Building a Smart Laboratory 2018
An introduction to the integrated lab
From the publishers of From the publishers of
www.scientific-computing.com/BASL2018 Pharma&Biotech
The MODA™ Platform The Missing Piece in Your Lab Systems Portfolio
MODA-EM™ Software for QC Micro – Implement, Validate, Integrate Seamlessly.
Lonza’s MODA-EM™ Software is purpose-built for QC Microbiology’s unique – Reduce investigation and reporting time with rapid analysis challenges. MODA™ Software delivers better business alignment with QC, and visualization of Micro data at a lower total cost of ownership versus customizing your existing LIMS – Manage, schedule and track large EM and utility sample volumes or other lab systems. – Reduce errors while improving compliance and data integrity – Deliver measurable return on investment from time, cost and Find the missing piece in your lab systems portfolio. compliance savings Visit www.lonza.com/moda or contact us at [email protected]
© 2015 Lonza www.lonza.com/moda
MODA IT Ad 0715 for Scientific Computing World
MODA_IT_Ad_Comps_final_selection.indd 1 7/27/15 5:27 PM Contents
WELCOME TO THE SMART LABORATORY
his year’s edition of Building a Smart Laboratory discusses the importance of developing a robust strategy for the deployment of paperless lab Ttechnology. As the article on page 6 discusses, in order to gain the most insight and value from paperless technology there needs to be a consistent and comprehensive approach that covers the four most important pillars; connect, manage, decide, archive. As laboratories seek to drive more value and to move from a cost centre to being a value 4 An introduction to building 26 Knowledge: Document proposition for an organisation, it is important a smart laboratory 2018 management that all knowledge can be used effectively to generate the largest return on investment. The This introduction sets out procedures to help lab How the smart laboratory contributes to the only way to truly achieve this is to adopt smart users implement paperless technologies in the lab requirements of a knowledge eco-system, and the – with a particular focus on data-intensive science practical consequences of delivering access and laboratory technology. and new trends preservation of knowledge that was traditionally This is a consistent theme throughout stored in paper archives the entire publication. Building a smart laboratory can provide huge benefits to an organisation in terms of increased productivity 6 The key layers to a paperless strategy 30 or value generation, but also through collection Beyond the lab management and archiving of data. However, Developing a robust strategy is a key concern in order to make the most of the investment when deploying paperless laboratory technology, How the smart laboratory can help to improve your business, through greater productivity in ‘smart’ technologies, it is imperative that a write Isabel Muñoz Willery and Roberto strategy is devised that can look at the needs Castelonovo of NL42 Consulting and efficiency, better integration with existing systems, better regulatory compliance, data to the lab and its users to properly adapt and integrity and authenticity configure the technology accordingly. Technology will not do the thinking for us, 10 Dealing with data but if properly constructed a smart laboratory Informatics providers share their experiences Practical considerations in can add considerable value. While this guide on the importance of using the latest laboratory 36 specifying and building the cannot provide all the answers, it does provide an technology smart laboratory introduction to everyone that faces the challenge of increasing productivity and data integrity for This chapter focuses on how to go about building the modern laboratory workflow. a smart laboratory with information relating to 12 Smart laboratories approaches to take, and potential roadblocks An introduction to the concept of a ‘smart’ The authors of the guide are: Peter Boogaard laboratory, based on the data/information/ Industrial Lab Automation knowledge triangle 40 Knowledge: Data analytics Siri Segalstad Taking the theme of knowledge management Segalstad Consulting AS beyond document handling into the analysis of Joe Liscouski Institute for Laboratory Automation 14 Data: Instrumentation data to help develop new products or improve Charlie Sodano existing ones eOrganizedWorld We look at the latest progress towards truly John Trigg digital laboratories, with a focus on the types of phaseFour Informatics Ltd laboratory instruments and their capabilities 41 Summary Isabel Muñoz-Willery Pulling together the various threads on how to NL42 Consulting SL make the laboratory ‘smart’ this chapter hopes to Roberto Castelnovo NL42 Consulting SL Information: Laboratory lay out the most important factors that must be 19 considered informatics tools Cover image and all other images: Shutterstock.com Building a Smart Laboratory is published by Europa Science, the An overview of laboratory informatics tools publishers of Scientific Computing World (ISSN 1356-7853). ©2018 – LIMS, ELN, LES and – how convergence is 42 References and further reading Europa Science Ltd. 4 Signet Court, Cambridge, CB5 8LA, UK. changing the informatics market All images Shutterstock.com Design: Zöe Andrews Tel: +44 (0)1223 221033. Fax: +44 (0)1223 213385. www.scientific-computing.com/BASL2018
www.scientific-computing.com/BASL2018 3 An introduction to: Building a Smart Laboratory 2018 Building a Smart Laboratory 2018
AN INTRODUCTION TO Building a Smart Laboratory 2018
It’s rare for a company to start with sharing, the barriers to implementing successful a clean slate when making decisions Paperless or less paper? electronic integrated processes often remain a about laboratory automation bridge too far. Data-intensive science is becoming far more mainstream; however, going digital in the The informatics journey laboratory has been a relatively slow process. More his chapter serves as an introduction than 75 per cent of laboratory analysis starts with to this guide Building a Smart The journey starts with data capture, data a manual process such as weighing; the majority Laboratory 2018. We hope to highlight processing, and laboratory automation. When of results of these measurements are still written the importance of adopting smart samples are being analysed, several types of down or re-typed. Tlaboratory technology but also to guide users scientific data are being created. They can be There are exceptions: probably the best through some of the challenges and pitfalls when categorised in three different classes. example of integrated laboratory automation designing and running the latest technologies in Raw data refers to all data on which decisions can be found in how chromatography data the lab. are based. Raw data is created in real-time from an handling systems (CDS) operate in modern For any laboratory a cost/benefit analysis instrument or in real-time from a sensor device. laboratories. The characteristics of such a system needs to consider the functionality already Metadata is ‘data about the data’ and it is used include repeatable, often standardised, automated provided by legacy applications – as well as for cataloguing, describing, and tagging data processes that create a significant amount of raw business justifications. This guide will help you resources. It adds basic information, knowledge, and processed data. understand what informatics processes are and meaning. Metadata helps organise electronic The paper versus paperless discussion is as needed in laboratories, and why the laboratory resources, provide digital identification, and old as the existence of commercial computers. In should not merely be seen as a necessary cost helps support archiving and preservation of the the 1970s, just after the introduction of the first centre. resource. personal computer, Scelbi (Scientific, Electronic Only by becoming smart – as this guide Secondary or processed data describes how and Biological), Business Week predicted that outlines – can lab managers change that mind-set raw data is transformed by using scientific computer records would soon completely replace and generate true value for their organisation. methodologies to create results. To maintain paper. It took at least 35 years before paperless Many laboratory operations are still data integrity, altering methods to reprocess will operations were accepted and successfully adopted predominantly paper-based. Even with the require a secured audit trail functionality, data in many work operations. Although they have enormous potential to reduce data integrity and access security. If metadata is not captured, been accepted in banking, airlines, healthcare, and for compliance, to make global efficiency gains the ability to find and re-use previous knowledge retail, they lag behind in science. in manufacturing and to increase knowledge from scientific experiments is eliminated. The journey from paper to electronic begins
4 www.scientific-computing.com/BASL2018 Building a Smart Laboratory 2018 An introduction to: Building a Smart Laboratory 2018 with the transition from paper to digital, which changing to a new model based upon a ‘pay-as- mainstream adoption. The acceptance of tablets includes both the transfer of paper-based you-go’ or philosophy (OPEX). CRM applications and mobile devices will expand exponentially in processes to ‘glass’ and the identification and such as SalesForce.com started this business the laboratory. adoption of information and process standards to model in the traditional enterprise business Laboratories will need to manage the harmonise data exchange. software segment. Popular applications such as challenges presented by new consumers of Photoshop, Microsoft Office 365 and Amazon scientific data outside traditional laboratory Think exponential are following these trends rapidly. It is expected operations. Non-invasive, end-to-end strategies that scientific software suppliers will be forced will connect science to operational excellence. Traditional mainstream LIMS will face challenges. to follow the same model in the years to come. Technology will be critical, but our ability LIMS has been a brilliant tool to manage Community collaboration and social networking to change our mind-set to enable this cross- predictable, repeatable planned sample, test is changing the value of traditional vendor help functional collaboration will be the real and study data flows, creating structured data desks. challenge. n generated by laboratories. In R&D environments, unpredictable workflows creating massive Reduce and simplify workflow complexities amounts of unstructured data showed that current The need to simplify our scientific processes Adapting to change LIMS systems lack the capability effectively to will have a significant impact on reducing data manage this throughput. ELNs are great tools to integrity challenges. For example, balance and Much of the change that drives new capture and share complex scientific experiments, titrator instruments may store approved and processes or methods in the laboratory is while an underlying scientific data management pre-validated methods and industry best practice based on regulation from that aims to more system (SDMS) is used to manage large volumes workflows in their firmware. tightly control the way in which data is of data seamlessly. collected, stored and handled. Adopt and use industry standards and processes Many laboratory users will be aware Data consumer vs data creator examples Initiatives such as the Allotrope Foundation are of previous regulations such as Title 21 working hard to apply common standards. The CFR Part 11, part of Title 21 of the Code For the researcher, the ability to record data, make Allotrope Foundation is an international not-for- of Federal Regulations that establishes the observations, describe procedures, include images, profit association of biotech and pharmaceutical United States Food and Drug Administration drawings and diagrams and collaborate with companies, building a common laboratory (FDA) regulations on electronic records and others to find chemical compounds, biological information framework for an interoperable electronic signatures (ERES).[1] structures – without any limitation – requires a means of generating, storing, retrieving, Part 11, of the document, as it is flexible user interface. For the QA/QC analyst transmitting, analysing and archiving laboratory commonly called, defines the criteria under or operator, the requirements for an integrated data and higher-level business objects. which electronic records and electronic laboratory are quite different. A simple, natural signatures are considered trustworthy and language-based platform to ensure that proper Consolidation and equivalent to paper records. procedures are followed will be well received. harmonisation of systems However new regulation around General Product innovation and formulators will Data Protection Regulation (GDPR) need the capability to mine data across projects, Most laboratories already depend on an and data integrity are new standards that analytical methods or formulations to create informatics hub comprising one or more of the laboratory users must now familiarise valuable insights. Transforming unstructured major tools: laboratory information management themselves with. For many users GDPR will scientific experimental data into a structured systems (LIMS); electronic laboratory notebooks not be applicable as it only relates patient equivalent will be mandatory to perform these (ELN); scientific data management systems data or companies that hold data of EU tasks. (SDMS); chromatography data-handling systems citizens. However, if in a clinical setting Organisations with a strong consumer (CDS) and laboratory execution systems (LES). GDPR could have a huge effect on the way marketing focus deal with data mining techniques The trend over recent years has been towards that you store patient data. [2] providing clear pictures of products sold, price, convergence, applying best practice industry In addition to GDPR lab managers must competition and customer demographics. standard processes to harmonise multisite also familiarise themselves with pending deployments. Cost reduction to interface regulation on Data Integrity (DI) which New trends harmonised processes to ERP (SAP), MES and hopes to improve completeness, consistency, CAPA results in lower maintenance and validation and accuracy of data recorded by The power of life cycle process improvement costs with a significant overall higher system laboratories [3]. In simple terms this means The scientist is no longer in the laboratory, but availability for end-users. abiding by principles such as ALCOA integrated in the overall quality process. Quality (attributable, legible, contemporaneous, should be built into the design throughout the Mobile computing original, and accurate). However it is advised specification, design, and verification process. that lab managers and users explore the Performance metrics on non-conformance While many other industries are implementing ramifications of this new regulation to see tracking are mandated and monitored by modern tools to connect equipment wirelessly, how it might affect daily workflows. regulatory authorities. Integrating laboratory many laboratories still write scientific results systems will add significant value by decreasing on a piece of paper, or re-type them into a References non-conformance. computer or tablet. Many modern ELN and 1. https://www.fda.gov/regulatoryinformation LES systems allow electronic connection to a guidances/ucm125067.htm New budgeting and licensing models (wireless) network. However, to integrate simple 2. https://www.ncbi.nlm.nih.gov/pmc/articles Managing operating budgets will be redefined instruments like a pH balance, titration and PMC5346164 3. https://www.fda.gov/downloads/drugs/guidances in the next decade. The days of purchasing Karl-Fischer instruments to mobile devices, a ucm495891.pdf software as a capital investment (CAPEX) are simpler approach is required in order to achieve www.scientific-computing.com/BASL2018 5 Planning your lab Building a Smart Laboratory 2018
‘eConnect, eDecide, eManage, eArchive’ The key layers of a laboratory paperless strategy
Isabel Muñoz-Willery and Roberto will be discussed in detail at the Paperless Lab manual transcriptions. The goal is to reduce Castelnovo, of NL42 Consulting, Academy 2018. The annual European event aims the manual documentation, the risk of human highlight the importance of developing to become a learning platform for anyone looking errors, and more importantly, to maintain the a robust strategy for the adoption to consolidate, integrate or simplify their data information about the source that has generated paperless laboratory operations management systems. the raw data. The raw data may be a critical part of the ‘eConnect’: effective workflows based activities performed in the systems of the upper on self-documenting data capture layers. Data management and the creation of n the new era of the internet of things strategies meaningful information and decisions should and artificial intelligence, the majority of be always taken with the possibility to go back to laboratories still have a long way to move Even if data integrity is a critical aspect of the the original data from the system in which it was from paper-based processes to paperless entire data life cycle, data capture requires a generated. Iones. strong focus from both the inspectors and Finally, while in this first stage of collecting The electronic data life cycle, as it is auditors. Most lab instruments are now offered data we should not obviate the ones coming described in several regulations and documents with intelligent software embedded into them. from collaborators. Collaborators are generators used in paperless projects, can be divided in Labware and sensors are beginning to embrace of data and potential sources of information. four layers of data, information and activities: the internet of things, ensuring the collection of If external organisations such as academic eConnect; eManage; eDecide and eArchive. the raw data and the related metadata which can contributors or outsourced services from CRO These keywords refer to initial capture of then be transferred to the next phase of the data and CMO are generating the data, it can create data, the data management to create useful life cycle. immediate security concerns. With the latest information, the decisions taken based on Several laboratories are using instruments GDPR considerations, we need to incorporate information and data available in the lower which are not able to connect the current data protection assessment at least on the most layers and, finally, the electronic data archiving platforms. While searching for the business vulnerable data. By May 2018, companies to ensure long-term availability of the justification for their replacement, intermediate will need to design their processes and also information and the related data. solutions should be considered to generate digital include serious considerations on cybersecurity Those are the four-main streams that inputs and reduce paper-based processes and protection to avoid any risk in losing data.
6 www.scientific-computing.com/BASL2018 LABWARE 7 LIMS and ELN together in a single integrated software platform. A laboratory automation solution for the entire enterprise.
