To contents
SFI Digital Food Quality Annual Report 2020
To contents
Her er en tittel
Aut latiusd andaeri quo eostistrum facero cum exernam dolorem explit erum nonsequas aliberchil mo iur, ut quas voles molorumquia sam et dit et hariam facienia cus consent. Nequi dusandis aciassimenes magnimi llautem sum aliberferia cuscimp eribus eos sintur sandi imus namendit faccus dendandit, cus duci nullupt ibeatur, cust qui officto cuptio modion sequis dolup etrav ethwq tiatur?
Ficiis a poris sitam doloriatur? Ique sit militaq uiatur, eumquod icimetur? Uptaquam etur? Fictum ut maximinit quo inumetu repelia natur, nihil imaiore volore ipictat emporup tatisim doluptatur, sam doloribus auta serspictaque cusandit eatur? Quide licabo.
Beatem et eos apedita tiorem enitat is aribus incte dolecatem hic tem ilit voluptat ut explanda sus enit alit, accus sit, tende se natae venimolupta cus, qui berspelene nusam exces ex explignatur, offic test, optur andis ape re por alit voluptas ulpa cus del estorro vitium quatem a doluptur rem ium eum aut eum accatemqui. Navn
Multiple authors. (2021). Annual Report 2020 SFI Digital Food Quality – DigiFoods.
Ås: Nofima, eds. Katinka Dankel, Senior Engineer in Nofima, makes in-line measurements on residual raw material. Design: Raquel Maia Marques @ Vitenparken • Photo/cc: Jens Petter Wold/Nofima
DigiFoods Annual Report 2020 2 Contents
1. Overall progress and summary for 2020 4 4. Scientific activities and results 28 Vision and Objectives 7 Pillar 1: Novel sensor systems and application development 28 Milk production is the backbone of Norwegian agriculture 8 Pillar 2: Robot and sensor integration 30 Pillar 3: Integrated in-line sensing solutions 32 2. Research Plan and Strategy 10 Pillar 4: Utilization of large-scale quality assessments 34 How robots find the best food 12 5. International collaboration 37 3. Organization 14 Partners 18 6. Recruitment, education and training 39 R&D Partners 18 Getting down to the nitty-gritty of food quality 40 User Partners 20 “There is no doubt that DigiFoods will play an important 7. Communication and dissemination 42 role in our innovation work in the years to come” 26 The devil is in the details 44
Contents DigiFoods Annual Report 2020 3 To contents
1. Overall progress and summary for 2020
The global food production is facing enormous challenges in terms of sustainability and food security for a growing population.
Climate reports from IPCC1 estimates that 21–37% • Photo/cc: Anders Hansen, Sintef of the total man-made greenhouse gas (GHG) emission is related to the food sector, while 25–30% of all produced food is lost or wasted. As a result, the Norwegian food industry has made an agreement to reduce food loss by 50% by 2030. To reach this goal, one major challenge is to effectively handle and utilize all available food resources of different origin and quality. The inherent variation in raw material quality reduces process efficiency and increases waste and variation in the final product. The key to reducing waste and increasing profitability is, therefore, to control and utilize the raw material variation in the best possible way.
SFI Digital Food Quality (DigiFoods) is a Center for Research-based Innovation (SFI), funded by the Research Council of Norway and the partners. DigiFoods will develop smart sensors for effective food quality assessment directly in the processing lines and in the field. Massive assessment of essential food qualities throughout the value chains will pave the way for a digital transformation of the food production.
1 Intergovermental Panel on Climate Changes, report 160819, 53-121 (2019). DigiFoods kick-off had to be digital, due to the pandemic. Nevertheless, the enthusiasm among the participants was great.
