensure an holistic of building public confidence. Leading on the dis- view of the project, reducing bias as much as pos- cussion in these areas will be the role of the new sible. It should be remembered that gender is not Centre for Data Ethics and Innovation that is the only bias that can creep in: we should also con- being set up in parallel with the AI Council. sider age, race, ethnicity, culture and many others. I have been in the computing world for over 30 Bias in all its form has to be tackled head-on. years and the number of women in this industry As a society, we also have to decide how we are has actually decreased over that time. A short- going to treat the people who are going to lose term stimulus to introduce AI skills into industry their jobs but cannot be retrained into this type of will have to rely on the computer scientists, math- work. What does a ‘welfare state’ involve in a ematicians, etc, leaving university today. There world with AI? ☐ will be hardly any women, so we are producing a workforce which will not be diverse, by definition. 1. www.gov.uk/government/publications/growing- Diversity is, though, a key feature of the review. the-artificial-intelligence-industry-in-the-uk The debate ow many jobs will be lost or affected by by the World Health Organisation. The deploy- After the main machine learning? Broad estimates ment of healthcare advances to patients could be presentations, have been considered by the Council for widened through the widespread use of AI-driv- the audience was HScience & Technology, but none can be regarded en apps on phones. These would make good qual- invited to give as definitive. The impact on employment may be ity medical advice more readily available to com- their views. Topics most significant for those aged 40-60, so the munities lacking access to doctors. raised included: potential to retrain in this age group should not be Mathematics (including probability) will be the impact on neglected. key to the pace of further progress. However, this employment; If NHS data is used in new ways, the NHS itself needs to be coupled with an understanding of benefits to should receive a benefit, at least by obtaining how real data could incorporate bias. New tech- data providers; access to the relevant products at a discount. Indi- niques to avoid bias are needed. avoiding bias; viduals who obtain their own health data could Although machine learning could eliminate and the human contribute that data into curated records, which the need for human involvement in many tasks, element. could be used for wider benefit. The manner in human understanding of the outcomes from AI which the Met Office has been forced to offer its will still be fundamental. Those responsible for data free to competitors has weakened it in an the expansion of apprenticeships should consider unhelpful way. how jobs in AI could be served. Universities There is a healthy overlap between AI and tra- could create degree apprenticeships in AI. ditional neuroscience, which goes back to the Greater efforts must be made to enthuse young origins of AI. These synergies will continue to be people about the potential of AI and the recent fruitful. Developments in AI could widen global initiatives to increase the teaching of computer inequalities: perhaps this should be investigated science in schools are most welcome. ☐ fst journal www.foundation.org.uk March 2018, Volume 22(2) 43 COMMENT The future of social care Norman Lamb