Offices worldwide supporting customers in more than 100 countries
www.labware.com Planning your lab Building a Smart Laboratory 2018
‘eManage’: generation of meaningful ‘eDecide’: Rapid decisions taken their fingertips. Moreover, these tools are able information from trusted data from meaningful information to dig into the underlying systems to view the information and related raw data, when needed. The ‘translation’ from data to information is the In the everyday activities of a laboratory, we are We will finally see one single screen open on key principle of this layer of activities typically getting used to perform them very rapidly and the computers of the managers instead of multiple performed in the most well-known systems decisions should be taken in short time. Little windows jumping from one system to another The real challenge in the new era of Internet of remains available for data review, approval of data in order to desperately collect all the necessary Lab Things (IoLT) is not about picking up the and creation of related documents. The request information required in a given moment for a right acronym for the lab. The challenge lies coming from laboratory’s customers, both given decision to be taken urgently. in identifying the right solutions that provide internal and external is a prompt answer. answers to a series of requirements: secured The removal of manual processes, of paper- ‘eArchive’: essentials to secure long- connectivity without large investment; usability based activities and mix of information sitting term multi-departmental archiving with limited customisation; ability to share in different systems is essential for taking faster information using the newest technologies; decisions. Only paperless processes shorten A key objective in operating with efficient mobile devices and web-access without the periods of review of the information and archival approach is to reduce the challenge of performing complex platform implementations; ease rapid decision-making which can then finding the right data. Considering the growing and the possibility to use the software as a service. be communicated immediately to the relevant digital universe, archiving can no longer be left We are observing a large market stakeholders. New approaches like the review by behind in a project and considered only once it is transformation in this area. The presence exception are helping to increase the efficiency of too late. Nowadays, we often hear about concerns of systems which are offering a large set of this process. on legibility and format consistency along the functionalities and product offerings based on The laboratories that are able to respond to time for a given retention time that might end up new technologies. these requests on time and with the adequate requiring access to obsolete technologies. Multiple software modules adapted to specific level of quality will transform from cost centers to Archiving should be approached and laboratory activities and software platforms allow value generators. designed to reduce multiple types of risk: the creation of personalised solutions with no need to customise but rather configure the system to the needs of the user. This revolution will generate large benefits for the laboratories because the selection of the The removal of manual processes, of paper-based activities solutions will be based on the needs rather than and mix of information sitting in different systems is essential for the capabilities. taking faster decisions These modules should respond to a few critical requirements in order to become part of the ‘solution’: easily connectable to the ‘eConnect’ layer; easily connectable to modules of the “ ‘eManage’ layer; easily accessible from browsers and mobile devices; and easily accessible from the Decisions should be taken according to the knowledge limited to one critical person, security ‘eDecide’ layer. available information. Today many software and loss of data. providers offer simple tools presenting the A comprehensive archiving protocol should What is the end goal? information in a graphical view, showing the eliminate the struggle to find the data to the point outliers, highlighting the areas of attention, of desperately looking for the person owning the On one hand, the final goal should be to interface allowing the ‘drill-down’ approach when needed. knowledge of where it is. the ‘solution’ with the multiple generators of raw Fact is that solutions providers, integrators A corporate master data management and data in order to enable the review directly at the and customers are joining efforts in organisations vocabulary model should support a correct source at any time. Additionally, the possibility to like the allotrope foundation, Pistoia Alliance, management and archival, facilitating a flawless exchange information between the modules of the SiLA consortium to consolidate outputs and track record of the data. ‘eManage’ layer, in a flawless manner, should allow tools, that could one day lead to the creation During the Paperless Lab Academy 2018, the access and interpretation of all data to generate of one single user interface, one single way several presentations will focus on this item that meaningful information. of showing the information in a unique and too often is approached too late in a ‘paperless’ The possibility to access the ‘modules’ from personalised dashboard. project. The archiving strategy requires a clear any remote location or even from mobile devices Simple reports created automatically definition of the business requirements and, also in order to manage all the information in the overnight and available in the ‘eDecide layer’ first the potential technical challenges. shortest period of time. The possibility to provide thing in the morning. A new ‘control room’ of the The ability to archive and then retrieve aggregated information to the next layer of laboratory where decisions are taken to correct unstructured data is becoming an urgent need systems where decisions are taken. situations not in line with the expectations, where which must be solved in R&D laboratories. Is this real? Absolutely. The technology has scheduling changes are adjusted to ensure that Solutions providers are dedicating resources evolved to the level that all these goals could be the activities are completed on time, on budget to this matter and positioning their data reached. and according to the customer expectations. management software to address the need for Numerous solutions are already Is this real? Yes, again. Great reporting and better archiving and retrieval. Above all, the implemented in various markets where they are business intelligence tools are now available ‘eArchive’ strategy is one that requires stronger using the newest technologies. The laboratory to integrate the information coming from alignment within the whole company in order informatics systems will have to be ready for this different systems and present in a simple and to build up a reference master data management new era too. graphical way. All what the managers need at strategy at an enterprise-level. n
8 www.scientific-computing.com/BASL2018
Dealing with data Building a Smart Laboratory 2018
Dealing with data
Informatics experts share their and they want to know what we could do using experiences on the implementing new that layer,’ said Gonzalez. ‘That is the technical technologies and manging change in question that we get the most from our existing the modern laboratory users. They don’t tend to ask about cloud because they have a running system. The IT department might be interested in moving to the cloud but since they have the system already running and they are not likely to want to change that in the short term,’ Gonzalez added. Gonzalez noted that mobile technology as a solution for laboratory users ‘is a solution that Mark Gonzalez needs to solve real-life problems’. Technical director at Labware ‘What we want to do is solve the right problems we don’t want to just throw out a bunch of technology that doesn’t really solve anything of What technologies are requested by any business value.’ laboratory users? One example that he noted was the ability to use mobile devices in untethered mode. This Mark highlighted that there are clear divisions could allow users to perform actions such as between the two primary groups of existing entering data without a continuous connection customers and potential users. to the LIMS server. Once the connection is ‘In terms of technology the question that re-established the data can be automatically existing users are asking about most often is sent to the LIMS system. ‘One value of mobile mobile. That is not to say that they have a clear technology is that people could work remotely to plan on how to use the technology but they have collect data, even if they don’t have a connection smart phones and tablets in their personal lives to the LIMS server,’ concluded Gonzalez
10 www.scientific-computing.com/BASL2018 Building a Smart Laboratory 2018 Dealing with data
and objectives in order to increase efficiency effort and money, so they aren’t eager to move. and productivity and finally it must facilitate When a Lab needs to be compliant to GMP, GLP, collaboration between scientists. etc it has many other points to manage: change ‘The main issue is to handle data, not only control, system validation, certification and audits.’ to store it but also to be able to use that data All of these aspects can make a move more effectively,’ stressed Acker. ‘Laboratories are challenging, but ultimately choosing not to producing and accumulating more and more upgrade impacts agility – and the speed and data from experiments, analysis, bibliography and quality of further laboratory operations. other areas. For instance, one screening campaign Another aspect that AgiLab was keen to stress could generate hundreds of thousands of results, a was that cloud deployments are increasingly seen query on a citation source like Pubmed can report as a good choice for many laboratories. However, Renaud Acker thousands of references. the move to cloud based informatics requires a Chief operating officer at Agilab ‘The challenge is to centralise data, to manage user to change their mindset as they move from and gather information, to generate knowledge silos of data to a more fluid model of shared data from data – and to keep track of what has been sets and collaboration. What are the main challenges that done, how it has been done, if it has worked or ‘Labs are still working in silos,’ added Acker. your users face when deploying digital not. Big data technologies will be very useful to ‘New R&D processes should break this logic in informatics technology? annotate, explore and exploit the whole set of order not only to exchange data but mostly to data generated in labs and gathered from external anticipate issues by gathering scientists working Renaud Acker explains that, for many of AgiLab’s public sources. on a project. Collaboration is essential for R&D customers, ‘change control’ is the main challenge: While there are clear benefits to using the project success. Cloud applications could help to ‘Processes have changed by using a new latest software, cost of investment can be a big exchange data and ideas between labs in different generation of software. Users must be trained, issue that prevents companies from replacing locations, between industrial, partners and standard operating procedures (SOPs) must be legacy infrastructure – but it is not the only academics.’ adapted, data handling and traceability must be reason, as Acker explains. Acker concluded that cloud-based laboratory managed in a different way. ‘There are at least two main reasons why labs informatics is growing due to a number of factors ‘This means that lab software should be don’t move easily to new lab software. Many including their robust security, the potential for user-friendly for daily use. Screens must be clear companies and labs have spent fortunes in their hosting management of services off-premise and with adapted vocabulary,’ stated Acker. ‘However, first generation of lab software. Secondly, they the use of cloud subscription models that can it must also be adapted to laboratory processes have customised these products with considerable reduce initial investment and running costs.
that iVention is managing in Europe that is solution was hosted for the client by iVention. consolidating as many as seven individual ‘I don’t think there are many of those rollouts implementations with their own custom software, completed successfully with a conventional LIMS with additional software connected to it. system,’ said Kox. ‘They cannot upgrade everything all at once,’ ‘They are a big company with their own he said. IT department and we are hosting it for them The presence of custom software in each because we have all the technology in place to implementation means that each installation is automate everything, so all the upgrades can be essentially a new piece of software. done automatically.’ He explained the success of ‘Now if you compare this to the capabilities of this rollout has meant this company is now using Oscar Kox a web-based system you can rollout to all of those iVention as a strategic partner for much larger Business delopment manager at Ivention sites without custom software – there is a big rollouts in the future. benefit,’ said Kox. Kox said: ‘I have seen organisations with ‘If there is a LIMS project that people who are very old software, which can be costly and time now looking for a new LIMS or ELN, the decision consuming to maintain and upgrade. Some IT How important are digital technologies they make now will affect them for the coming directors would say the upgrade would cost more to the modern laboratory? five to 10 years, because that is the investment than the original installation, so they either try and that you are looking at.’ run for a few more years or select a new system.’ ‘There is a lot of innovation available in the Kox stressed customers should ask He said one of the main challenges when market but I don’t think many labs are picking it themselves: will this big conventional LIMS dealing with legacy LIMS or ELN systems is a lack up as early adopters,’ said Kox. vendor help me to innovate? ‘That is where the of maintenance and upgradability: ‘The biggest ‘People should ask themselves how important gap comes in. There’s a lot of innovation out thing I see is customers paying maintenance and is it adopt new technologies – to innovate in the there but can I adopt it right now, because of the they cannot upgrade. Support cannot help them lab. Having worked in this industry for more than systems I have in place?’ because they have an old version and in many 20 years – of course it is important. You want to He explained that iVention has installed cases this support money is wasted because the see new technology getting into the laboratory systems across very large organisations. He gave system is too old to be properly supported. either because you want to reduce FTE, you want an example of a pharma client who wanted to roll ‘I would strongly recommend firms look at to increase throughput or improve quality.’ out a system for 300 users across seven countries, their maintenance contacts and ask themselves Kox gave an example of large implementation over eight months. Cox also mentioned that this “what are we getting back from it?”’ n
www.scientific-computing.com/BASL2018 11 The smart laboratory Building a Smart Laboratory 2018
The smart laboratory
This chapter discusses what we mean oday the landscape for laboratory a common problem facing many laboratories – by a ‘smart laboratory’ and its role in technologies is broad and varied. This data generated through ‘dumb’ instrumentation an integrated business. We also look is true purely in terms of the variation such as pH meter or weighing scales. Instruments at the development of computerised of management systems and other that are not connected directly to a (Laboratory laboratory data and information software packages but also due to the proliferation Informatics Management System) LIMS or management; the relationships T of additional technology such as cloud, mobile Electronic Laboratory Notebook (ELN) type between laboratory instruments technologies and more recently the IoT. management system present opportunities to and automation (data acquisition); There is no specific definition of a ‘smart introduce error through human data entry but laboratory informatics systems laboratory’. The term is often used in different there are multiple ways to solve this problem. (information management); and higher- contexts to imply either that a laboratory is One would be to buy new scales for example. level enterprise systems and how they designed to optimise its physical layout, that it Purchasing a new instrument with smart align with knowledge management incorporates the latest technology to control the capabilities could feed that data directly into the initiatives. laboratory environment, or that the laboratory is LIMS reducing the chance for error. Another using the latest technology to manage its scientific approach would be the use of mobile devices The progressive ‘digitisation’ of the activities. For the purposes of this publication, it is which could be used to capture the data at the laboratory offers an unprecedented the latter definition that applies. bench another would be to use a raspberry Pi like opportunity not only to increase Using technology to manage scientific device connected to the internet to take the result laboratory efficiency and productivity, endeavours is conceptually a straightforward and feed it into the LIMS. The choice around but also to move towards ‘predictive task but the subtlety lies in choosing the right whether mobile, IoT or new instruments is one science’, where accumulated explicit combination of technologies that can be adapted that can only be answered on a case by case basis knowledge and computer algorithms to suit the use case of a specific laboratory which – there is no one size fits all solution for every can be exploited to bring about greater may be dictated by geography and personnel laboratory. understanding of materials, products, as much as it is driven by the availability of The introduction of industrial R&D and processes technology. As such the right answer to setting laboratories heralded a new era of innovation and up a smart laboratory is not to adopt all possible development dependent on the skills, knowledge technological features but to identify which and creativity of individual scientists. The areas of the laboratory need to be accelerated or evolution has continued into the ‘information improved upon. age’ with a growing dependence on information A simple example of this could be found in technology, both as an integral part of the
12 www.scientific-computing.com/BASL2018 Building a Smart Laboratory 2018 The smart Laboratory
FIG 1 Information structure The two primary areas of technology that apply to a smart laboratory can be broadly categorised as laboratory automation and laboratory informatics. In general, laboratory automation refers to the use of technology to streamline or substitute manual manipulation of equipment and processes. The field of laboratory automation comprises many different automated laboratory instruments, Programmes devices, software algorithms, and methodologies Document management used to enable, expedite, and increase the efficiency and effectiveness of scientific research Projects in labs. Laboratory informatics generally refers to Project management the application of information technology to the handling of laboratory data and information, and Experiments optimising laboratory operations. Laboratory notebook In practice, it is difficult to define a boundary between the two ‘technologies’ but, Interpreted/processed data in the context of this publication, chapter three SDM/LIMS (Data) will provide an overview of laboratory instrumentation and automation, predominantly Raw data data capture. Laboratory instrumentation Chapter four (Information) will look at the four major multi-user tools that fall into scientific process, and as a means of managing model (see Figure 1) that defines the conceptual, the ‘informatics’ category, identifying their scientific information and knowledge. multi-layered relationship between data, similarities, differences and the relationship Laboratory information has traditionally information, and knowledge. between them. Chapters three and four, therefore, been managed on paper, typically in the form The triangle represents the different layers of focus on the acquisition and management of of the paper laboratory notebook, worksheets abstraction that exist in laboratory workflows. data and information, whereas chapter five and reports. This provided a simple and These are almost always handled by different (Knowledge) will provide guidance about the portable means of recording ideas, hypotheses, systems. The ‘experiment’ level is the focal point long-term retention and accessibility of laboratory descriptions of laboratory apparatus and for cross-disciplinary collaboration: the point knowledge through online storage and search laboratory procedures, results, observations, and at which the scientific work is collated and algorithms that aim to offer additional benefits conclusions. As such, the lab notebook served as traditionally handled by the paper laboratory through the re-use of existing information, the both a scientific and business record. However, notebook. avoidance of repeating work, and enhancing the the introduction of digital technologies to the Above the experimental layer is a management ability to communicate and collaborate. laboratory has brought about significant change. context that is handled by established groupware The underlying purpose of laboratory From the basic application of computational and document management tools at the automation and laboratory informatics is to power to undertake scientific calculations at ‘programme’ level, and by standard ‘office’ tools increase productivity, improve data quality, to unprecedented speeds, to the current situation at the project level. Below the experiment level reduce laboratory process cycle times, and to of extensive and sophisticated laboratory there is an increasing specialisation of data types facilitate laboratory data acquisition and data automation, black box measurement devices, and tools, typically encompassing laboratory processing techniques that otherwise would be and multiuser information management instrumentation and multi-user sample and test impossible. Laboratory work is, however, just one systems, technology is causing glassware and management systems. The triangle also represents step in a broader business process – and therefore, paper notebooks to become increasingly rare the transformation of data to knowledge, the in order to realise full benefit from being ‘smart’, in the laboratory landscape. The evolution of journey from data capture to usable and reusable it is essential that the laboratory workflow is sophisticated lab instrumentation, data and knowledge that is at the heart of the smart consistent with business requirements and is information management systems, and electronic laboratory. integrated into the business infrastructure in order record keeping has brought about a revolution The introduction of ELNs therefore opens up for the business to achieve timely progress and in the process of acquiring and managing the possibility of a more strategic approach, which, remain competitive. laboratory data and information. However, the in theory at least, offers the opportunity for an Chapter seven (Beyond the laboratory) will underlying principles of the scientific method integrated and ‘smart’ solution. examine the relationship between laboratory are unchanged, supporting the formulation, A frequently articulated fear about the processes and workflows with key business testing, and modification of hypotheses by relentless incorporation of technology in issues such as regulatory compliance and means of systematic observation, measurement, scientific processes is the extent to which it can patent evidence creation, and will also address and experimentation. In our context, a smart de-humanise laboratory activities and reduce productivity and business efficiency. laboratory seeks to deploy modern tools and the demand for intellectual input, or indeed, any Chapter eight (Practical considerations in technologies to improve the efficiency of fundamental knowledge about the science and specifying and building the smart laboratory) the scientific method by providing seamless technology processes that are in use. The objective is therefore devoted to the process of making integration of systems, searchable repositories of this publication is to present a basic guide to the the laboratory ‘smart’, taking into account the of data of proven integrity, authenticity and most common components of a ‘smart laboratory’, functional needs and technology considerations reliability, and the elimination of mindless and to give some general background to the benefits to meet the requirements of the business, and unproductive paper-based processes. they deliver, and to provide some guidance to how addressing the impact of change on laboratory At the heart of the smart laboratory is a simple to go about building a smart laboratory. workers. n www.scientific-computing.com/BASL2018 13 Data: Instrumentation Building a Smart Laboratory 2018