1. Overall progress and summary for 2020 DigiFoods Annual Report 2020 4 To contents
The obtained quality information will be used for We have followed the centre’s innovation model, The DigiFoods innovation model is based on each research optimization of processes and value chains making meaning that in each research project we have sought task being assigned active partners from all three groups: food the food industry more efficient and sustainable. In to assign active partners from the three groups: food companies, technology providers and R&D institutions. Together they will i) consider the needs and business cases, ii) develop DigiFoods, six research partners, eight food companies companies, technology providers and R&D institutions. and evaluate technology and iii) implement and commercialise. and thirteen technology providers are eager and Together these will consider the needs and business motivated to contribute to actual innovations within cases, develop and evaluate technology, and hopefully food processing, sensor technology and data analytics. implement and commercialise. The projects cover all the objectives of the centre and span several This annual report covers activities during the first essential research themes: Sensor development, three months of the centre, October 1st to December robotic operation of sensors, strategies for sensor 31st 2020. implementation in industry, and process optimisation. Food companies In December, the Board approved the DigiFoods In this report, we provide a brief presentation of each Working Plan for 2021, consisting of 10 research of these research projects. In 2020, research activities projects, which will initiate the process of realizing were already conducted in the project RAMAN, which the research that was proposed in the SFI application. is a successful collaboration between technology The plan marks the start of a long-term and ambitious providers. It is our opinion that this first working plan Research Project collaboration, and addresses interesting and relevant will ensure a solid and exciting start for DigiFoods Technology research with great innovation potential. and contribute significantly to reach the main goals providers of the centre. During October, the leader team had individual meetings with all partners, where suggested activities were presented and discussed. These meetings were indeed inspiring. It was important for us to find a good R&D balance between long term research and low hanging institutions fruits that also provide important and motivating results in the short term.
1. Overall progress and summary for 2020 DigiFoods Annual Report 2020 5 To contents
A vital part of the centre is the engagement and education of PhD-students and post docs. We have defined suitable tasks for the seven scholarship holders who will work in DigiFoods from 2021, and six of them have already been recruited. We are looking forward to including them in the research projects and, thereby, securing the foundation for successful years ahead in the centre.
The autumn of 2020 was affected by the Corona pandemic. We looked forward to kicking off the centre early in December by gathering all of the partners in Norway. Since this became impossible, we had a good digital meeting. A very important purpose for DigiFoods is, however, to be an innovation hub and a network where the partners can meet, exchange and discuss ideas. For this to happen we need to meet physically. We are looking forward to this in 2021.
Jens Petter Wold Centre Director, DigiFoods
Jens Petter Wold, the Centre Director of SFI Digital Food Quality, shortnamed DigiFoods, is a senior scientist at Nofima. • Photo/cc: Jon-Are Berg-Jacobsen, Nofima
1. Overall progress and summary for 2020 DigiFoods Annual Report 2020 6 To contents
Vision and Objectives The goal of SFI Digital Food Quality is to develop smart, sensor-driven solutions that deliver the essential food quality information required for successful process optimisation and digitalization of the food industry.