ocial care, and its interactions with the health was rejected by the Labour Government on the service, has been a political ‘hot potato’ for grounds that it would carry “a very substantial cost” decades. Yet the UK has still not been able to and “would not necessarily improve services”. settleS on a coherent strategy. Instead, many people Further proposals were developed during the find themselves feeling terribly let-down and vul- Labour years; in particular, more generous nerable just when they need access to these services. means-testing, a ‘National Care Service’, a two-year How can we move forward as a society on this cap on paying for social care from 2014, and care urgent issue? Technology and innovation have vital free-at-the-point-of-use at some point after 2015. The Rt Hon Norman Lamb parts to play in addressing the challenge, yet they are MP is a member of the not the only elements in an effective solution. Personal budgets Council of the Foundation for For me, the aim of social care is to enable our On other fronts, changes had started to happen. Science and Technology as well as Chair of the House of citizens to lead good, happy and fulfilling lives – in Personal budgets, a concept which had its origins Commons Select Committee the way they want to lead them. That involves peo- in the United States, were introduced and the right on Science and Technology. ple having power and control, rather than having to a personal budget was later enshrined in legisla- He has been the Liberal things done to them. Then, at the end of life, the tion by the Care Act of 2014. Democrat MP for North Norfolk since 2001. In 2005 really important priority is for individuals to have a With the arrival of the Coalition came a new he was appointed Shadow dignified death. This is denied to many people. initiative. The Dilnot Commission was estab- Secretary of State for Trade Our society has made enormous advances in lished to consider a partnership model between and Industry. In 2006, he keeping people alive with medical technologies individuals and the state. Its report concluded that became Liberal Democrat Shadow Health Secretary. and therapeutics. However, at times the mainte- “the current system is confusing, unfair and unsus- Following the 2010 General nance of life seems to have become the sole aim, tainable”. It highlighted that people were left Election, he served first as sometimes irrespective of the dismal quality of “unable to plan ahead to meet their future care Chief Parliamentary Adviser that life. Just staying alive is not what I or, I suspect, needs”. Echoing one of the conclusions of the to Nick Clegg, the Deputy Prime Minister, and then most of us want. Royal Commission, it said that “a major problem is as a junior minister at the Atul Gawande is a surgeon living and working in that people are unable to protect themselves Department of Business, the USA but his book, Being Mortal, has universal against very high care costs” in old age. Innovation and Skills (BIS), relevance. It presents a stark critique of where we Dilnot’s central recommendations were for a before he was promoted to Minister of State for Care and have gone wrong; approaches are too often focussed £35,000 lifetime cap for over 65s and lower caps for Support at the Department on minimising risk, restricting the lives of vulnera- those in younger age groups, along with a more gen- of Health. ble people, often institutionalising them. He makes erous means-test. the point that we permit children to take more risks The Care Act, which I took through Parliament, than a person approaching the end of their life. enshrined the three principles of choice, indepen- The history of the past two decades indicates a dence and prevention. Promoting the individual’s very clear public policy failure in the UK. The wellbeing must be at the heart of all decision-mak- omens looked good when, coming into office in ing by the local authority (and, for the first time, 1997, the first Blair Government delivered on a key that was to include the wellbeing of the carer). manifesto commitment by establishing a Royal The cap on care costs was initially set as a life- Commission on the funding of long-term care. time cap of £72,000 for everyone over the age of 25 The Commission identified a range of failings in with a more generous means-test for support for the social care system: “It is too complex and pro- those who had not reached the cap. This meant that vides no clarity as to what people can expect. It too more people could get a contribution towards the often causes people to move into residential care cost of care. All this was due to be implemented in when this might not be the best outcome. Help is April 2016. available to the poorest, but the system leads to the Within weeks of the election of a Conservative impoverishment of people with moderate assets Government in May 2015, the cap was postponed before they get any help.” until 2020. The Conservative General Election These words could be written today. Manifesto in 2017 abandoned the cap and, instead, For all its merits, though, the Royal Commission proposed protection for just the last £100,000 of a ultimately failed to resolve these fundamental person’s assets. There was to be no pooling of risk. issues. Its chief recommendation – of free personal This was immediately dubbed a ‘dementia tax’. The care on the basis of need – divided its members and proposal was changed just a few days later.