DATA Instrumentation
This chapter will consider the Simple laboratory instruments technologies in instrumentation significantly different classes of instruments and improves both their utility and the labs’ workflow. computerised instrument systems to evices such as analytical balances be found in laboratories and the role and pH meters use low-level Computerised instrument systems they play in computerised experiments processing to carry out basic and sample processing – and the functions that make them easier to The improvement in workflow becomes more steady progress towards all-electronic work with. The tare function on a balance avoids evident as the level of sophistication of the laboratories. D a subtraction step and makes it much easier to software increases. It is rare to find commercial However, the choice of best-of-breed weigh out a specific quantity of material. instrumentation that doesn’t have processing laboratory instruments and instrument Connecting them to an electronic lab capability either within the instruments’ systems can present challenges when notebook (ELN), a laboratory information packaging or, through a connection to an it comes to getting everything to work management system (LIMS), a lab execution external computer system. together in a seamless way. The final system (LES), or a robot, adds computer- The choice of dedicated computer-instrument part of this chapter will look at the issue controlled sensing capability that can significantly combinations vs. multi-user, multi-instrument of standard data interchange formats, off-load manual work. Accessing that balance packages is worth careful consideration. The most the extent of the challenge, and some of through an ELN or LES permits direct insertion common example is chromatography, which has the initiatives to address them of the measurement into the database and avoids options from both instrument vendors and third- the risk of transcription errors. In addition, the party suppliers. informatics software can catch errors and carry One of the major differences is data access out calculations that might be needed in later and management. In a dedicated format, each steps of the procedure. computer’s data system is independent and has The connection between the instrument and to be managed individually, including backups to computer system may be as simple as an RS-232 servers. connection or USB. Direct Ethernet connections It also means that searching for data may or connections through serial-to-Ethernet be more difficult. With multi-user/instrument converters can offer more flexibility by permitting systems there is only one database that needs to access to the device from different software be searched and managed. systems and users. The inclusion of smart If you are considering connecting the systems
14 www.scientific-computing.com/BASL2018 Standardize Analytical Data across techniques and vendors in a single informatics platform
Process, review, and store data in context Assure data integrity Provide live, on-demand access Enhance regulatory compliance Simplify and expedite knowledge sharing Drive innovation ACD/Spectrus Platform Integrate live, standardized analytical data into your IT stack
www.acdlabs.com/Standardize
IVENTION LAB EXECUTION SYSTEM MAIL OR CALL US FOR A FREE PILOT MORE THAN WITH PRODUCT DEMONSTRATION EUROPE LIMS WITH LES [email protected] +31 (0)38 4528375
UNITED STATES [email protected] A FRESH APPROACH IN LAB AUTOMATION +1 (502) 909 0317 LIMS / ELN / (S)DMS TECHNOLOGY FOR MOBILE / APP / WEB STREAMLINE WORKFLOWS / FEWER CLICKS FREE AUTOMATED UPGRADES
ivention.nl
iVention_Adv_213x135mm.indd 2 26-11-15 09:16 Data: Instrumentation Building a Smart Laboratory 2018
to a LIMS or ELN, make the connections as FIG 2 Analogue data acquisition simple as possible. If an instrument supported by the software needs to be replaced, changing the connection will be simpler. Display Licence costs are also a factor. Dedicated formats require a licence for each system. Shared- Electrical Property access systems have more flexible licensing circuit to be considerations. Some have a cost per user and converting A/D Control Communications measured properly to processor connected instrument; others have a cost per (detector) active user/instrument schedules. voltage In the latter case, there are eight instruments and four analysts, of which only half may be Product Digital I/0 simultaneously active, licenses for only four packaging (switches, LEDs, etc.) instruments and two users are needed. One factor that needs attention is the education of laboratory staff in the use of computer-instrument systems. While instrument software systems are Instrument data management are created – and transfer the format to that capable of doing a great deal, their ability to needed by the instrument. Robotic arms – still function is often governed by user-defined The issue of instrument data management is appropriate for many applications – have been parameters that affect, at least in chromatography, a significant one and requires considerable replaced with components more suitable to the baseline-corrections, area allocation for planning. Connecting instruments to a LIMS task, particularly where liquid handling is the unresolved peaks, etc. Carefully adjusted and or ELN is a common practice, though often dominant activity, as in life science applications. tuned parameters will yield good results, but not an easy one if the informatics vendor Success in automating sample preparation problems can occur if they are not managed and hasn’t provided a mechanism for interfacing depends heavily on thoroughly analysing the checked for each run. equipment. Depending on how things are set process in question and determining: up, only a portion of the information in the n Whether or not the process is well instrument data system is transferred to the documented and understood (no Type of A/D Capability informatics system. undocumented short-cuts or workarounds Successive These are general- If the transfer is the result of a worklist that are critical to success), and whether approximation purpose devices suitable execution of a quantitative analysis, only the improvements or changes can be made for a wide range of final result may be transferred – the reference without adversely impacting the underlying applications. They have limited resolution, data still resides on the instrument system. The science; but have amplifiers for result is a distributed data structure. In regulated n Suitability for automation: whether or not low-level signals, and environments, this means that links to the backup there are any significant barriers (equipment, can sequentially access information have to be maintained within the etc.) to automation and whether they can be multiple input channels. Their resolutions are LIMS or ELN, so that it can be traced back to the resolved; up to 18 bits (262,144 original analysis. n That the return in investment is acceptable steps) and sampling The situation becomes more interesting and that automation is superior to other speeds of up to five when instrument data systems change or are alternatives such as outsourcing, particularly million samples per retired. Access still has to be maintained to for shorter-term applications; and second (sps). The higher n the resolution, the slower the data those systems hold. One approach is That the people implementing the project the sampling speed. virtualising the instrument data system so that the have the technical and project management operating system, instrument support software, skills appropriate for the work. Integrating Good for low speed and the data are archived together on a server. The tools available for successfully implementing sampling (<100 sps), high resolution >14 (Virtualisation is, in part, a process of making a a process are clearly superior to what was bits, single channel copy of everything on a computer so that it can available in the past. Rather than having a robot inputs, with good noise be stored on a server as a file or ‘virtual container’ adapt to equipment that was made for people rejection. Often used in and then executed on the server without the to work with, equipment has been designed chromatography. need for the original hardware. It can be backed for automation – a major advance. In the life Sigma-Delta Up to 24 bits of up or archived, (so that it is protected from loss). sciences, the adoption of the microplate as a A/D resolution, single channel In the smart laboratory, system management is standard format multi-sample holder (typically input – may not be a significant function – one that may be new to 96 wells, but can have 384 or 1,536 wells – denser efficient for multi-channel inputs, low speed, may many facilities. The benefits of doing it smartly forms have been manufactured) has fostered replace integrating A/Ds. are significant. the commercial availability of readers, shakers, washers, handlers, stackers, and liquid additions Flash Single channel input, Computer-controlled experiments and systems, which makes the design of preparation 8-bit conversion, approximately 1 billion sample processing and analysis systems easier. Rather than SPS. Good for very processing samples one at a time, as was done in high-speed applications, Adding intelligence to lab operations isn’t limited early technologies, parallel processing of multiple where low resolution is to processing instrument data, it extends to an samples is performed to increase productivity. not a problem. You can earlier phase of the analysis: sample preparation. Another area of development is the ability to digitise electrical noise. Robotic systems can take samples – as they centralise sample preparation and then distribute
16 www.scientific-computing.com/BASL2018 Building a Smart Laboratory 2018 Data: Instrumentation the samples to instrumentation outside the processed by the instrument would wait until examples of integration methods that enabled sample prep area through pneumatic tubes. the data system told it to go ahead. The LIMS the user to extend the basic capability and have This technology offers increased efficiency by has the expected range for valid results and the ready access to a third-party market of useful putting the preparation phase in one place so that acceptable limits. If a result exceeded the range, components. It also allowed the computer solvents and preparation equipment can be several things could happen: vendors to concentrate on their core product and easily managed, with analysis taking place n The analyst would be notified; satisfy end-user needs through partnerships; each elsewhere. This is particularly useful if safety is n The analysis system would be notified that vendor could concentrate on what they did best an issue. the test should be repeated; and the resulting synergy gave the users what they Across the landscape of laboratory types n If confirmed, standards would be run needed. and industries, the application of sample to confirm that the system was operating Now these traditional methods are being preparation robotics is patchy at best. Success and properly; and surpassed by the IOT or wireless connected commercial interest have favoured areas where n If the system were not operating according devices but the argument for connecting devices standardisation in sample formats has taken to SOPs, the system would stop to avoid still remains the same – is the value added worth place. wasting material and notify the analyst. the investment? The answer depends on the The development of microplate sample The introduction of a feedback facility would instrument, but generally it is more effective to formats, including variations such as tape systems significantly improve productivity. connect the most widely used instruments such that maintain the same sample cell organisation At the end of the analysis, any results that are as PH meters and weighing scales. in life sciences, and standard sample vials for outside expected limits would have been checked Connections are only part of the issue. The more significant factor is the structure of the data that is being exchanged: how it is formatted; and the organisation of the content. In the examples Building a smart laboratory needs to look beyond commonplace above, that is managed by the use of standard approaches and make better use of the potential that exists in device drivers or, when called for, specialised informatics technologies device handlers that are loaded once by the user. In short, hardware and software are designed for integration, otherwise vendors find themselves at a disadvantage in the marketplace. “ Laboratory software comes with a different auto-samplers, are common examples. Standard and the systems integrity verified. Making this mindset. Instrument support software was sample geometries give vendors a basis for happen depends on connectivity and the ability designed first and foremost to support the successful product development if those products to integrate components. vendor’s instrument and provide facilities that can have wider use rather than being limited to weren’t part of the device, such as data analysis. niche markets. Instrument integration Integration with other systems wasn’t a factor. That is changing. The increasing demand Putting the pieces together In order for the example described above to for higher productivity and better return work, components must be connected in a way on investment has resulted in the need for It’s not enough to consider in isolation sample that permits change without rebuilding the systems integration to get overall better systems preparation, the introduction of samples into entire processing train from scratch. Information performance; part of that measure is to reduce instruments, the instruments themselves, and technology has learned those lessons repeatedly the need for human interaction with the system. the data systems that support them. Linking as computing moved from proprietary products Integration should result in: them together provides a train of tasks that can and components to user friendly consumer Ease-of-use: integrated systems are lead to an automated sample processin systems. expected to take less effort to get things done; system as shown in Figure 3. Consumer level systems aren’t any less capable Improved productivity, streamlined The control/response link is needed to than the earlier private-brand-only systems, they operations: the number of steps needed to synchronise sample introduction and data are just easier to manage and smarter in design. accomplish a task should be reduced; acquisition. Depending on the nature of the Small Computer Systems Interconnect, n Avoiding duplicate data: no need to look work, that link can extend to sample preparation. Firewire and Universal Serial Bus are a few in multiple places; The end result is a system that not only provides higher productivity than manual methods, but does so with reduced operating costs (after the An automated sample processing system initial development investment). FIG 3 However, building a smart laboratory needs to look beyond commonplace approaches and Control/response make better use of the potential that exists in informatics technologies. Extending that train of elements to include a LIMS, for example, has additional benefits. The initial diagram above Instrument would result in a worklist of samples with the test results that would be sent to a LIMS for incorporation into its database. Suppose there was a working link between Data acquisition, analysis, a LIMS and the data system that would send Sample Sample reporting, storage, etc sample results individually, and that each sample preparation introduction www.scientific-computing.com/BASL2018 17 Data: Instrumentation Building a Smart Laboratory 2018 n Avoiding transcription errors: integration n In the 1990s, efforts by instrument vendors standard specification under the ASTM (www will result in electronic transfers that should\ led to the development of the andi standards astm.org/DATABASE.CART/WORKITEMS be accurate; this avoids the need to enter and (Analytical Data Interchange) which WK23265.htm) that is designed to be widely verify data transfers manually; resulted in ASTM E1947 – 98(2009) Standard applicable to instrument data. Initial efforts n Improving workflow and the movement Specification for Analytical Data Interchange are planned to result in implementations for of lab data: reducing the need for people to Protocol for Chromatographic Data, which chromatography and spectroscopy. make connections between systems – uses the public domain netCDF data base A concern with the second, third and fourth integration facilitates workflow; and structure, providing platform independence. points above is that they are primarily aimed n More cost-effective, efficient lab operations. This standard is supported in several vendor at the pharmaceutical and biotech industries. The problem of integration, streamlining products but doesn’t see widespread use; While vendors will want to court that market, operations, and better productivity has been n SiLA Rapid Integration (www the narrow focus may slow adoption since it addressed via automation before: in sila-standard.org). The website states: ‘The could lead to the development of standards manufacturing applications and clinical labs. SiLA consortium for Standardisation in for different industries, increasing the In the 1980s, clinical lab managers recognised the Lab Automation develops and introduces implementation and support costs. Common only way they were going to meet their financial new interface and data management standards, issues across industries and applications lead to objectives was to use automation to its fullest allowing rapid integration of lab automation common solutions. capability and drive integration within their ystems. SiLA is a not-for-profit membership The issue of integrating instruments with systems. corporation with a global footprint and is informatics software is not lost on the vendors. The programme came under the title ‘Total open to institutions, corporations and Their product suites offer connection capabilities laboratory automation’ and resulted in a series of individuals active in the life science for a number of instrument types to ease the work. Planning a laboratory’s information handling requirements should start from the most critical point, a LIMS for example, and then on to In the 1980s, clinical lab managers recognised that the only way they were support additional technologies. going to meet their financial objectives was to use automation to its fullest capability and drive integration within their systems Chapter summary
The transition from processing samples and experiment manually to the use of electronic “ systems to record data is a critical boundary. It moves from working with real things to standards that allowed instrument data systems lab automation industry. Leading system their digital representation in binary formats. to connect with Laboratory Information Systems manufacturers, software suppliers, system Everything else in the smart laboratory (equivalent to LIMS) and hospital administrative integrators and pharma/biotech corporations depends on the integrity and reliability of that systems. have joined the SiLA consortium and transformation. Those standards were aggregated under contribute in different technical work groups The planners of laboratory systems may HL7 (www.hl7.org), which provides both with their highly skilled experts’; never have to program a data acquisition message and data formatting. While hospital n The Pistoia Alliance (www.pistoiaalliance system, but they do have to understand how and clinical systems have the advantages of a org) states: ‘The Pistoia Alliance is a global, such systems function, and what the educated more limited range of testing and sample types, not-for-profit precompetitive alliance of life lab professional’s role is in their use. Such making standardisation easier, there is nothing science companies, vendors, publishers and preparatory work will enable the planners to take in their structure to prevent them being applied academic groups that aims to lower barriers full advantage of commercial products. to a wider range of instruments, such as mass to innovation by improving the One key to improving laboratory spectrometry. interoperability of R&D business processes. productivity is to develop an automated process An examination of the HL7 structure suggests We differ from standards groups because we for sample preparation, introducing the sample that it would be a good foundation for solving bring together the key constituents to identify into the instrument, making measurements, integration problems in most laboratories. the root causes that lead to R&D inefficiencies and then forwarding that data into systems for From the standpoint of data transfer and and develop best practices and technology storage, management, and use. Understanding communications, the needs of clinical labs match pilots to overcome common obstacles’; the elements and options for these systems is those in other areas. The major changes would n The Allotrope Foundation (www the basis for engineering systems that meet the be in elements, such as the data dictionaries, and allotrope.org): ‘The Allotrope Foundation needs of current and future laboratory work. field descriptions, which are specific to hospital is an international association of The development of the smart laboratory is at and patient requirements. biotech and pharmaceutical companies a tipping point. As users become more aware of A cross-industry solution would benefit building a common laboratory what is possible, their satisfaction with the status vendors as it would simplify their engineering information framework (‘Framework’) for quo will diminish as they recognise the potential and support, provide a product with wider an interoperable means of generating, storing, of better-designed and integrated systems. market appeal, and encourage them to implement retrieving, transmitting, analysing and Realising that potential depends on the it as a solution. archiving laboratory data and higher-level same elements that have been successful in Most of the early standards work carried out business objects such as study reports and manufacturing, computer graphics, electronics, outside the clinical industry has focused on data regulatory submission files’; and and other fields: an underlying architecture for encapsulation, while more recent efforts have n The AnIML markup language for analytical integration, based on communications and data included communications protocols: data (animl.sourceforge.net) is developing a encapsulation/interchange standards. n