Food processes are extremely complex and challenging 6. Foster innovations, patents and spin out companies The DigiFoods vision: Extensive food quality assessment enables to measure due to the inherent high level of biological by the project partners from food industry, new insights and radical changes from farm to fork. variation in raw materials. The development of advanced technology and research solutions that are built on a fundamental understanding 7. Disseminate knowledge to the industrial sector, of food science, will allow the food industry to effectively the research community, and to the general public measure and handle these variations, enabling a ground-breaking digital transformation of the industry. DigiFoods will radically change food production Machine learning by enabling optimization, control and differentiation Big data The Primary objective of DigiFoods is to develop digital based on measurements of food quality. The results solutions for food quality assessment as cutting-edge will lead to a more efficient and sustainable food Information connection technological basis for optimal food value chains. industry, internationally competitive Norwegian technology companies, and enhanced knowledge
Besides this there are seven Secondary objectives: transfer and researcher training. Decision Process Product 1. Develop novel in-line sensor systems and support control differentiation applications for measuring critical food The DigiFoods objectives range from fundamental quality parameters technology knowledge to practical industry and 2. Develop automation and robotic solutions for market implementations, which are equally important enhanced sensor operations in process and in field for achieving successful innovations. We aspire 3. Develop solutions and strategies for successful to bridge the gap between research and industry sensor implementation in the food production by building a strong, business-oriented research 4. Develop data-driven strategies for process, network of innovation-oriented companies, and product and value chain optimisation based national and international R&D institutions. These on extensive food quality measurements expected impacts are in line with the centre goals 5. Build and transfer competence in industry and the overall objectives for the SFI scheme. and academia and educate master students, 9 PhDs and 3 post docs
1. Overall progress and summary for 2020 DigiFoods Annual Report 2020 7 Article To contents Milk production is the backbone of Norwegian agriculture by Wenche Aale Hægermark, Nofima
“Happy, healthy cows, grazing along the fjords and in the mountains, along the coast and inland are some of the most delightful things about Norway.” These are the words of Anne Cathrine Whist, who is responsible for Product Development and Research at TINE Meierier and is also the Chairperson of DigiFoods.
It is almost 20 years since Anne Cows – and the wellbeing are milked, how long they spend Cathrine qualified as a veterinarian. of farmers chewing and how long they spend She realised early on that she loved The important role played by dairy lying down in their stalls. large animals. Especially cows – and production in Norwegian agricul- helping dairy farmers. Travelling ture was a key factor when Anne “The sensors provide farmers with around Norway’s regional areas Cathrine chose her field of research. an idea about whether or not the and experiencing how Norwegian In her PhD she examined the rela- feed is optimal and how the cows cows constitute the backbone of tionship between udder health and are doing. New technologies and «Without cows and dairy Norwegian agriculture has done milk quality. She has been leading sensors, data collection and sys- something to her. TINE’s research on sustainable milk tematisation will provide even production our regional production for several years, and she more opportunities. DigiFoods has “Without cows and dairy produc- is proud of what has been achieved great expectations in this respect,” areas would have seen little tion our regional areas would have by TINE together with the other cat- says Anne Cathrine. The next step seen little settlement and our cul- tle industry partners (GENO, Anima- is to ensure direct communication settlement and our cultural tural landscape would have been lia and Nortura) in Norway: strong, between the data produced by the a thing of the past. Like much of healthy cows that require minimum farms and data at the dairy. landscape would have Sweden, Norway had been regen- usage of antibiotics. erated. Milk production still shapes Sustainability has three been a thing of the past.» our landscape and our country.” Many Norwegian dairy farmers have dimensions – Environment. She points out that this applies to already entered the digital age. Sen- Society. Economics. the whole settlement pattern, which sors have been installed in many Dual purpose Norwegian cattle pro- includes small local hamlet schools, robotically operated cowsheds. vide both milk and meat and have a sports clubs and local grocery stores These measure conditions such as lower CO2 footprint than pure meat – because people actually need how cows move in the cowshed, or dairy breeds. something to live off. how often they eat, how often they