44 March 2018, Volume 22(2) fst journal www.foundation.org.uk COMMENT

The Budget in November 2017 failed even to Those in younger adult life – starting a family, try- Faced with such a mention social care. The process is expected to start ing to buy their first home or renting – cannot be dismal landscape, it all over again with a Green Paper in the summer. expected to meet the full burden. would be easy to It is worth reflecting on the extraordinary Kate Barker’s Commission for the King’s Fund conclude that this is demographic changes since the late 18th century has suggested a range of ways of raising extra an intractable which are still gathering pace. In 1982, there were resources. Many of these have merit. problem. That, about 600,000 people over the age of 85. By 2007, Critically, additional resources for health and though, would be the this had more than doubled to 1.3 million and by care must be used to shift the emphasis to prevent- wrong conclusion. 2032 it will be 3.1 million. ing ill-health and so potentially reducing the Half of people over the age of 75 live alone. Social increase in demand for health services. isolation is known to affect health and wellbeing. Loneliness is said to be more damaging to health An irrational divide than smoking 15 cigarettes a day. In the past, a fam- There is an irrational and highly damaging divide ily would all live on the same street while now peo- between health and social care budgets. Too ple are left stranded as extended families are dis- many people fall through the gap, caught in the persed far and wide. crossfire between the NHS and local government. The economy’s ability to fund services is also The drive for a pooled budget in places like Great- increasingly difficult. The ratio of people of work- er Manchester has massive potential to deliver ing age (those who work and pay taxes) to people in better outcomes for people. Pooled budgets for retirement is changing. For every 1,000 people of localities, with a single commissioner for local working age, the number in retirement is projected services who has some democratic legitimacy to rise from 305 in mid-2016 to 370 by mid-2041. (accountable to the community rather than to In addition, around 300,000 working-age Whitehall) would be a significant step in the right adults rely on the social care system today. Finan- direction. cial pressure from the increasing care needs of We also need to see greater local experimenta- younger adults with disabilities or mental health tion, giving regions the freedom to innovate and problems is rising. shape a health and care system that works for local Social care has always lost out to the National populations. Health Service. Since 2010, first under the Coali- Pooling of risk (as proposed by Dilnot) through tion and then under the Conservatives, cuts to local a partnership model – shared responsibility authority funding has resulted in social care being between the individual and the state – makes emi- cut by 8% from a peak in 2009-10. This has put the nent sense. It also reduces the divide between the sector under immense pressure – and the impact on free NHS and means-tested social care. those in need of care, some of the most vulnerable A recent report from Stanford University in our society, is stark and distressing. School of Medicine argues that “a focus on data in These pressures show no signs of abating. The the coming years has the potential to make health King’s Fund, The Nuffield Trust and the Health care more preventive, predictive and person- Foundation have all warned of a looming £2.5 bil- alised, meaningfully reduce health costs and lead lion annual funding gap in adult social care by the to better patient outcomes.” Data has the poten- end of this decade. tial to transform the future of health and social Faced with such a dismal landscape, it would be care. Information is becoming easier to collect, easy to conclude that this is an intractable problem. analyse and understand, opening the door for That, though, would be the wrong conclusion. major advances in preventive care, quality of care Crucially, though, this will not be solved through and cost of care. At both the individual and pop- the normal process of partisan politics. A ulation level, data is helping to prevent ill-health, cross-party process is needed to develop a sustain- provide earlier diagnosis and more effective able, long term settlement for the NHS and social treatment. care. Such an approach also has to engage with the One example is in the dramatic increase in the public and with staff. use of wearable devices which collect healthcare Necessary additional resources have to be raised data. Globally, there were an estimated 274 mil- in a progressive way, based on ability to pay, but it lion sold last year. Fitness bands are the most pop- must also be fair between the generations. Serious ular. Data produced can be linked directly to consideration should be given to a hypothecated health professionals. We have the ability to track Health and Care Tax, perhaps based on a reformed temperature, pain and stress through ther- National Insurance system. Reform of NI would be mo-electric pulses. The promise of wearables is necessary to ensure that people on high earnings that they will help to detect and treat illness at an above retirement age make a fair contribution. earlier stage. fst journal www.foundation.org.uk March 2018, Volume 22(2) 45 COMMENT

At both the individual At present the market is dominated by recre- hospital. They get people out of their homes. They and population level, ational technology and fitness companies but do the shopping. Another brilliant example of the data is helping to health and care organisations are already seeing the power of communities is Shared Lives Plus. Fam- prevent ill-health, potential to make use of medically robust and rele- ilies take people with a learning disability or men- provide earlier vant data. Remote home monitoring is growing, tal ill-health into their homes. They are paid for diagnosis and more although the adoption of technologies in the UK is the care they provide but the cost is a fraction of effective treatment. very slow. the cost of institutional care. Some other European countries are further Within the membership of Shared Lives there are ahead – Norway has developed a national pro- also community-based services which share the gramme to drive the use of technology. Spain is same values and ethos. This is a very different exploring the use of incentives to make it worth- approach which helps put people in control and while for providers to invest in technology. It is ini- helps them lead ordinary lives. tiating the procurement of a service over a four-year There needs to be a recognition at all levels of period that includes risk sharing. society of the importance for people to be in control The UK could learn from this. When the Gov- of their care and their lives. Although the Care Act ernment gets behind a technology, such as electric enshrined in law the right to a personal budget, care cars, it makes things happen. We could see dramat- provision is continually and relentlessly trimmed ic advances in health and care if the Government down to the bare basics. Too many councils still pay similarly got behind assistive technologies in these lip service to transferring power to citizens. This has sectors. There is potential to empower people to to be challenged. take control of their lives, maintain independence And at the end of life, people should be able to die for longer and stave off dependency. where they want to, their priorities respected. Too Technology is not the whole answer, though. many are trapped in the alien environment of a busy Ultimately, human contact, kindness and attention acute hospital. We have the ludicrous situation that are essential for making life meaningful. While half care is free in the place the person generally least of over 75-year olds live alone, couples can be lonely wants to be, while not free at home. too – particularly where someone is caring for a None of this is rocket science. This challenge is loved one perhaps with dementia. solvable. Yet today, many people feel horribly let We surely all have a responsibility to each other. down. Money is wasted. People are denied their On every street, there are people living alone. In dignity. As one of the biggest economies in the Potter Heigham on the Norfolk Broads, a group of world, we can do better. ☐ local people make sure that no one in the village need be lonely. They take people to the doctor or to Email: [email protected]