18 www.scientific-computing.com/BASL2018 Chapter summary
Building a Smart Laboratory 2018 Information: Laboratory informatics tools
INFORMATION Laboratory informatics tools
This chapter will look at the four aboratory informatics is the corresponds to an application-centric major laboratory informatics tools – specialised application of portfolio of ‘systems’ that were not laboratory information management information technology aimed at necessarily designed to work together, and systems (LIMS), electronic laboratory optimising laboratory operations by for which interoperability is hampered notebooks (ELNs), laboratory execution the application of information technology by the lack of standards and so has to systems (LES) and scientific data L to the handling of laboratory data and be customised. A smart laboratory is an management systems (SDMS) – their information. It encompasses four major ‘integrated’ laboratory that is modular, based differences and how they relate to each multi-user systems: laboratory information on standards, and is designed to facilitate other. Each of these systems functions management systems (LIMS), electronic connectivity, data sharing and collaboration. at or around the ‘Information’ layer laboratory notebooks (ELNs), laboratory Over the past few years, the informatics (see Figure 1) and typically serves to execution systems (LES) and scientific data market has undergone two interesting collate data and information about the management systems (SDMS). developments; firstly, the previously separate laboratory’s operations There is a very good reason why the LIMS and ELN sub-markets have started use of a generic term such as ‘laboratory to overlap, causing a certain amount of informatics’ is important: we need to confusion to the application-centric mind- get away from an application-centric set; secondly, mergers and acquisitions have approach and think of a fully integrated reshaped the vendor line-up, specifically in laboratory and its interaction with other the ELN field. company systems. The deployment of an The origins of the LIMS market can ELN generally represents the final step in be traced back several decades to the making a laboratory fully electronic, and point where the increasing prevalence of hence raises the demand to connect up all computers in the laboratory, coupled with laboratory systems. Being fully electronic their increasing processing power, led and being fully integrated are two different enterprising scientists to develop simple, things. custom computerised workflow systems to For most labs, being fully ‘electronic’ operate in conjunction with data acquisition www.scientific-computing.com/BASL2018 19 The initial evolution of the ELN market was centred on the provision of functionality to support small molecule chemistry
Information: Laboratory informatics tools Building a Smart Laboratory 2018 and data processing. In the early 1980s, first-generation commercial LIMS started to appear, usually based on minicomputers, The initial evolution of the ELN market was centred on the provision of supporting sample and test management, functionality to support small molecule chemistry and reporting of results. A second generation of commercial LIMS started to appear in the late 1980s, typically taking advantage of relational “ databases to provide more sophisticated functionality. The development of client- additional requirements gradually became and electronic laboratory notebooks (ELNs). server based systems represented the apparent. Firstly, there was a need to Functionally, the LIMS products have next (third) generation of commercial transfer data from laboratory instruments become increasingly sophisticated, to the systems, taking advantage of the evolution directly to the LIMS, to avoid transcription point that the dividing line between LIMS of the personal computer. The fourth errors; secondly, the need to manage the and other informatics products has become generation emerged as the internet and instrument data files from which data stored less clear. wireless connectivity developed, offering in the LIMS was derived; and thirdly, the The ELN market has grown and opportunities to extend the reach of LIMS need to handle unstructured data, graphical developed rapidly over the past decade, it beyond the confines of the laboratory. data, and to collate sample data. These still exhibits some instability with a large As LIMS products were increasingly requirements led to the development of number of vendors (there are more than 30 adopted by laboratories, three specific scientific data management systems (SDMS) purveyors of products that purport to be an ELN) competing for market share. As a consequence, the market suffers from some The Gartner Hype Cycles degree of ‘hype’ (see Figure 4). FIG 4 Just where ELNs sit on the Gartner Hype Cycle[1] is probably somewhere around the ‘Trough of Disillusionment’, although Visibility individual vendors may occupy positions either side of this point. The ‘Trough of Disillusionment’ can be considered as the turning point past the hype and when the focus is on delivering true benefit. Chemistry-based and generic ELNs are probably already beyond this point, as indeed are the majority of LIMS products. Commercial ELNs have evolved from two approaches: discipline-specific; and generic. Generic software provides the architecture and tools to create and search content, and to work collaboratively in a way that satisfies the needs of almost any science- related industry. Discipline-specific ELNs are aimed at a particular market segment, Technology Peak Trough of Slope of Plateau of trigger of inflated disillusionment enlightenment productivity such as chemistry, biology, or analytical. expectations These systems are usually tailored to work Maturity with other discipline-specific software tools. Most of the commercial ELNs offer a combination of generic and discipline- Technology trigger: The first phase of a Hype Cycle is the ‘technology trigger’ or specific functionality. breakthrough, product launch or other event that generates significant interest. The initial evolution of the ELN market Peak of inflated expectations: In the next phase, a frenzy of publicity typically generates was centred on the provision of functionality over-enthusiasm and unrealistic expectations. There may be some successful applications of a to support small-molecule chemistry. Most technology, but there are typically more failures. of the experimental processes associated Trough of disillusionment: Technologies enter the ‘trough of disillusionment’ because they with synthetic chemistry are well established, fail to meet expectations and quickly become unfashionable. Consequently, the press usually reasonably consistent, and are well supported abandons the topic and the technology. by desktop software tools. Integrating these Slope of enlightenment: Although the press may have stopped covering the technology, functions into an ELN that can create, some businesses continue through the ‘slope of enlightenment’ and experiment to understand manage and store a full experimental record the benefits and practical application of the technology. was a logical progression. As a consequence, chemistry-based ELNs are well established Plateau of productivity: A technology reaches the ‘plateau of productivity’ as its benefits and exhibit a good deal of maturity. If there become widely demonstrated and accepted. The technology becomes increasingly stable and evolves in second and third generations. The final height of the plateau varies according to is segmentation in this part of the market, whether the technology is broadly applicable or benefits only a niche market. it is determined to some extent by the origins and scope of the available products.
20 www.scientific-computing.com/BASL2018 Building a Smart Laboratory 2018 Information: Laboratory informatics tools
workflow requirements. Some organisations have chosen to implement generic ELN functionality within the framework of their standard IT tools, such as Lotus Notes and SharePoint. In the academic community, blogging tools have been used to record experimental work and thus provide the basic features of an ELN, with a strong emphasis on sharing and collaboration and in the form of a laboratory journal. The convergence in the informatics market is now confusing potential customers. The table, right, identifies the core differences in the major tools. Initially, each of these tools addressed a well-defined, functional requirement, but the increasing level of sophistication of the underlying information technologies has made it easier to extend functionality in ways that mean that there is now considerable overlap between the different tools. At one stage it was considered unlikely that a single ELN could provide the necessary functionality to support chemistry, biology and analytical requirements. Those days are over, and this should make the task of finding a suitable ELN easier. But the extent of the overlap with LIMS, SDMS, and LES can generate confusion, and for someone looking to address laboratory information management requirements, the task seems to be more challenging.
ELN Experiment-centric: an authoring tool that handles unstructured Some, for example, will be an enterprise- another ELN domain has emerged, that data and offers generic and wide solution, others will focus on utility of QA/QC and the regulatory world. A few specific functionality to support and personal productivity, while others vendors have concentrated specifically different scientific disciplines. Supports IP protection, will provide a generic ELN capability that on this area, with products that are knowledge re-use, productivity integrates third-party software tools. strongly aligned to laboratory workflows, and collaboration. Biology, however, has presented a bigger following the step-by-step execution of challenge to the ELN vendors. The diverse SOPs or test methods. LES Procedure or experiment- and complex nature of biological processes The products are more structured than centric: basically able to and outcomes creates a need to capture not a ‘conventional’ ELN and in some respects handle structured data and just the data, but also the interrelationships appear to be functionally closer to a LIMS. some unstructured data. between the data. This, coupled with a This particular segment of the market has Specifically designed to meet the requirements of the GxP diverse portfolio of biology-specific software seen a number of LIMS vendors extending environment. Simplifies tools, begs the question: do biologists just the functionality available in their LIMS repeated operations. Supports need a generic ELN that will integrate with products to embrace some of the more electronic SOPs. their existing software tools, or do they unstructured requirements associated with need a complete suite of functionality that experimentation. It could be argued that LIMS Sample-centric: primarily is embedded in the ELN? The issue for the such products may be better labelled as designed to handle structured biologists is whether there is a commercial laboratory execution systems (LES) as data, and offers sample and test management, batch ELN that addresses their specific and diverse they follow a very prescriptive approach operations, and industry-specific requirements. Furthermore, for those applicable to those communities engaged in workflows. Secure laboratory companies that need to support chemists regulatory based testing. information hub. Supports and biologists, the question is whether it To replace a paper notebook, all that compliance. is possible to find a single vendor solution could be required could be a simple that addresses the requirements of both authoring tool capable of generating a SDMS Data-centric: handles data files disciplines, or whether to choose the best of compound-document. However, additional from laboratory instruments, breed for each discipline. capability will be needed for storing and meta-data, documents, and the relationships between them. Within the past two or three years, searching documents, and for addressing www.scientific-computing.com/BASL2018 21 Information: Laboratory informatics tools Building a Smart Laboratory 2018
What is a laboratory workflows by eliminating errors due to Most laboratory processes have evolved over information management system manual data entry and transcription time to meet local laboratory requirements (LIMS)? errors. This is achieved through interfacing rather than being specifically designed to laboratory instruments to the LIMS for meet wider organisational requirements. Any A laboratory information management system two-way communication of sample IDs, LIMS implementation must simplify and (LIMS) provides the basic functions for worklists, and results, and by integration streamline the process rather than automate sample and test management, and has become with other laboratory systems such as an inefficient, paper-based status quo. the standard tool for analytical and QC electronic laboratory notebooks (ELNs) The commercial systems on the laboratories for registering samples, assigning and scientific data management systems marketplace have become increasingly tests, gathering and managing results, and (SDMS). sophisticated over the years. The major issuing reports. Most LIMS now provide a A LIMS also acts as a major repository challenge in choosing a LIMS is identifying more integrated solution to support workflows of the records of analytical testing and can how an out-of-the-box solution is aligned and processes customised to a range of be a source of historical data associated with to the organisation’s needs. Most systems industry-specific requirements. the organisation’s products and production are highly configurable and avoid the need The basic functions to be found in a processes. In addition, the transactional for any custom code to be written to meet LIMS are: nature of a LIMS enables a secondary record specific requirements. n The registration of samples and system to be maintained as an audit trail associated data, such as provenance, to track date, time, user – and, if necessary, What is a scientific data customer, due dates, etc.; what change was made within the system. management system (SDMS)? n The assignment of tests to the sample; This data may then be used to satisfy quality n Scheduling and tracking of the sample assurance requirements in terms of data A scientific data management system (SDMS) and tests; integrity, and can also be used to generate a is, in its basic form, a database application n Recording the test procedure, equipment wide variety of management reports on the that manages electronic records generated by and materials used during testing; laboratory’s performance. laboratory instruments. Typically, an SDMS n The review, approval, and aggregation A pre-requisite before implementing a will provide long-term data preservation, of test results for the sample, including LIMS, or indeed any major computerised accessibility and retrieval. It is complementary specification checking ; and system, is to map and optimise the to other laboratory informatics systems, such n The preparation and communication of laboratory processes that the LIMS as LIMS and ELNs, in the sense that it can customer reports. will automate. The laboratory needs to provide a common repository for experiment- The major business benefits of a LIMS are understand the process and to identify any and sample-related data files. In this way typically associated with more efficient bottlenecks and their underlying causes. it provides a more consistent approach to managing laboratory data than local repositories, and off-line media (CDs, DVDs, tape, etc.) The lines between a LIMS, ELN and an SDMS are at times blurred through the incorporation of additional features to complement the core functionality. An SDMS is a means of collecting data files from a wide range of laboratory instruments and storing them, along with metadata, in a uniform way in a database; in other words, it is a laboratory content management system. By adding workflow elements and providing facilities for the management and storage of other documents associated with laboratory operations (worksheets, SOPs, safety information, reports, PDFs, office documents, images, etc.), an SDMS can in practice evolve into a more comprehensive single informatics solution for some laboratories. However, an SDMS is essentially an ‘event-driven’ system that gathers data, which may limit some of its capabilities relative to the other informatics tools, and is therefore more frequently seen as a system that is complementary to a LIMS or an ELN. Nevertheless, the principle on which the SDMS is based is that it aggregates records into a logical collection associated with a specific entity such as a programme, project, experiment, product, or sample, to provide a readily accessible collection of relevant
22 www.scientific-computing.com/BASL2018 Building a Smart Laboratory 2018 Information: Laboratory informatics tools information. Embedded into an SDMS will be configured with appropriate data entry FIG 5 Broad vs. deep also be the means to provide appropriate points, with data checking; and unique security of the records by means of access workflows can be mapped to support control, audit trail, authorisation, and change repetitive and routine procedures. management. As with other laboratory informatics Broad functions systems, the underlying information Records, patents, What is a laboratory technologies can extend an LES to a broader cross-discipline collaboration execution system (LES)? range of capabilities. For this reason, the LES can, in some cases, serve as an alternative Fig. 5: Broad vs. deep A laboratory execution system (LES) sits to a LIMS, an ELN, or an SDMS. As with somewhere between an ELN and a LIMS each of the major laboratory informatics in terms of the functionality it delivers, tools, purchasing and implementation but its existence is typically targeted at decisions require a thorough understanding analytical service and quality control of the laboratory’s functional requirements. laboratories where high-volume workflows However, it is more likely that the LES will and regulatory compliance are primary be seen as complementary to ERP and QM business requirements. In a very basic sense, systems where high-throughput QA is an the underlying logical structure of an LES essential step in a business process. Analysts Chemists is almost identical to a LIMS, but the user Biologists interface is procedure-centric, rather than What is an electronic Other discipline Other discipline the usual sample-centric approach found in laboratory notebook (ELN)? a LIMS. This allows a standard laboratory operating procedure (SOP) to be executed In its simplest form, an electronic laboratory in an automated way, usually by interaction notebook can be considered to be a direct with laboratory instruments interfaced to replacement for the paper lab notebook. the LES (where possible) in order to capture In this instance, it can provide the generic data without the need for transcription. functionality (‘paper on glass’) to support Calculations on the captured data can be scientific documentation for patent evidence, addressed by the ‘broad’ layer, whereas the performed in the system, and thus the cross-discipline collaboration, and general discipline-specific functionality penetrates automated approach can eliminate two record keeping. However, the integration the ‘interpreted/processed data’ layer in potential sources of error. The concept of a capabilities raise the possibility of a tighter Figure 1. ‘paperless lab’ is a specific objective of the coupling of other laboratory systems into From a patent perspective, the LES, eliminating the use of paper either for the ‘electronic laboratory notebook’. In other ‘experimental layer’ of Figure 1 is crucial intermediate recording of data, or for longer words, can the information that is currently as it captures what the scientist is thinking term record keeping and archival purposes. printed from other laboratory systems, cut and doing, and therefore will provide the The LES is designed to adhere to out and pasted into the paper lab notebook, evidence of conception and reduction laboratory workflows and provides a more be electronically entered or linked directly to to practice of the ‘invention’. In broader repeatable and structured approach to the electronic laboratory notebook? intellectual property (IP) terms, it is the ‘experiment’ layer that constitutes a record of the laboratory’s work and as such contributes to the scientific knowledge repository. For as long as this repository resides on In the academic community, blogging tools have been used to record paper, the ability to access, collaborate and experimental work and thus provide the basic features of an ELN share scientific knowledge is constrained. The implementation of an ELN therefore offers a significant opportunity to bring about greater efficiencies. “ But a clearly defined understanding of the role that the ELN is going to play in a quantitative testing procedures to help For example, systems that provide given organisation is absolutely essential ensure compliance. The user interface chemical structure drawing, structure and at the start of an electronic laboratory usually takes the form of an electronic sub-structure searching, and compound notebook project. As discussed, an equivalent to the paper version of a registration are an integral part of the electronic laboratory notebook supports laboratory standard operating procedure chemistry laboratory’s process, and therefore the ‘experiments’ layer, and also contains or worksheet. This type of interface is often would be expected to become part of an abstractions from the lower data levels (see referred as ‘paper on glass’, a term also electronic solution. Similarly, other scientific Figure 1). used for a generic electronic laboratory disciplines will have specific requirements So the CENSA[2] definition of an notebook. Most LES applications can be consistent with their particular laboratory electronic laboratory notebook as ‘a system readily configured to support alternative processes. to create, store, retrieve and share fully laboratory workflows in a way that relates Figure 5 illustrates the relationship electronic records in ways that meet all closely to traditional paper-based processes. between ‘broad’ (generic) and ‘deep’ legal, regulatory, technical and scientific Worksheets can be converted to electronic (specific) systems. In this context, the requirements’ is all encompassing and can forms; standard operating procedures can ‘notebook’ functionality (see Figure 1) is mean different things to different people. www.scientific-computing.com/BASL2018 23 Information: Laboratory informatics tools Building a Smart Laboratory 2018
Configuration versus customisation
The difference between customisation and configuration is very simply the difference between writing additional code and setting (configuring) in-built parameters in order to achieve some desired functionality. Customisation is generally considered a poor choice as it increases costs, complexity, and risk, and makes it more difficult and more expensive to upgrade software in the future. In a regulated environment, custom code will require extra validation steps. It may often be a symptom of bigger problems, including a mismatch with a company’s requirements or a lack of project controls during implementation. Most laboratory informatics systems are designed to be configurable, and a major activity during implementation is to undertake the entire required configuration to meet functional requirements. Once configured, system upgrades will automatically carry through existing configuration.