OBITUARY: THE RT HON SIR BRIAN NEILL

Council Member of the Foundation for attempt to publish the diaries of the Labour the foundations for decades of later effort in Science and Technology and former Lord minister Richard Crossman. He also worked the field. For almost 40 years, Sir Brian led the Justice of Appeal on a series of high-profile libel cases. charge for the modernisation and digitisation 2 August 1923 - 24 December 2017 In 1978 he was appointed a High Court of the courts and the work of judges.” judge and he served as Lord Justice of In 2003, to coincide with his 80th To the world at large, Sir Brian Neill was best Appeal from 1985 to 1996. Shortly after, birthday, a festschrift was published known as the barrister who represented the he became President of the Court of Appeal entitled Essays in Honour of Sir Brian Sunday Times when the Government tried in Gibraltar, a position he held until 2005. Neill: The Quintessential Judge. In a to prevent publication of an article about Early on, though, he had recognised the joint introduction, Lord Bingham (then the drug thalidomide. However, his more possibilities that computing could bring to Senior Law Lord) and Lord Woolf (then enduring contribution to the legal system the legal world and he was one of the co- Lord Chief Justice) spoke of Sir Brian as may prove to be his work in encouraging the founders of the Society for Computers and “an advocate who was highly erudite, use of computers and automation. Law (SCL), the inaugural meeting of which very intelligent, immensely well-prepared After serving in the Army during the held in December 1973. and so persuasive that it always seemed Second World War, Brian Neill studied law In May 1979, he set out some initial unreasonable to disagree with him”. They at Oxford, being called to the bar at the thoughts on ‘Computers and the Courts’ went on to speak of his distinguished Inner Temple in 1949 and becoming a QC in in the Society’s newsletter. As the current contribution as a judge whose “court was 1968. In addition to the thalidomide case, President of the SCL, Professor Richard a showcase for the British system at its he represented the Sunday Times in its Susskind, remarked: “In that paper, he laid exemplary best”.

46 March 2018, Volume 22(2) fst journal www.foundation.org.uk EVENTS

Presentations and audio recordings from all meetings of the Foundation for Science and Technology are ­available at: www.foundation.org.uk

UKRI leaves the starting blocks: the Paul Wheelhouse MSP, Minister for What constitutes an effective industrial management of government funding of Business, Innovation and Energy and strategy for the UK? research and innovation Member for South Scotland, Scottish 10 May 2017 28 February 2018 Parliament Professor Graeme Reid, Specialist Adviser Sir Mark Walport FRS FMedSci HonFRSE, to the House of Lords Select Committee on Chief Executive, UKRI Searching for the Holy Grail of a science and Science and Technology Sir Alan Wilson FBA FRS, Chief Executive, innovation strategy that makes a difference Professor Dame Ann Dowling DBE FRS the Alan Turing Institute 18 October 2017 FREng, President, Royal Academy of Kirsten Bound, Executive Director of The Lord Hennessy of Nympsfield FBA, Engineering Research Analysis and Policy, NESTA House of Lords Andrew Barker, Head of Investor Relations, [Panellist] The Rt Hon the Lord Heseltine CH, House International Airlines Group Jonathan Neale, Chief Operating Officer, of Lords Anthony Lilley OBE, Chief Executive and McLaren Technology Group Ltd [Panellist] The Rt Hon the Lord Willetts, Executive Chief Creative Officer, Magic Lantern Chairman, The Resolution Foundation Dr Andrew Harter FREng FIET FBCS, The Hackitt Review of Building Regulations Chair, Cambridge Network and Founder and Fire Safety Cancer diagnostics: can cancer be and CEO, RealVNC [Panellist] 24 January 2018 diagnosed earlier and if yes what are the Dame Judith Hackitt DBE FREng, Chair, consequences? What needs to be done to meet urban air Hackitt Inquiry into Building Regulations 11 July 2017 quality targets and what are the and Fire Safety Sir Harpal Kumar, Chief Executive, Cancer consequences if the targets are not met? Graham Watts OBE, Chief Executive, Research UK 26 Apr 2017 Construction Industry Council (CIC) Billy Boyle, Chief Executive Officer, Eliott Treharne, Air Quality Manager, Peter Baker, Director, Construction Owlstone Medical Greater London Authority Division and Chief Inspector of Dr Clare Turnbull, Clinical Lead, Genomics Dr Stephen Bryce, Vice-President, Fuels Construction, Health and Safety Executive England 100,000 Genomes Cancer Technology, Shell Projects and Technology Dr Peter Bonfield OBE FREng, Member, Programme Professor Frank Kelly, Professor of Grenfell Expert Panel, Ministry of Housing, Environmental Health, King’s College Dr Suzanne Jenkins, Diagnostics Expert Communities and Local Government and London (Director), Personalised Healthcare and Chief Executive, BRE Group [Panellist] Dr Christa Hasenkopf, Chief Executive and Biomarkers, AstraZeneca [Panellist] Turlogh O’Brien CBE, Chairman of the Co-Founder, OpenEQ [Panellist] Sara Hiom, Director of Early Diagnosis and Governing Board of the Chartered Institute Health Professional Engagement, Cancer of Housing and Post-Grenfell Expert How can skill levels be raised to meet the Research UK [Panellist] Working Group, Construction Industry needs of society and the economy? Council (CIC) [Panellist] 1 March 2017 The impact of demographic and medical Sir Mark Walport FRS FMedSci, trends on the health and social care The impact on society of machine learning Government Chief Scientific Adviser, systems of the UK – an opportunity or a threat? Government Office for Science 21 Jun 2017 14 November 2017 Sir Adrian Smith FRS, Chair of the Smith Dr Mike Lynch OBE FRS FREng, Founder, Professor Chris Whitty CB FMedSci, Chief Inquiry into mathematics education for 16 Invoke Capital Scientific Adviser, Department of Health, to 18 year olds and Vice Chancellor of the Dr Claire Craig CBE, Director of Science Deputy Government Chief Scientific Adviser University of London Policy, The Royal Society Sir Robert Lechler PMedSci, President, Dame Judith Hackitt DBE FREng, Chair, Amir Saffari, Head of AI, BenevolentAI Academy of Medical Sciences EEF (formerly the Engineering Employers’ Dame Wendy Hall DBE FRS FREng, Regius Professor Marcel Levi, Chief Executive, Federation) Professor of Computer Science, University University College London Hospitals NHS Stephen Metcalfe MP, Chair, House of of Southampton Foundation Trust Commons Select Committee on Science and Professor Chris Bishop FRS FREng, Technology [Panellist] Laboratory Director, Microsoft Research, Making cities work - the application of Cambridge [Panellist] technology, science and infrastructure Making good use of science and innovation improvements to create a place where in overseas development programmes A business strategy for Scotland citizens wish to live 14 December 2016 6 November 2017 24 May 2017 Professor Charlotte Watts FMedSci, Chief Professor Iain Gray CBE FREng FRSE, Vice Professor The Lord Mair CBE FRS FREng, Scientific Adviser and Director Research President for Business, The Royal Society of Sir Kirby Laing Professor of Civil and Evidence Division, Department for Edinburgh Engineering, Department of Engineering, International Development Nora Senior CBE, Chair, Scottish University of Cambridge Jon Ridley, Head, M-KOPA Labs, M-KOPA Government’s Strategic Board for Enterprise Tom Saunders, Principal Researcher, Solar and Skills International Innovation, Nesta Rowan Douglas CBE, Chief Executive, Capital, Dame Susan Rice CBE FRSE, Chair, Scottish Councillor Peter Marland, Leader, Milton Science & Policy Practice and Chair, Willis Water Keynes Council Research Network, Willis Towers Watson