An ELN can serve the organisation in three ways: it can take advantage of the capabilities of IT to improve the The wide range of commercial systems on the marketplace has become ability to acquire, manipulate, share and increasingly sophisticated over the years store data (productivity); it can facilitate communication and sharing in real-time across multi-disciplinary and multi-site teams (collaboration); it can provide a “ scientific knowledge repository that can be easily accessed to recover records of the laboratory’s work (content/knowledge of lawyers and patent attorneys to gamble talking about an electronic laboratory rather management). on the legal acceptance of electronic than an electronic laboratory notebook. The way in which laboratory notebooks records in patent interferences and patent are used is largely dictated by the United litigation without any case law; and the Chapter summary States’ patent system which, unlike the rest lack of confidence in our ability to preserve of the world, is based on ‘first to invent’. The electronic records over several decades. The four major laboratory informatics need to be able to demonstrate who really One of the challenges to a successful systems serve different basic functional was first to invent requires the laboratory ELN implementation is identifying exactly requirements, but convergence and notebook to be an authentic and trustworthy what role the ELN will play. The term increasingly sophisticated technologies are record that describes the concept and ‘electronic laboratory notebook’ is inherently creating a good deal of overlap between the its reduction to practice, and for it to be ambiguous. In most cases, the ELN is systems. signed by the author and corroborated by expected to do more than just replace the So when it comes to choosing the right an impartial witness. There are two factors paper lab notebook. The paper lab notebook solution, it’s far better to start by defining why the migration away from paper lab is a simple authoring tool, and any electronic an objective or describing the problem to be notebooks has taken so long: the reluctance authoring tool capable of generating solved, rather than placing the initial focus a compound document will serve as a on a ‘solution’. replacement. Just deciding ‘we need an ELN’ or ‘we An ELN system (like the bound For some companies this has proven to need a LIMS’ should not be the starting laboratory notebook) has several roles: be the case. The combination of Microsoft point; it’s far better to think about the big Office, SharePoint services and a means picture, i.e. the end-to-end business process A place to do science – a working environment; of preserving documents (e.g. in PDF – that embraces the role of the laboratory, portable document format) has proven to the specific workflows, the communication A place to write up the experimental work; be an adequate replacement for paper. But and collaboration requirements, and the if more functionality than this is needed – integration requirements. A record of the work; and for example, integrating various chemistry Once these requirements are defined, A long-term preservation mechanism. or biology-centric functions, or other then the task of finding a solution is more discipline-specific tools – then we are really straightforward. n
24 www.scientific-computing.com/BASL2018 Simplifying Your LABORATORY Journey to an INFORMATICS Integrated GUIDE 2018 Smart Laboratory
Want to learn more about laboratory informatics? Laboratory Automation Data Integrity Process Optimization Data and Systems Integration Access our Knowledg•e Management Laboratory Data Analysis Benchmarking
Informatics • Guide 2018 now e
FROM THE PUBLISHERS OF e-Science Solutions SCIENTIFIC COMPUTING Digital Laboratory Consulting WORLD [email protected] www.e-sciencesolutions.com www.scientific-computing.com/lig2018 Knowledge: Document Management Building a Smart Laboratory 2018
KNOWLEDGE Document management
This chapter considers how the cientists have been involved in the human, and cognitive issues associated with the smart laboratory contributes to development of artificial intelligence creation, capture, transfer, and use of knowledge the requirements of a knowledge (AI) for decades. The modern version in organisations. eco-system, and the practical of AI, which sought to create an Today the value of tapping into an easily consequences of joined-up science. Sartificial human brain, was launched in 1956 accessible collection of information, such as a Knowledge management describes but had been fermenting many years before smart laboratory’s assets, is appreciated more the processes that bring people and that, following the discovery that the brain was than in the past. The amount of data and information together to address the an electrical network of neurons that fired in information generated by instruments and acquisition, processing, storage, use, pulses. Over the years, there were apparent scientists increases exponentially, and staff and re-use of knowledge to develop breakthroughs that were followed by troughs turnover is rising to the point where someone understanding and to create value of despair caused by hardware and software with seven years of service with the same limitations. Among the highlights of this era employer is now considered an old-timer. was a Siri-like machine called ELIZA which, in Undocumented know-how and locations 1966, could be asked natural language questions of information resources are now issues. and provided voice-appropriate – albeit canned Reinventing the wheel is becoming more – answers. commonplace – not a desirable occurrence, The birth of the discipline of knowledge because the costs of drug development are ever- management (KM) was in the early 1990s. increasing and fewer blockbuster products are A tidal wave of KM consultants appeared, hitting the market. heralding the birth of this newer AI version Many software solutions offered to assist and the emergence of supporting hardware and information management are specialised and software. A few years later, several scholarly fragmented. Often, disparate divisions of an journals appeared as forums for advancing the organisation select their own local software understanding of the organisational, technical, solutions, and IT departments often dictate
26 www.scientific-computing.com/BASL2018 Building a Smart Laboratory 2018 Knowledge: Document Management requirements that restrict the scope of possible Retention schedules of custody if the record is moved. Copies can vendor solutions. Excluding small, single be made and distributed, but the ‘original’ is location labs, it is rare to see a smart laboratory Not everything can be kept for ever – but how always in the vault. It’s pretty much the same where all associated information resides under long is sensible? There is some consensus that with electronic records: the documents are one roof. information supporting a patent should be stored on a server where users can view them or There are general solutions to support retained for the life of the patent, plus several make copies. The official, ‘original’ record stays the processing of large data sets in a distributed years before and after to cover eventualities. in its slot. Chain of custody is maintained when computing environment. One of the best known Most pharmaceutical companies have settled the record is migrated to another location or is is Hadoop, sponsored by the Apache Software on a 40- to 65-year retention for intellectual converted into other formats. Foundation. Hadoop makes it possible to run property. Records to support regulatory applications on systems with thousands of compliance sometimes need to be retained for Long-term archiving: nodes, involving Petabytes of data. Its distributed as long as 25 years. At the end of their retention paper and microfilm records file system facilitates rapid data transfer among period, records should be evaluated for their nodes and allows the system to continue disposition. Should they be destroyed, or There is a general perception that records will operating, uninterrupted, in case of a node perhaps kept for a few more years? Scheduled be easy to find, retrieve, and view in the distant failure. The Hadoop framework is used by major examinations of records have a bonus of future. Paper and microfilm records that are players including Google, Yahoo and IBM, providing information that could be applied to stored in a clean, temperature- and humidity- largely for applications involving current issues. Looking through the supposed controlled environment could be readable for search engines and advertising. ‘rubbish’ can be a very good thing. more than 100 years. However, finding and There are at least two good reasons for retention retrieving them requires some strategic planning. Organisation is everything schedules. First, there is the smoking gun. In the At the very least, they should be organised by event of legal or regulatory investigations and/ year. Additional sub-categories or folders can be When organising information one needs to or audits, there’s bound to be information that is added to facilitate retrieval. The ideal solution decide what is important and what is not. erroneous, that conflicts with established facts, involves the assignment of a unique identifier to Traditionally, in the paper notebook era, experiments, results, and comments were systematically entered to show diligence in pursuing a potential patent on an invention. Nothing could be removed; only subsequently People like to use familiar visual signs to navigate. It’s natural and usually noted or re-explained. Supporting data from results in finding what you want plus additional associated materials instruments was retained with the notebook entries, and this practice led to the warehousing of innumerable papers as well as electronic records that might or might not be needed “ to support patent claims or meet regulatory requirements. or serves no particular purpose. Observations each record; the identifier containing or linking The volume of instrumental data today is and comments that arae taken out of context to relevant metadata to aid in searching. For much larger. Is it prudent to keep everything, can also be misleading. This is not a licence large collections, this information should be or perhaps classify the data into two piles – one to cook the books; the aim is to throw out the stored in a database. A plan must be developed that directly supports a conclusion and another junk and items that have no real contribution to migrate this information from its existing that is perhaps more generic? All electronic data to the organisation. It’s better to identify what hardware and software, after it becomes obsolete, suffers from aging, not unlike human aging. needs to be retained before any of these issues to newer systems. We’ll talk about media and file format aging a occur. The non-records should be destroyed little later, but we should also consider relevance as quickly as possible and the declared records Long-term archiving: aging. Should a particular spectral analysis file evaluated after a pre-prescribed time (retention electronic records be kept or should the sample be re-run five years period). Keeping non-records and records past from now using updated equipment? their retention dates costs money. The cost of We are all aware of the extremely short half- Information needs to be categorised into a hardware associated with information storage life of computer hardware and software. The small number of groups, preferably in a central continues to decrease but the amount of labour software authoring tools in use today will blink location to facilitate retrieval. needed to support large collections has increased out of existence and be replaced by Start with two piles and gradually split them sharply. tools that have more capabilities or are appropriately. It is sensible to imagine how a compatible with current operating systems. One researcher in the future would look for things, Authentication can only speculate regarding the hardware and having no knowledge of past notations and data storage media we will be using in the future. conventions. How can records that are created within an There will probably be no practical Rosetta Stone People like to use familiar visual signs to organisation be authenticated? They don’t all to help translate codes used in legacy software. navigate. It’s natural and usually results in need to be notarised, but it would be nice if there Maintaining authenticity and minimising data finding what is needed plus additional, was an easy way to come close to this. So here corruption needs to be addressed. associated materials. Search engines may give are the concepts to use. Appoint a designated There have been attempts to maintain a more precise results but may omit important records manager who will have full control of the museum of hardware and software that could things that are part of the navigation journey. records. People have been doing this for years help in viewing legacy records. These mostly Scientists appreciate the role of serendipity in with paper records – it works. The custodian failed, most notably an effort by the National drug discovery. authenticates the author and maintains a chain Aeronautics and Space Administration (NASA). www.scientific-computing.com/BASL2018 27 Knowledge: Document Management Building a Smart Laboratory 2018
NASA lost many of its electronic records from intended. The emerging standard for this least two good reasons for applying retention the early 1960s and then took steps to ensure purpose is PDF/A, an ISO-standardised version schedules. In the event of legal or regulatory that it would not happen again. of the portable document format (PDF). Finally, motivated investigations, and/or audits, there’s This resulted in the 2001 launch of the Open the immutable file is further protected from bound to be information that is erroneous, Archival Information System (OAIS) reference tampering by digital encryption. conflicts with established facts, or serves no model, sponsored by a global consortium of Electronic storage media is a moving target. particular purpose. space exploration agencies concerned with data It is quite unlikely that media being employed If there is a risk that observations and preservation. today will be used beyond the next 20 years. The comments can be taken out of context, Other global consortia have come together storage of electronic information on magnetic items that have no real contribution to the to develop preservation strategies. The tape, pioneered by IBM in the 1970s, is not only organisation’s business should be thrown out. International Research on Permanent Authentic the storage method of choice today, but its usage Records that are past their retention dates Records in Electronic Systems (InterPARES) is increasing. should also be discarded to avoid maintenance aims at ‘developing the knowledge essential Tape is far cheaper and more reliable than costs. to the long-term preservation of authentic any other medium used for archiving data. This The cost of hardware associated with records created and/or maintained in digital does not mean that records from a 20-year-old information storage continues to decrease, but form and providing the basis for standards, tape can be retrieved readily unless a compatible the amount of labour needed to support large policies, strategies and plans of action capable of drive, which could retrieve its content, has been collections has increased sharply. A records ensuring the longevity of such material and the saved in a museum. manager should be designated and given full ability of its users to trust its authenticity.’ To keep electronic records for more than control of the records. Finally, Australia’s Victorian Electronic 10 years, a migration strategy needs to be The basic guidelines are as follows: Records Strategy (VERS) provides a framework developed and implemented now, before the n Understand the legal implications of within which to capture and archive electronic museum closes. electronic records; records in a long-term format that is not n Establish a file plan; dependent on particular hardware or software. Chapter summary n Establish an electronic records preservation The concepts that these global data initiatives file plan; use for long-term preservation are the same. The best approach to organising information n Establish an electronic records manager or First, capture the content and metadata, then is to decide what is important to keep and management team; protect them with an immutable file format that what is not. How would a researcher in the n Establish and communicate policies; preserves the text, images, charts and tables and future look for things, having no knowledge of n Avoid point solutions; and renders them readable in the way the authors past notations and conventions? There are at n Don’t keep electronic records forever. n
28 www.scientific-computing.com/BASL2018 Subscribe for free* Do you compute? The only global publication for scientists and engineers using computing and software in their daily work
*Registration required
Do you subscribe? Register for free now! SCIENTIFIC COMPUTING scientific-computing.com/subscribe WORLD
Realising the potential of data emil Europa Market through accurate research, Intelligence Limited lead generation and client-led community creation
Scope fresh markets
Bespoke lead generation
Niche marketing
Make your data work harder
Research industry events
GDPR ready?