fst journal www.foundation.org.uk March 2018, Volume 22(2) 47 EVENTS

The opportunities for and threats to the What is the value to the economy of the Using science to authenticate, verify or research and innovation communities from finance and insurance sectors? assure the identity of people and things Brexit 6 July 2016 2 March 2016 16 November 2016 Anne Richards CVO CBE FRSE, Chief Sir Mark Walport FRS FMedSci, Government Sir Venki Ramakrishnan PRS FMedSci, Executive, M&G Investments Chief Scientific Adviser President, The Royal Society John Nelson, Chairman, Lloyd’s of London Dr Derek Craston, Government Chemist and Professor Louise Richardson FRSE, Vice- Professor John Kay CBE FRSE FBA, Managing Director of Science and Chancellor, University of Oxford Economist and Financial Times Columnist Innovation at LGC The Rt Hon the Lord Willetts, House of Professor Dame Sue Black DBE FRSE, Lords How should universities and Research Professor of Anatomy and Forensic Anthropology at the University of Dundee Dr Hermann Hauser KBE FRS FREng, Councils proactively respond to gender bias Co-Founder, Amadeus Capital Partners in success rates in grant applications? Bringing science to the heart of government: [Panellist] 22 June 2016 Professor Madeleine Atkins CBE, Chief the Nurse Review of the Research Councils Professor Paul Boyle CBE FBA FRSE, Executive, Higher Education Funding 12 January 2016 President and Vice-Chancellor, University Council for England [Panellist] Sir Paul Nurse FRS FMedSci, Chair, the Nurse of Leicester Review of the Research Councils, and The vision for UK Research and Innovation Professor Henrietta O’Connor, Deputy Director, The Francis Crick Institute (UKRI) Head of College of Social Science, Arts and Professor Phil Nelson FREng, Chair, RCUK 9 November 2016 Humanities and Professor of Sociology, Executive Group and Chief Executive, Sir John Kingman KCB, Chair, UK Research University of Leicester Engineering and Physical Sciences Research and Innovation, Department of Business, Linda Holliday, Director of Capacity and Skills Council Energy and Industrial Strategy Development, Medical Research Council Gareth Davies, Director General, Business and Science, Department for Business, Professor Dame Julia Goodfellow DBE Innovation and Skills FMedSci, President, Universities UK and Is a paradigm shift taking place in the ways Professor Dame Jocelyn Bell Burnell DBE FRS Vice-Chancellor, University of Kent individuals and organisations access, FRAS PRSE, President, The Royal Society of Phil Smith, Chair, Cisco UK & Ireland, Chair, analyse and protect data? Edinburgh [Panellist] Innovate UK and Chair, The Tech Partnership 25 May 2016 Professor Sir Nigel Shadbolt FREng, Chairman Closing the US/UK productivity gap: Health, happiness and wellbeing: supporting and Co-Founder, The Open Data Institute connecting innovation and research to the transition from adolescence to adulthood Dr Mike Lynch OBE FRS FREng DL, Founder, economic output 26 October 2016 Invoke Capital 2 December 2015 Dr Joanne McLean, Research and Professor David Hand OBE FBA, Chief Dr Ruth McKernan CBE, Chief Executive, Development Manager, Scotland, Mental Scientific Adviser, Winton Capital Innovate UK Health Foundation Baroness O’Neill of Bengarve CH CBE FBA Professor Jonathan Haskel, Professor of Dr Helen Sweeting, Reader, MRC/CSO HonFRS FMedSci, House of Lords [Panellist] Economics, Imperial College Business School Social and Political Health Sciences Unit, Tony Harper, Head of Research and Advanced University of Glasgow The pros and cons of EU membership for UK Systems Engineering, Jaguar Land Rover Lord Layard FBA, Director, Wellbeing research programmes in private enterprises Programme, Centre for Economic and public sector organisations Responding to a changing Arctic: The House Performance, London School of Economics 3 May 2016 of Lords Arctic Select Committee Report and Political Science The Lord Hennessy of Nympsfield FBA, 4 November 2015 Catherine Calderwood FRCP, Chief Member, House of Lords Science and The Lord Teverson, Chair, House of Lords Medical Officer for Scotland, Scottish Technology Select Committee, House of Lords Select Committee on the Arctic, House of Government [Panellist] Viscount Ridley FMedSci FRSL, Member, Lords Jane Rumble, Head, Polar Regions The National Flood Resilience Review: the House of Lords Science and Technology Select Committee, House of Lords Department, Foreign and Commonwealth lessons learned from recent flood events in Office Professor Dame Jocelyn Bell Burnell DBE FRS the United Kingdom Professor Dame Julia Slingo DBE FRS, Chief FRSE FRAS FInstP, President, The Royal 12 October 2016 Scientist, Met Office Professor Dame Julia Slingo DBE FRS, Society of Edinburgh Chief Scientist, Met Office Sir Emyr Jones Parry GCMG FInstP FLSW, The Accelerated Access Review for the Dr Doug Wilson, Director, Scientific & President, The Learned Society of Wales Department of Health (the Taylor Review) Evidence Services, Environment Agency 26 October 2015 Simon Warsop, Chief Underwriting Officer, Building effective and efficient Sir Hugh Taylor KCB, Chair, Accelerated Personal Lines, Aviva infrastructure for the UK Access Review, Department of Health Professor Charles Godfray CBE FRS, Chair, 27 April 2016 Sir Leszek Borysiewicz FRS FRCP FMedSci Defra Science Advisory Council and Tony Meggs, Chief Executive, Infrastructure FLSW, Vice-Chancellor, University of University of Oxford and Projects Authority, Cabinet Office Cambridge Professor Bas Jonkman, Professor of The Rt Hon The Lord Adonis, Chair, National Integral Hydraulic Engineering, Delft Infrastructure Commission University of Technology Sir Terry Morgan CBE, Chairman, Crossrail Katharine Hammond, Director, Civil Darren James, Managing Director, Contingencies Secretariat, Cabinet Office Infrastructure, Costain [Panellist]

48 March 2018, Volume 22(2) fst journal www.foundation.org.uk MAJOR SUPPORTERS IN 2017/2018

A I Royal Society of Chemistry Academy of Medical Sciences IBM Royal Society of Medicine Arts and Humanities Research Council ICP London S Association of Innovation, Research and Imperial College London Science and Technology Facilities Technology Organisations Innovate UK AstraZeneca Shell Institute of Biomedical Science Atkins Limited Sir William Francis CBE FREng Institute of Materials, Minerals & Mining Society and College of Radiographers B Institute of Mathematics and its BAE Systems Applications Society of Maritime Industries Biotechnology and Biological Sciences Society of Operations Engineers Research Council Institute of Quarrying Sovcomflot (UK) Ltd BP International Ltd Institution of Chemical Engineers IPA T BPE Solicitors LLP The Haskel Family Foundation BRE Group J Japan Society for the Promotion of The IET British Academy Science The Kohn Foundation British Geological Survey Johnson Matthey Plc The Medical Schools Council BSI Group K The Peter Jost Memorial Foundation Keele University C The Nautical Institute Canterbury Christ Church University King's College London The Royal Academy of Engineering Cardiff University Knowledge Transfer Network Limited The Royal Commission for the Exhibition Chartered Association of Building L Engineers Lloyd's Register Foundation The Royal Society Chartered Institute of Credit The Royal Society of Edinburgh Management M Medical Research Council The Wellcome Trust Chartered Institute of Linguists Met Office U Chartered Institute of Plumbing and Heating Engineering Microsoft Research Limited University Alliance Cogent Skills Ltd N University College London National Physical Laboratory Comino Foundation University of Birmingham Natural Environment Research Council Cranfield University University of Bristol Natural History Museum University of Chichester D Defence and Security Accelerator Network Rail University of Dundee Department of Health Nottingham Trent University University of East Anglia E O University of Edinburgh Office for National Statistics Economic and Social Research Council University of Glasgow EEF P Parliamentary and Scientific Committee University of Hull Energy Institute Phrase University of Kent Engineering and Physical Sciences Research Council Plymouth University University of Leeds University of Leicester ERA Foundation Q Queen's University Belfast University of Nottingham G Genomics England R University of Reading Recruitment and Employment GlaxoSmithKline Confederation University of Sheffield H Risk Solutions University of Southampton Heads of University Centres of Biomedical Science (HUCBMS) Rolls-Royce plc University of Warwick Health and Safety Executive Royal Academy of Engineering University of York High Value Manufacturing Catapult Royal Geographical Society (with the W IBG) Willis Towers Watson Higher Education Funding Council for England Royal Society of Biology Winton Philanthropies

The Foundation is grateful to these companies, departments, research bodies and charities for their significant support for the debate programme. The Journal of The Foundation for Science and Technology

The Foundation for Science and Technology 10 Carlton House Terrace London SW1Y 5AH

Telephone: 020 7321 2220 Email: [email protected]

www.foundation.org.uk