Europa Market Intelligence Ltd Tel: +44 (0) 1354 610188 [email protected]
www.europamarketintelligence.com Think smart. Think small. Beyond the laboratory Building a Smart Laboratory 2018
FURTHER AFIELD Beyond the laboratory
This chapter considers who cares here is a quote that states: ‘A couple delve deeper into the knowledge base to visualise about how smart the laboratory is, of days in the laboratory can save a and interpret relationships and correlations is and why? It also looks at the broader couple of hours in the library’. This growing. ‘Big data’ is the popular term being business requirements and their impact sentiment once typified the attitude assigned to the data problem, as it applies to on the laboratory, with an emphasis on Tof a lot of scientists, but those days are over. It all walks of life. This puts the emphasis on productivity and business efficiency, may not have reached the point where a couple ensuring that we have confidence in the integrity, integration with manufacturing and of hours on the computer can save a couple of authenticity, and reliability of the data going business systems, patent evidence days in the laboratory, but it’s heading that way. in, and that the appropriate tools are available creation, regulatory compliance, and The laboratory is part of a business process and, to search, analyse, visualise, and interpret the data integrity and authenticity as such, it is subject to the same productivity and information coming out. efficiency targets that apply to other parts of the These requirements are driven by the business. requirements of the laboratory’s customers Most laboratory informatics projects focus on for robust, reliable, and meaningful scientific the return on investment, typically quantified by information and data that is delivered in a timely streamlining data input through the elimination and cost-efficient way. The time-honoured of bottlenecks, by interfacing systems, and by principles of the scientific method provide the removing manual processes involving paper. basis for the integrity, authenticity, and reliability Most projects will also specify long-term gains of scientific data, but those principles need through establishing a knowledge repository, but to be reinforced in the context of regulatory this is where quantitation becomes difficult. It’s compliance and patent evidence creation. not unusual in the early days of an informatics deployment, for example, for a simple search to Productivity/business efficiencey uncover prior work that can save reinventing the wheel. But, as time goes on, it becomes the The basic objective in deploying laboratory norm to check before starting an experiment. informatics systems is to improve laboratory There is, however, another side to exploiting productivity and business efficiency. To the knowledge base – which we’re only just maximise the benefits, it is important to consider starting to come to terms with. The need to the wider laboratory and business processes
30 www.scientific-computing.com/BASL2018 Chapter summary
Building a Smart Laboratory 2018 Beyond the laboratory
that may be affected by the new system. It is therefore, far more difficult to measure as the the evidence – not on the medium that holds easy to fall into the trap of just ‘computerising’ value will be determined by behavioural changes. the evidence. One important factor is the data an existing laboratory function, rather than There is a growing body of evidence being integrity, which must be possible to prove in looking at the potential benefits of re-engineering presented at conferences on electronic laboratory court if necessary. Bound paper logbooks are a business process. The use of tools such notebooks (ELNs) by numerous companies that still being used to a large extent, as most legal as 6-Sigma or Lean can help considerably. have implemented them, showing that the short- advisors don’t feel comfortable with electronic Nevertheless, it is prudent to be careful with term time savings associated with the electronic data. It may be smart to talk to patent lawyers the use of these tools, depending on the nature solution are significant. These organisations also before starting to create electronic lab data. of the lab. For example, high-throughput, routine-testing laboratories, which basically follow standard operating procedures, are more receptive to process improvement. Discovery/ research laboratories however, which are less Building a good business case requires a thorough and systematic structured and are dependent on more diverse approach to understanding current limitations as well as future requirements and uncontrolled processes, are less likely to for the business benefit from formal process re-engineering. Productivity and business efficiency are usually measured in financial terms, although this may be translated into time-savings or, “ in some cases, the numbers of tests, samples, list a number of other non-quantifiable, long- How can we prove that the experiments completed. It is necessary, therefore, term benefits such as: IT system is good enough? to be able to quote ‘before and after’ figures for n Scientists spending more time in the any deployment project. Establishing a baseline laboratory; The answer is, of course, validation. Actually, metric is an important early step in the project. n It is easier to find information in a searchable validation of processes is nothing new; that has The tools can facilitate improvement through archive; been a part of the GMP and GLP regimes since well thought-out deployment, but also offer the n It is easier to share information; they were introduced. An IT system is a part of capability to monitor and improve processes. n Increased efficiency through the elimination the process and must therefore be validated as of paper; well. Costs/return on investment n A reduced need to repeat experiments The industry asked the US Food and Drug (knowingly or unknowingly); Administration (FDA) how it would handle Any organisation considering the n Improved data quality; electronic signatures, and accordingly 21 CFR implementation of a new informatics or n A smooth transition when people leave the Part 11 saw the light in 1997. The surprise was automation system will want to investigate the company; and that most of the two-page document was about return on investment (ROI), or cost/benefit. n Online use in meetings. electronic data, and only a little about signatures. This is usually extremely difficult, since many This, however, does make sense. How can of the projected benefits will be based on Regulatory compliance scientists use an E-signature if they are not a certain amount of speculation and faith. sure that the data is (and will be) valid? They However, there are some important points to The early research phases in the pharmaceutical can’t; they need to have control of your data consider in building the cost/benefit case. The industry comprises the testing of large numbers before they can sign it electronically. The EU costs associated with managing paper-based of chemicals to see if any of them have potential also came up with an equivalent to 21 CFR processes (e.g. notebooks, worksheets, etc.) as a new drug. Only the best will go on to more Part 11, namely the EU GMP Annex 11. This through their full lifecycle in the lab are not extensive testing. There has been a ‘consensus’ was revised in 2011 but does not improve on always fully visible or understood. that regulatory work does not start until the the first version. But a really good document Apart from the material costs, and the chemical has been chosen. Then adherence to covering electronic data and signatures is costs of the archive process, there is a hidden GMP[3] (good manufacturing practice) and yet another document numbered 11, the cost – and the time taken in writing by hand, GLP[4] (good laboratory practice) starts, and PIC/S PI 011.[7] This is a 50-page document cutting, pasting, transcribing, and generally the IT systems need to be in compliance with with the same requirements as Part 11, but it manipulating paper, as well as approval and the local requirements for IT systems. In the US, includes also a lot of explanations. PIC/S is the witnessing processes, all contribute to this this is 21 CFR Part 11[5] and in the EU it is GMP organisation for European pharma inspectors. hidden cost. It is normal in building the cost/ Annex 11.[6] While this may be at least partially They do stress that this document is not a benefit equation to look at how much of a correct, the fact is that the data, and of course the regulatory requirement, only an explanation scientist’s time is spent managing the paper- IT systems that hold the data, need to be under to the inspectors on how to handle IT systems. based processes, and to use this as a basis control for another business reason: patents. How that cannot be a requirement document, for potential time-savings with an electronic The US patent system is based on ‘First to is hard to understand, however. The main solution (see Figure 6). Although the start-up Invent’, and that means it must be possible to difference between Part 11 and Annex 11 is costs are high for an electronic solution, the prove the date of the invention. Traditionally, this that the latter also includes risks. IT validation incremental cost of adding new users and has been done using bound paper notebooks, shall be based on risks; high-risk systems need increasing storage space is modest. where the entries have been dated and signed, more validation than low-risk systems.[8] ROI tends to focus on the short term: how and co-signed by a witness. Paper notebooks This follows the same line of thought that soon can one get a return on the money invested can be admitted as evidence if they can be the FDA started in the early 2000s: know your in deploying a new system? But the true value demonstrated to be relevant. Electronic records processes, and base the work on the risks they of the system may be in the long term and, are equally relevant, as the judgment is made on encompass. www.scientific-computing.com/BASL2018 31 Beyond the laboratory Building a Smart Laboratory 2018
cannot be configured or that use default How do we validate our Assess the risks of E-records: configuration. Run-time parameters may be IT system? n The impact and likelihood/probability of entered and stored, but the software cannot problems being detected/happening: is it be configured to suit individual business The best answers are in a guidebook called high/medium/low? processes. GAMP5[9]. GAMP also has a few more Implement controls to manage risks: n Example: firmware-based applications, detailed sub-books. n Modify processes, modify the system design; COTS, instruments. What is written in this guide to a smart apply technical or procedural controls. n Validation: the package itself. Record version laboratory is of course just an overview. Please Monitor effectiveness of controls: and configurations, verify operations against see GAMP5 for more details. n Verify effectiveness, consider if unrecognised user requirements. Consider auditing the The GAMP5 way of validating the IT hazards are present; assess whether the vendor. Risk-based tests of application: test systems is as follows: estimated risk is different and/or the original macros, parameters, and data integrity. n Risk management to decide how important assessment is still valid. the system is in the process; Category 4: Configured products n Categories of software to decide what needs GAMP5 software categories n Definition: software, often very complex, to be done; that can be configured by the user to meet n Combination of risks and categories to the specific needs of the business process. decide what to do for this system; and Category 1: Infrastructure software Software code is not altered. n Testing guide for how to test the system. n Definition: layered software (i.e. upon which n Example: LIMS/SCADA/MES/MRP/EDMS/ applications are built). Software used to clinical trials, spreadsheets and many others Risk management manage the operating environment. (See GAMP5). n Example: operating systems, database n Validation: life cycle approach. Risk- engines, statistical packages, programming based approach to supplier assessment and Identify regulated E-records and E-signatures: languages. other testing. Record version and n Is the record required for regulatory n Validation: record version and service pack. configuration, verify operation against user purposes? Is it used electronically? Is a Verify correct installation. requirements. Make sure SOPs are in place signature required by GMP/GLP/GCP? for maintaining compliance and fitness for\ Assess the impact of E-records: Category 2: This category is no longer in use intended use, as well as for managing data. n The classification of potential impact on patient safety and/or product quality: is it Category 3: Non-configured products Category 5: Custom applications high/medium/low? n Definition: off-the-shelf solutions that either n Definition: software, custom-designed and
32 www.scientific-computing.com/BASL2018 Building a Smart Laboratory 2018 Beyond the laboratory
coded to suit the business processes. n Example: it varies, but includes all internally/ externally developed IT applications, custom There is a growing level of interest in how consumer technologies can firmware and spreadsheet macros. Parts of enhance the user experience of working with laboratory Category 4 systems may be in this category. informatics tools n Validation: the same as Category 4, plus more rigorous supplier assessment/audit, full life cycle documentation, design and source code review. “ brief test will show that the system is up and order to get a really smart laboratory with the running. information required. System validation But validation is never done. It’s important OQ – Operation qualification to prove that the system is still validated, even The validation itself usually needs to be This may be defined differently in different after changes in and around the system. Having divided into more manageable pieces. One organisations, but a common definition is to appropriate procedures to explain how to keep way is to use the ‘4Q method’. This comprises test each function separately. This is often what the validated state, and documentation to prove development qualification (DQ), installation the supplier already has done, so the user may that procedures have been followed, are a must. qualification (IQ), operation qualification not have to do this if documentation and/or a These procedures need to cover whatever is (OQ), and performance/process qualification supplier audit has shown that this has been done. appropriate, including: (PQ). What the chosen manageable pieces n Error handling, including corrective action or phases are is up to the individual, but it PQ – Performance or process qualification and preventive action; must be described in the validation plan. This The PQ is also defined differently in different n Change management; defines the phases, the input and output of the organisations. Basically, the PQ is testing that the n Validation/qualification of changes; phases, and which documents will be created implemented system is according to business n Backup and recovery; during the phases. Typically, this will be a processes. This includes indirectly testing that the n Configuration management; phase plan including test plans, the testing separate functions work as intended. This may n Disaster recovery and business continuity; itself and the test documentation, and the also be called the system testing. n E-signatures; phase report. If the supplier’s OQ testing is unavailable, n Environmental conditions; more of the functions may have to be included in n Risk assessment and management; DQ – development qualification testing. It is perfectly fine to combine the OQ and n Security and user access; This includes writing the user requirements PQ into one combined phase. n Service level agreements; specification, choosing the system, auditing the It is important to qualify or validate all the n System description; supplier if the risk assessment says that this is functions needed for your workflows. n Training; needed, and implementing the system. A thorough description of IT validation n Validation and qualification; can be found in the book, International IT n Supplier audit; IQ – Installation qualification Regulations and Compliance. This book also has n Daily use; Installing the system is usually just a matter of chapters on LIMS and instrument systems, and n Implementation of data in the system; following the description from the supplier. A the tips there are useful to read and follow in n Qualification/validation of implemented data in the system; and n Data transfer between systems. Some of these SOPs will be generic in the organisation, and some will be system specific. FIG 6 System costs of paper notebooks and ELNs Validation is a never-ending job, but with a validated system the user can be sure that the system works as intended and that the data is Paper secured inside the system. That means that the user can prove beyond doubt that the data was entered on a given date and that the system will show that data has been corrected later. Electronic Patent-related issues
The US patent system is based on ‘First to Invent’ and, in order to help determine who System cost was first to invent, most companies engaged in scientific research create and preserve evidence that they can use to defend their patents at a future date. Traditionally, this evidence has been in the form of the bound paper laboratory notebook. In a patent dispute, any Number of users inventor is assumed to have an interest in the outcome of the case, so their testimony must www.scientific-computing.com/BASL2018 33 Beyond the laboratory Building a Smart Laboratory 2018
to maintain the integrity of each system, and the consistency of the content between them. Similarly, the use of generic systems for such a task can increase discovery concerns and also increase the likelihood of problems. Further guidance should be sought from records management personnel and legal advisors within the organisation, in order to determine policy. A recommended approach to help uncover and resolve legal/patent concerns is to work with the company’s lawyers and patent attorneys to simulate the presentation of ELN evidence in the courtroom, and then work back to the creation of that evidence in the laboratory.
The America Invents Act – implications
Patent-reform legislation, in the form of the Leahy-Smith America Invents Act 2011, changed the US system from First to Invent to First to File in March 2013. It is very tempting be corroborated. Most organisations require that electronic records are admissible under to view this change as an opportunity to relax these notebooks to be signed by the author (‘I the Federal Rules of Evidence. The weight to be some of the procedural requirements of ELNs have directed and/or performed this work and given any particular record necessarily must be used in research laboratories. adopt it as my own’) and also by an impartial determined on a case-by-case basis.’ However, there are clauses in the Act that witness (‘I have read and understood this In terms of admissibility, paper and would suggest it’s wise not to make such an work’).[10, 11] electronic records are therefore equivalent. assumption. It is likely that patent interferences Evidence in US patent interferences is The judgment is made on the evidence, not and interfering patent actions will continue for subject to the Federal Rules of Evidence. There the medium in which it is presented. However, many years for patents and applications filed are a number of important hurdles that need it is important to understand the factors that after March 2013.[13] to be overcome, in particular the Hearsay Rule impact upon the authenticity of electronic There are specific circumstances described (by definition, if the author cannot be present, records and that in the adversarial nature of the in the America Invents Act that, for example, then the evidence is hearsay) and the Business courtroom, the opposing side may attempt to require proof of inventive activities to remove Records Exception. discredit the record, the record-keeping system, prior art for joint research activities, or to The Business Records Exception is an and the record-keeping process. The integrity preserve the right to an interference if the exception to the hearsay rule, which allows of the system and the process used to create and application contains, or contained at any time, a business records such as a laboratory notebook preserve records are therefore paramount. claim to an invention filed before March 2013. to be admitted as evidence if they can be Many organisations still require their Until the act becomes effective, and there is demonstrated to be relevant, reliable and scientists to keep bound laboratory notebooks. clarification about the implications of the new authentic. The following criteria must be met: This is because there isn’t the case law or other legislation, there is no reason to change in-house n Records must be kept in the ordinary course experience for most legal advisors to feel as procedures for keeping laboratory notebooks, of business (e.g. a laboratory notebook); comfortable with electronic records as they are or for vendors to revise the procedures and n The particular record at issue must be with paper. The issue is not one of admissibility, workflows in their ELN products. The more one that is regularly kept (e.g. a laboratory but of the weight that the record will have in immediate concerns are: notebook page); court. Unfortunately, we are unlikely to see a n There is a loophole that will allow people n The record must be made by or from by a suitable body of case law for many years. to prosecute a patent under the old First knowledgeable source (e.g. trained scientists); The high-stakes nature of the problem, to Invent rules for many years to come. n The record must be made lack of experience, and long-term accessibility First to File isn’t dead even after 16 March contemporaneously (e.g. at the time of the concerns have caused a number of organisations 2013 – there are some changes that mean experiment); and to adopt a hybrid solution, using an electronic proof of inventive activities will be especially n The record must be accompanied by lab notebook (ELN) front-end tool to create important for joint research activities. The testimony by a custodian (e.g. company records, and then preserving the resulting retention of other documentation related to records manager). records on paper. This gives the benefits of paper joint research projects may need to improve; Any doubt about the admissibility of electronic records (for the lawyers) while providing the and records was largely removed by this statement scientists with the benefit of new tools. A fully n Derivation proceedings will also require from the Official Gazette (10 March 1998 [12]: electronic system will require scientists to sign proof of inventorship. ‘Admissibility of electronic records in documents electronically, and the resulting To add further uncertainty, there’s always a interferences: Pursuant to 37 CFR 1.671, record to be preserved electronically. chance (or indeed probability) that things are electronic records are admissible as evidence Using multiple systems for patent evidence going to end up in the US Supreme Court to in interferences before the Board of Patent creation and preservation can expose an examine the constitutional implications of Appeals and Interferences to the same extent organisation to increased risk, due to the need a move away from First to Invent. So it does
34 www.scientific-computing.com/BASL2018 Building a Smart Laboratory 2018 Beyond the laboratory
appear that the new Act makes legally robust, records. An electronic signature is a generic n Enduring – they must not be recorded on the signed, and witnessed records of inventive term used to indicate ‘an electronic sound, back of envelopes, cigarette packets, or the activities (generally in the form of lab notebooks) symbol or process attached to or logically sleeves of a laboratory coat but in laboratory even more critical. With a move to ‘First to File’ associated with a record, and executed or note books and/or electronically by the there’s the additional pressure of getting to the adopted by a person with the intent to sign the chromatography data system and LIMS; and Patent Office quickly, which means it is necessary record.’ n Available – for review and audit or inspection to start paying attention to the patent filing A digital signature is a specific sub-set of an over the lifetime of the record. process, which has historically not been under electronic signature that uses a cryptographic It is important that laboratory staff understand much time pressure. technique to confirm the identity of the author, these criteria and apply them in their respective based on a username and password and the time analytical methods regardless of working on Data integrity, authenticity and management
Whenever electronic records are used There is a growing level of interest in how consumer technologies can within the framework of legal or regulatory enhance the user experience of working with laboratory compliance, data integrity and data authenticity informatics tools are fundamental requirements of the computer systems used to create, manipulate, store and transmit those records. These requirements may also apply to in-house intellectual property “ (IP) protection requirements. It will therefore be necessary for a laboratory informatics at which the record was signed. The requirements paper, hybrid systems or fully electronic systems. implementation project to very carefully for an informatics project will be somewhat To support the human work, we also need to consider the specific requirements of their dependent on the nature of the organisation’s provide automation in the form of integrated organisation in this business and internal requirements, but laboratory instrumentation with data handling area. [14] security, access control and electronic signatures systems and laboratory information management The characteristics of trustworthy electronic are factors that must be given appropriate systems (LIMS) as necessary. In any laboratory, records are: consideration. this integration needs to include effective audit n Reliability – the content must be trusted as There are a number of ways to ensure data trails to help maintain data integrity and monitor accurate; integrity and authenticity. The first is to develop changes to data. n Authenticity – records must be proven to be clear, written policies and procedures of what Supervisors and quality personnel need to what they purport to be, and were created is expected when work is carried out in any monitor these audit trails to assess the quality and transmitted by the person who purports laboratory; the integrity of the data generated of data being produced in a laboratory – if to have created and transmitted them; in the laboratory is paramount and must not be necessary a key performance indicator (KPI) or n Integrity – must be complete and unaltered, compromised. This is the ‘quality’ aspect of the measurable metric could be produced. physically and logically intact; and quality management system (QMS) that must be n Usability – must be easily located, retrieved followed. Chapter summary presented and interpreted. There is the parallel need to provide initial Data integrity, in a general sense, means that and ongoing training in this area. The training From a broader business perspective, the data cannot be created, changed, or deleted should start when somebody new joins the introduction of computerised tools for without authorisation. Put simply, data laboratory, and should continue as part of the managing laboratory information comes at a integrity is the assurance that data is consistent, individual’s ongoing training over the course of perceived higher cost, and challenges the user correct and accessible. Data integrity can be their career with the laboratory. to consider very carefully the consequences of compromised in a number of ways – human To help train staff, we need to know the moving from a paper-based existence to one error during data entry, errors that occur basics of laboratory data integrity. based on technology. when data is transmitted from one system to The main criteria are listed below: The return on investment equation is another, software bugs or viruses, hardware n Attributable – who acquired the data or critical in obtaining the initial go-ahead for an malfunctions, and natural disasters. performed an action, and when? informatics project, but the transition to There are many ways to minimise these n Legible – can you read the data and any digital from paper represents a major threats to data integrity including backing up laboratory notebook entries? upheaval to long-established and well- data regularly, controlling access to data via n Contemporaneous – was it documented at understood information management security mechanisms, designing user interfaces the time of the activity? processes. that prevent the input of invalid data, and using n Original – is it a written printout or Computer systems used in regulated error detection and correction software when observation or a certified copy thereof? environments need to be validated; the user transmitting data. n Accurate – no errors or editing without needs to be confident that computerised systems Data authenticity is the term used to documented amendments; can deliver productivity benefits, and data reinforce the integrity of electronic data n Complete – all data including any repeat or integrity and data authenticity can be guaranteed by authenticating authorship by means of reanalysis performed on the sample; in a digital world. electronic signatures and time stamping. n Consistent – do all elements of the Lawyers and patent attorneys need Generally speaking, electronic signatures chromatographic analysis, such as the to be confident electronic lab notebooks are considered admissible in evidence to ensure sequence of events, follow on and are they can be presented as evidence in patent the integrity and authenticity of electronic date- or time-stamped in expected sequence? submissions, interferences and litigation. n www.scientific-computing.com/BASL2018 35 Practical considerations Building a Smart Laboratory 2018
PRACTICAL CONSIDERATIONS Specifying and building the smart laboratory
This chapter looks at how to build a Functional/user requirements n Reporting requirements; and smart laboratory; what approaches to n Performance requirements. take; and how to deal with potential athering user or functional problems. Inevitably, becoming ‘smart’ requirements is one of the The criteria that define required performance takes time, not only due to the level of key tasks, usually assigned to may include: investment required, but also because the project team, to provide a n Access control and security; of the impact of change and the need specification against which potential solutions n Look and feel; G n to consider legacy requirements. can be evaluated. The task involves uncovering Robustness; The rate of change in computer and understanding user-needs, distinguishing n Scalability; technologies is far greater than in the them from ‘wants’ and ‘nice to haves’, and n Ease of use; laboratory and in business, and this aggregating the needs into a requirements n Technical performance/response times; and unavoidably means that the computing specification. In this context, reference to n Technical support. experience in the laboratory will lag ‘users’ includes not just end-users of the All of these requirements are normally collated behind the consumer experience. proposed system, but anyone who will interact into a request for proposal (RFP) that will be Additionally, the constraints of IP, with the system, or be involved with inputs submitted to potential vendors. The RFP should regulatory and legal compliance do not or outputs to the system. In order to do this, also provide more general information, including lend themselves to risk-taking when various methods may be used to gather needs an introductory description of the organisation deploying new technologies. New and to prioritise them. and the major objectives of the project, as well as laboratory informatics projects demand diagrams showing relevant workflows. The RFP a carefully managed and risk-averse The requirements may include, but are not may be preceded by a request for information approach limited to: (RFI) – a means of gathering information about n General business requirements; a potential vendor’s products and services, which n User/functional requirements; may be used to fine-tune a final list of vendors to n IT requirements; whom the RFP may be submitted. n Interface requirements; Unfortunately, users are notoriously bad n Regulatory issues; at stating what they need. Most systems are n Data management requirements; specified or designed by a team or committee n Error handling; and the team/committee members tend to be
36 www.scientific-computing.com/BASL2018 Chapter summary
Building a Smart Laboratory 2018 Practical considerations volunteers who are committed to the concept of Why do we need a new system? n Are there any specific policies an the system, enthused about the improvements it n What is the problem that needs to be solved? restraints relating to the introduction of can bring, and are able to envision the potential. n Is there any quantitative data that illustrates the ITsystems? Unfortunately, the committee process can create problem? Establish how the laboratory is currently complex systems and reflect compromises, and it n Which laboratory areas will be involved in the working, paying specific attention to the use is often the case that most problems come from project? and effectiveness of manual systems such as people who don’t volunteer for the committee! n Who makes the go/no-go decision? worksheets, paper lab notebooks, and data By definition, the members of the team are n What are the issues relating to IP (internal/legal/ management. Also identify major ‘electronic’ more committed to the success of the project patent)? and systems used for the acquisition, processing and than those who are not directly involved. In n Are there any regulatory compliance management of data, and ask what happens to their deliberations, project teams often develop requirements? this data – where is it stored and for how long? Is a concept of a solution that is much more Clarify why the organisation thinks it needs a it communicated or transferred elsewhere – if so, sophisticated than might be needed or, indeed, is new system. This is best achieved by developing how? Is it backed up and/or archived? Can it be economically justifiable. a problem statement that quantifies a specific found? Typically the requirement-gathering phase problem, or set of problems, about the laboratory’s Is laboratory data the responsibility of the involves harvesting needs, wants and ideas from productivity and/or knowledge management laboratory, or does IT have any involvement? the potential user community, and then engaging performance. What level of involvement does IT have in the in a prioritisation exercise to reduce the list to a The scope and scale of the problem (and purchase and implementation of laboratory specific set of requirements that form the basis of hence, the solution) should be identified. The key systems? a request for proposal (RPF) to be presented to vendors. It is important that the business requirements are fully clarified first of all; this ensures that the scope of the project is defined and can therefore Building a good business case requires a thorough and systematic help exclude some of the more exotic ‘needs’ that approach to understanding current limitations as well as future might arise. Any single item on the requirements requirements for the business list should justify itself not only financially, but also in terms of its usefulness and ease of use. Anecdotal experience suggests that some requirements specifications could be shrunk by “ between 25 and 50 per cent by the removal of decision-makers/budget-holders should also be Future laboratory processes and systems ‘wish list’ items – bringing cost-savings and lower identified, plus any other interested party who n Based on interviews with laboratory cost of ownership, as well as easier user adoption. may have influence over a go/no-go decision. It is managers and laboratory staff, a model It is important for the project team and sponsors important to know what business level constraints should be developed to illustrate the major to be able to define what business problem the may apply in terms of internal, legal or regulatory relationships between laboratory data and electronic laboratory notebook (ELN) will solve, compliance. information; and to ensure that user requirements are kept n Construct data workflow and laboratory simple and are focused on solving the problem. Laboratory/company background process diagrams; The formal RFI and RFP approaches can n Use organisation charts to clarify roles and n Identify any conflicts in nomenclature and only go so far, and it is essential that candidate responsibilities and organisational relationships; establish an agreed taxonomy; systems be demonstrated and assessed with n Identify the nature and scientific disciplines of some preliminary configuration to establish the laboratory work and how they relate to each n Identify the role (scope and scale) of existing and evaluate not only whether the system meets other; and laboratory systems in the model and the functional requirements, but also whether n Establish whether outsourced agencies (contract diagrams; and it provides an acceptable user experience. labs) are involved? n Test the model and diagrams against In some respects, it makes sense to consider Establish the way in which the laboratory is each of the laboratory areas and other functional requirements and user requirements organised, the nature of the work it undertakes and interested parties (IT, legal, QA, records as separate criteria. how it relates to internal and external organisations management). with whom it collaborates. A high-level plan, showing the relationships, Business case development and project processes and data flows that describe a future management Current laboratory processes and systems state for the laboratory, should be developed. n Which laboratory systems are already in use? This should include an identified role for each Building a good business case requires a (Are there SOPs?) of the laboratory systems and should clarify the thorough and systematic approach to n Which data acquisition systems are already specific functions of each. Any problems with understanding current limitations as well in use? laboratory terminology should be resolved. as future requirements for the business. It n Which teamwork/collaboration systems are The plan should be tested by presentation and is important to see laboratory informatics already in use? discussion with the interested parties. as a component in a laboratory ecosystem n Which document management systems are (technology, processes and people), rather already in use? Business plan development than ‘just another laboratory application’. The n Who is responsible for the management and n Quantify the benefits of the proposal, in following points should all be considered in support of these systems? particular productivity gains, ROI and formulating the case for a new informatics n Is there a (electronic) records management knowledge management, and support these system. policy? and estimates with case studies; www.scientific-computing.com/BASL2018 37 Practical considerations Building a Smart Laboratory 2018 n Undertake a risk assessment, paying attention User adoption is often considered one n Perceived usefulness (PU): ‘The degree to process, technology and people-related of the most critical success factors of an IT to which a person believes that using a risks. Align the risk assessment to the set of project, and paying appropriate attention to user particular system would enhance his or her user requirements; and requirements will enhance the likelihood of job performance’; and n Prepare, and include in the business case, success. Key to this is the recognition that people n Perceived ease-of-use (EOU): ‘The degree a high-level implementation plan that are more likely to comply with a request when: to which a person believes that using a addresses any specific requirements and/or n A reason is provided; particular system would be free from effort.’ risks that have been identified. n There is give and take; The technology acceptance model assumes that, Quantitative benefits should be identified, n They see others complying; when someone forms an intention to act, they along with all risks. An implementation plan n The request comes from someone they will be free to act without limitation. In the real should address known risks and/or potential respect or like; and world there will be many constraints such as problems, in particular the strategic approach n The request comes from a legitimate source of limited ability, time constraints, environmental to roll out, e.g. a progressive deployment, authority. or organisational limits, or unconscious habits the composition of the project team, change Concerns about user adoption can be reduced that will limit the freedom to act. management and user support. by carefully choosing the project team to Concentration on the positive aspects of ensure that these criteria are addressed, rather ‘usefulness’, both to the organisation and to Human factors than just announcing a new system and the the individual, and ‘ease of use’ will help users n What practical problems do laboratory training course schedule. Typically, putting a develop a positive attitude. It is in this area that workers experience with existing laboratory strong emphasis on user requirements and user the early adopters can have a powerful influence processes and data workflows? adoption by engaging users throughout the on their conservative and pragmatic peers. n How well will laboratory workers process tends to brand the implementation as a accommodate change? and ‘laboratory’ project, rather than an ‘IT’ project, Technology considerations n Are there any cultural, political or other and this can make it easier for scientists to accept internal relationships that could have an the proposed change. Multi-user informatics systems are typically impact on the project? The Technology Acceptance Model[17] (see based on two- or three-tiered structures in Potential problems associated with change Figure 8) is an information systems theory that which the application software and database management show be identified. This models how users come to accept and use a may share a server or be located on separate may be at an individual level or at an technology. The model suggests that, when users servers, and the client-side software deployed organisational level. are presented with a new software package, a on a local desktop, laptop or mobile device. number of factors influence their decision about Traditionally, the servers are based in-house, Internal culture and technology how and when they use it. The main ones are: but hosted services (cloud/SaaS) are generating adoption FIG 7 Crossing the Chasm The introduction of multi-user IT systems into organisations has a mixed track record. Multi-user systems are usually specified by a project team and often contain a number of Pragmatists Conservatives Looking for Believe in compromises and assumptions about the way an improvement tradition Visionaries people work. High-level business objectives Looking for a can therefore be put in jeopardy if users do not breakthrough successfully adopt the new system. However, Sceptics most case studies on electronic laboratory Not Technology Enthusiasts looking notebook implementations indicate a positive Looking for some user take-up. This may be attributed to a neat technology growing understanding of aspects of technology adoption, originally reported by Everett Rogers [15] in his book, The Diffusion of Innovations, The Chasm and developed further by Geoffrey Moore in Crossing the Chasm[16]. Moore’s ‘Chasm’ (see Innovators Early Early majority Late majority Laggards Figure 7) is the gap between the early adopters adopters and the mainstream market. The early adopters are a relatively easy market. Targeting them Innovators 2 to 3 per Technology enthusiasts: want to be first to try new initially is important, but the next phase of the cent technology; want one of everything. marketing strategy must target the conservative Early adopters 10 per cent Visionaries: able to align technology with strategic and pragmatic majority. The early adopters can opportunities; willing to take risks; horizontally oriented. play a central role in this. Since the electronic laboratory notebook (ELN) project team is Early majority 36 per cent Pragmatists: cautious with risk and money; loyal; vertically oriented. likely to be formed from the early adopters, they can play a pivotal role not only in specifying Late majority 36 per cent Conservatives: opposed to discontinuous innovation; believe and selecting a solution, but in articulating in tradition rather than progress. the rationale for the ELN, provide training Laggards 15 per cent Sceptics: negative attitude towards technology; identify and ongoing support to the conservative and discrepancies between what’s promised and what’s delivered. pragmatic majority.
38 www.scientific-computing.com/BASL2018 Building a Smart Laboratory 2018 Practical considerations
increasing interest, based on potential business FIG 8 Technology acceptance model benefits. From the user perspective, the client-side options fall into two categories: thick client and The degree to which a person believes that using thin client. The thick client is usually a substantial a particular system would enhance his or her job software installation on a local computer in which performance a good deal of the data processing is undertaken Individual user’s positive or negative feelings about before passing the output to the database server. performing the target behaviour This has the advantage of distributing the total Perceived usefulness processing load over a number of clients, rather than the server, and may also allow a certain External Attitude Behavioural amount of personalisation of the client software variables toward intention to support individual users’ needs. The downside is that system upgrades Perceived can become time-consuming and potentially ease-of-use A measure of the strength of one’s intention to troublesome, depending on the local perform a specific behaviour configuration – although centrally managed The degree to which a person believes that using a systems are now making thick client systems particular system would be free from effort easier to deploy, maintain and support. Thin clients typically access the application and database server(s) through a browser. No local processing power is used, so the server transition from paper systems to electronic is that data must be held locally on the device. and network performance are critical factors the loss of portability of, for example, a paper lab Wireless connectivity to a hosted system in providing good performance. The use of a notebook. (SaaS or cloud) has the benefit of direct browser can significantly reduce deployment The form factor of a laptop computer goes access to the system. and upgrade costs, but may restrict or limit user part way to resolving this concern, but the From a business perspective, the configurability. emergence of compact, lightweight tablets holds cloud offers an effective solution to the far more potential. Although tablets are often increasing demand for the implementation With regard to devices, successful deployments considered to be ‘data consumers’ – great for of collaboration tools across multiple have been made with: reviewing data but less effective for data entry – departments, multiple sites and different n Small form-factor PCs on the laboratory careful design of the user interface can optimise geographies – including outsourced operations bench; their potential for narrow, dedicated functions. where the practicalities of deployment are n ‘Remote desktop’; Typically, mobile devices offer significant largely limited to configuration rather than n Citrix; and potential for accessing data from remote physical installation of hardware and software. n A KVM switch operating between a desk- locations, or for capturing certain types of The benefits of a thin client – access from bound processor unit with keyboards and data in the field. The user experience can be anywhere, low start-up costs and centralised screens on the desk and in the laboratory. enhanced by the use of mobile devices that feature support – has both financial and functional There is a growing level of interest in how simple, gesture-based interactions for on-screen attractions. consumer technologies can enhance the navigation, consistent with typical consumer Pitted against this are concerns about access user experience of working with laboratory applications. Furthermore, the adoption of web control, security, and data integrity. informatics tools. With their focus on sharing, technologies creates the opportunity to design Some informatics vendors already offer this type of service. Cloud services generally fall into one of two categories: public clouds and private clouds. Public clouds utilise a single There is a growing level of interest in how consumer technologies can code base for the service to multiple clients. enhance the user experience of working with laboratory The single code base limits customisation informatics tools and integration, but helps keep costs down. A private cloud will typically offer a code base specific to an individual client, and will accommodate customisation and “ integration, but will normally come at a higher management cost. collaboration, interaction and ready access to a platform that supports all types of end-user information, consumer technologies exhibit devices, making critical laboratory data available Chapter summary considerable synergy with the high-level criteria anytime, anywhere, on any device in a global associated with current business requirements. wireless and mobile environment. The purchase and implementation of a laboratory Primarily, these focus on ‘mobile’ (portable The adoption of mobile devices for informatics system represents a major cost to devices), ‘cloud’ (access from anywhere), ‘Big informatics-based tasks raises a further the laboratory. It also represents the start of a Data’ (the need to be able to access and interpret question about how the host system is deployed relatively long-term relationship with the vendor. vast collections of data) and ‘social’ (collaborative and, in particular, how the mobile device Deploying a new system changes the working lives tools). communicates with the host. Synchronisation of laboratory workers and, as is the case with any The big attraction of mobile devices for end is one option, which has the advantage of not significant change, planning takes on a critical users is portability. A common complaint in the requiring remote connectivity, but it means role in the process. n www.scientific-computing.com/BASL2018 39 Knowlege: Data analytics Building a Smart Laboratory 2018
KNOWLEDGE Data analytics
This chapter takes the theme of analytics and data mining are often thought of with any other laboratory software, defining knowledge management beyond in the same context, often in connection with functional and user requirements are essential document handling into the analysis ‘Big Data’, they have different objectives. steps in making the right choice. Key areas to and mining of data. Technology by Data mining can broadly be defined as a focus on are that the tools have appropriate itself is not enough – laboratory staff ‘secondary data analysis’ process for knowledge access to laboratory, and other data sources; need to understand the output from discovery. It analyses data that may have that they provide the required statistical the data analysis tools – and so originally been collected for other reasons. tools; and that they offer presentation and data analytics must be considered This differentiates it from data analytics, visualisation capabilities that are consistent holistically, starting with the design of where the primary objective is based on either with broader company preferences and the experiment exploratory data analysis (EDA), in which standards. new features in the data are discovered, or Data analytics plays an important role ata analytics is the term applied confirmatory data analysis (CDA), in which in the generation of scientific knowledge to the process of analysing and existing hypotheses are proven true or false. and, as with other aspects of ‘knowledge visualising data, with the goal In recent years, some of the major management’, it is important to understand the of drawing conclusions and laboratory informatics vendors have started relationship between technology, processes, Dunderstanding from the data. Data analytics to offer data analysis and visualisation and people. In particular, staff need to have is becoming increasingly important as tools within their product portfolios. These the appropriate skills to interpret, rationalise, laboratories have to process and interpret the tools typically provide a range of statistical and articulate the output presented by the data ever-increasing volumes of data that their procedures to facilitate data analysis; and visual analysis tools. To take full advantage of data systems generate. output to help with interpretation. Alongside analytics, it should be considered as part of a In the laboratory, the primary purpose the integrated data analytics tools, more and holistic process that starts with the design of of data analytics is to verify or disprove more vendors offer generic tools to provide the experiment. existing scientific models to provide better software that can extract and process data from A quote attributed to Sir Ronald Fisher, understanding of the organisation’s current simple systems through to multiple platforms ca 1938, captures this point: ‘To call in the and future products or processes. and formats. The benefit of integrated data statistician after the experiment is done may Data mining is a related process that utilises analysis tools is that they will provide a be no more than asking him to perform a software to uncover patterns, trends, and seamless means of accessing data, eliminating post-mortem examination: He may be able to relationships within data sets. Although data concerns about incompatible data formats. As say what the experiment died of.’ n
40 www.scientific-computing.com/BASL2018 Chapter summary
Building a Smart Laboratory 2018 Summary
Summary
In this guide we have attempted to With a growing emphasis on right-first- and learning (data analytics) and experimental coalesce much of the information time, error-reduction, and productivity, it is design. Further afield, they will contribute to required in order to design and a management challenge to take the time to predictive science. implement as smart laboratory or, at review and assess successes and failures. Nevertheless there needs to be some the very least, to begin the process of The role of the informatics tools within a space for ‘right brain’ thinking, alongside laboratory automation. While it may smart laboratory, or ‘knowledge ecosystem’ those systematic and structured approaches seem like a challenging prospect, the (Figure 9) is important. They are strong in for increasing efficiency and productivity. underlying principles are simple and terms of capturing, recording, and organising Innovation depends on knowledge and focused on crafting a strategy that will data, and increasingly, they provide facilities understanding. Although technology can enable more productivity and insight to for sharing information. But further assemble and look after the data, making be generated from scientific research opportunities arise with regard to evaluation sense of it is down to human assessment
FIG 9 Knowledge processes he five nodes in Figure 9 represent a generalisation of the major knowledge processes, and it is quite Sense/explore/discover Mostly done as an individual activity –people evident that technology, in the looking for, or encountering new items of knowledge. Tform of the laboratory informatics tools, has Evaluate/learn an enabling role in a laboratory knowledge Intergrate, organise and test with other or pre-existing knowledge ecosystem. (either tacit or explicit) to generate new facts, ideas, knowledge, Capture/record From laboratory informatics to knowledge concepts. Save the ‘knowledge’ somewhere; e.g. lab notebook, management, technology is predicated technical report, etc. on logical and systematic processes. But serendipity has always had a significant role in science. Many scientific advances have originated from ‘what if’ moments, chance Share Organise Communicate with others Codify, index, make readily observations, and things that went wrong. (could be explicit or tacit). This accessible. Could be forces structure (organisation) accomplished or assisted by Failure often has more to teach than success! and further integration. others, or by automation.
www.scientific-computing.com/BASL2018 41 Chapter summary
Summary Building a Smart Laboratory 2018
and understanding. In the main, so-called be to adopt a one-size-fits-all approach. This and to user acceptance, and is a responsibility knowledge management ‘solutions’ are no could generate disaffection amongst users. that falls to management in its objective of more than data or information management The current informatics market is moving building a smart laboratory. It requires a solutions. towards more modular solutions, which have sympathetic view of the requirements of the It’s only when the human component a generic core, and optional discipline-specific different disciplines, and the way in which is added that knowledge management can modules. This creates a better opportunity these functions are managed and provided for, flourish, and even then, it needs the right to find a single-source solution. The shared even when organisational demands push for environment – hence the concept of a functions can be separated from the scientific increased uniformity and consistency. ‘laboratory ecosystem’, or smart laboratory. functions which are closer to the heart and The concept of a smart laboratory will Although management may want to see such soul of the scientist’s laboratory work. Shared vary from organisation to organisation an ecosystem, it can buy only the tools, it functions would include such issues as depending on the nature of its business, and needs to create the right environment if the document authoring, approval/witnessing, the technological choices it makes. Discovery knowledge ecosystem is to be nurtured and file and document management, and legal and development are increasingly recognised cultivated. and regulatory compliance – all of which fall as two steps in a holistic product life-cycle The ecosystem is dependent on an open into the ‘bureaucratic’ category and which process rather than stand-alone functions. and collaborative culture and supportive lend themselves to process improvement The focus of this guide has been on leadership; not secrecy, discipline or rigid opportunities more readily than the technology, with due consideration to the management. Participants need to opt in; not scientific aspects. It may still be a one-size- laboratory processes to which it can be be forced in. One worry is that the digital fits-all approach, but it can be designed to applied. It has also touched on some aspects revolution may be driving a lot of thinking to accommodate the requirements of multi- of culture and technology adoption, but it be ‘digital’, with the risk that random, analogue disciplinary laboratories, and to standardise must be remembered that user acceptance is a mindsets and gut feelings may be seen as and improve common sub-processes, rather critical success factor in almost every system irrelevant and inconsistent with modern than making compromises. or project. concepts of science. Ideally, laboratory informatics tools should Technology on its own cannot overcome This way of looking at things may shed not be perceived as an intrusive bureaucratic challenges in the laboratory. The take-home some light on why the early ELN market process, but rather as something that message is that to become ‘smart’ the lab was sub-divided into different solutions for facilitates the scientific method and doesn’t users’ and managers needs to understand its chemistry, biology and QA. intrude on the social and intellectual processes role in the organisation’s end-to-end business The risk for a multi-disciplinary laboratory that are essential to the science. Achieving processes and optimise its technologies to fulfil that is looking to implement an ELN would this objective is essential to joined-up science those requirements. n
References
1. The Gartner Hype Cycles: www.gartner.com Systems, February 2008, International Society for Segalstad, S. H., (2008) International IT Regulations and 2. CENSA: The Collaborative Electronic Notebook Systems Pharmaceutical Engineering (ISPE), Fifth Edition, ISBN Compliance: Quality Standards in the Pharmaceutical and Regulated Association 1-931879-61-3: www.ispe.org Industries, Wiley-Blackwell 3. Good Manufacturing Practice (GMP): The US has one set in the 9. Good Automated Manufacturing Practice Guidelines version 5, McDowall, R. D., (1987) Laboratory Information Management Federal Register 21 CFR, and the EU has its own, as do other International Society for Pharmaceutical Engineering, Tampa Systems, Sigma Press geographical areas and organisations like the Organisation for FL, 2008 Laboratory Notebook Guidelines: BookFactory, LLC, 2302 Economic Co-operation and Development (OECD) McDowall, R.D., (2009) Spectroscopy Focus on Quality, p23 S. Edwin C. Moses Blvd, Dayton, OH 45408. Available at US FDA, 21 CFR Part 211 Current Good Manufacturing 10. Using Electronic Records in Patent Proceedings, article by www.bookfactory.com Practice for Finished Pharmaceuticals, 2005, FDA: www.fda.gov Damien McCotter and Peter Wilcox. Originally published in Mahaffey, R. R., (1990) LIMS: Applied Information Technology for European Union Volume 4: Good Manufacturing Practices – Managing Intellectual Property’s World IP Contacts Handbook, the Laboratory, Nakagaw Medicinal products for human and veterinary use, 1998, 153 14th edition, 2007. Available at www.mondaq.com Sellen, A. J., and Harper, R. H. R., (2003) The Myth of the Paperless pages, incl. Annex 11 covering computerised systems 11. IP Expert Advice: Tips on creating a lab notebook that contains Office, The MIT Press 4. Good Laboratory Practice (GLP): The US has one set in the ‘convincing evidence’: www.edn.com/article/CA6445886. Franklin, C., (2003) Why Innovation Fails, Spiro Press Federal Register 21 CFR; EU has its own, and also other html?industryid=47048 Kanare, H. M., (1985) Writing the Laboratory Notebook, An 12. Admissibility of Electronic Records in Interferences, Bruce H. geographical areas American Chemical Society Publication Stoner Jr., Chief Administrative Patent Judge, www.uspto.gov/ US FDA, 21 CFR Part 58 Good Laboratory Practice for Non- web/offices/com/sol/og/con/files/cons119.htm Clinical Laboratory Studies, 2005, FDA: www.fda.gov 13. Private communication: Colin Sandercock (Perkins Coie LLP) European Union, Council Directive of 7 June 1988 on the September 2011 eOrganizedWorld: www.eorganizedworld.com inspection and verification of Good Laboratory Practice (GLP) 14. The ABCs of Electronic Signatures, Nettleton, D., Lab Manager Free online LIMS training courses: www.LIMSuniversity.com (88/320/EEC) Magazine, 9 September 2010: www.labmanager.com/?articles. The Generally Accepted Recordkeeping Principles: European Union, Council Directive – of 24 November 1986 - view/articleNo/3800/title/The-ABCs-of-Electronic-Signatures www.arma.org/garp/index.cfm 86/609/EEC – on the approximation of laws, regulations and 15. Rogers, E. M., Diffusion of Innovations, The Free Press. New York Independent, non-commercial LIMS user’s group: administrative provisions of the Member States regarding the 16. Moore, G. A., Crossing The Chasm, Capstone Publishing www.LIMSforum.com protection of animals used for experimental and other scientific 17. Bagozzi, R. P., Davis, F. D., and Warshaw, P. R., (1992). Industrial Lab Automation: www.industriallabautomation.com purposes Development and test of a theory of technological learning and 5. US FDA, 21 CFR Part 11 Electronic records; electronic usage. Human Relations, 45(7), 660-686 The Institute for Laboratory Automation: www.institutelabauto.org signatures, 1997, FDA: www.fda.gov 18. Multi-Ontology Sense Making, David Snowden, http:// The Integrated Lab: www.theintegratedlab.com OECD Series on principles of good laboratory practice and cognitive-edge.com/uploads/articles/40_Multi-ontology_sense_ Journal of Information & Knowledge Management (JIKM): compliance monitoring number 10 makingv2_May05.pdf www.worldscientific.com/worldscinet/jikm 6. European Union Volume 4: Good Manufacturing Practices – LIMSfinder: www.LIMSfinder.com Medicinal products for human and veterinary use, 1998, 153 NL42 Consulting: www.NL42.com pages, incl. Annex 11 covering computerised systems Further reading and websites 7. PIC/S, PI 011-03 Good practices for computerised systems in Online encyclopaedia for laboratory, scientific and health informatics: www.LIMSwiki.org regulated ‘GxP’ environments. 25 September 2007, PIC/S: Stafford, J. E. H., (1995) Advanced LIMS Technology: Case studies www.picsscheme.org and business opportunities, Springer PDF/A standard: http://en.wikipedia.org/wiki/PDF/A 8. GAMP 5 (Good Automated Manufacturing Practice) Guide: Christensen, C. M., (1997) The Innovator’s Dilemma: When New Scientific Computing World: www.scientific-computing.com A Risk-Based Approach to Compliant GxP Computerized Technologies Cause Great Firms to Fail, Harvard Business School Segalstad Consulting: www.limsconsultant.com
42 www.scientific-computing.com/BASL2018 Smart Labs need a PLATFORM SMART INFORMATICS PLATFORM MODULES Join the world’s forward-thinking R&D organizations who use IDBS’ E-WorkBook Cloud to securely capture, manage, share and exploit their data. ELN
ADVANCE
CHEMISTRY
BIOLOGY
INVENTORY
CONNECT
REQUEST THE E-WORKBOOK CLOUD: best-in-class, enterprise scientific knowledge management platform. It’s time to smarten up your lab. INTEGRATIONS
WWW.IDBS.COM
IDBS Scientific Comp World BASL Ad 282x213 V3.indd 1 23/02/2018 14:28 Scale up your expectations, Chromeleon CDS does more.
From workstation to global deployment, Thermo Scientific™ Chromeleon™ CDS scales to any business – from research through development to QC – delivering proven productivity gains and easy tracking and defense of your data. One cloud-ready CDS that does it all – all backed by a product team dedicated to keeping you one step ahead of the next regulatory or market change.
Learn more at thermofisher.com/Chromeleon
© 2018 Thermo Fisher Scientific Inc. All rights reserved. All trademarks are the property of Thermo Fisher Scientific and its subsidiaries unless otherwise specified. AD72657-EN 0218S