1. Overall progress and summary for 2020 DigiFoods Annual Report 2020 8 Article To contents • Photo/cc: TINE “We are interested in all three The composition of the milk is affect- “Although we haven’t completely dimensions of sustainability, and ed by what the cows eat. For exam- thought out the next step, this could these are represented in TINE’s ple, the fatty acid composition is involve connecting data from sen- sustainability strategy. It is import- completely different if the cows only sors on feed wagons and farm tanks ant that both the animals and the eat grass. This affects both the taste with data from incoming transport, farmers thrive. That people living in and consistency of TINE’s products. delivery checks and installations. Norway’s rural areas have some- This will provide us with better trace- thing to live off. That operations are Technology will bring the ability and control – and the oppor- both viable and environmentally dairy and farmers closer tunity to make even better use of all friendly. The greatest greenhouse together the raw materials”, concludes Anne gas emissions occur in primary TINE’s state-of-the-art, high-tech Cathrine. industry. We focus on circular rea- dairy at Jæren produces white soning and obtaining an overview cheese, prim (cream cheese), butter, of those factors that have the great- margarine and milk powder. “There est impact on methane emissions,” we will be testing and using the sen- says Anne-Cathrine. sors and other technology that will be developed at DigiFoods,” says Grass or concentrate feed? Anne Cathrine. “Sustainability is also about eco- nomics. It is important that we find The cheese production lines will be an optimal solution for cows, the equipped with in-line sensors that environment and the financial sit- will measure temperature and pH uation of dairy producers when dis- values, etc. Data from the sensors cussing feed and feed composition. will be used for monitoring whether Proper feeding is important for the or not cheese processing is proceed- fertility, health and welfare of the ing correctly or if there are any devi- cows and not least the financial sit- ation that would require the process uation of farmers,” she says. to be stopped and new samples to be taken before the process can be TINE has great ambitions in respect recommenced. This will provide TINE of growing feed for dairy cattle in with assurances that the cheese will Norway and is working constantly be just as it should be, with a perfect to ensure that the cows will eat as taste and consistency – always. much top quality roughage as possi- ble which has been harvested at the Anne Cathrine Whist is the Chairperson of DigiFoods. She has been leading TINE’s research on sustainable milk most optimal time. production for several years, and she is proud of what has been achieved by TINE together with the other cattle industry partners in Norway; strong, healthy cows that require minimum usage of antibiotics.
1. Overall progress and summary for 2020 DigiFoods Annual Report 2020 9 To contents
2. Research Plan and Strategy
The main research hypothesis of DigiFoods is that in-line food quality measurements can be used to understand, optimize and radically change food value chains.
The innovations in DigiFoods will be accomplished activities will straddle two pillars or more and others by combining basic and applied research. A major might over time progress from one pillar into another. difference from traditional research in this area lies in the scientific method; prototypes will be tested Pillar 1 will develop novel sensor systems that address at the end-users at an early stage, as part of the critical in-line challenges and industrial needs. Pillar 2 technology development. This includes large-scale will design novel integrations of robotics and sensors. trials in fields, onboard fishing boats and in industrial Pillar 3 will develop strategies for successful food process lines, and secures relevance and industry implementation of in-line sensors in processes. involvement from year one. The research activities In Pillar 4, the in-line food quality measurements are organized in four pillars, and involves value chains will be placed in a broader perspective and for fish, meat, vegetables, dairy and bio-processing combined with other relevant data sources to realize Partner companies representing the major food value chains will (Figure 3). These pillars are not at all silos; some improvements at farm, industry and value chain level. define relevant research activities for the four research pillars.
Pillar 1 Pillar 3 Pillar 4 Fish Novel sensor systems and Integrated in-line sensing solution Utilication of large-scale quality Meat application development assessments
Vegetables Pillar 2 Dairy
Bio-processing Robot and sensor integration
Novel sensors and robotics Novel strategies for in-line success Novel solutions for process and designed for in-line use value chain optimization
2. Research Plan and Strategy DigiFoods Annual Report 2020 10 To contents
Most of the experimental work in Pillar 3 and 4 will and envisioned innovations, targeting gaps take place in the food industry, in the field or onboard in knowledge and technology. All partners have been fishing boats. These will serve as important research involved in the planning of the projects, ensuring facilities for securing relevance and usefulness of the relevance and securing in-kind contributions through technology, and for collecting extensive amounts active involvement in the upcoming work. of food quality data. Projects in Pillars 1 and 2 will collaborate to develop All activities will as far as possible include participants prototype solutions and these will be evaluated for from all three partner groups (food companies, industrial use in Pillar 3, together with already existing technology providers and R&D institutions) to ensure sensors. Results from Pillar 3 will also be fed back to practical relevance, interdisciplinary and relevant Pillar 1 and 2 to optimise and improve the solutions competence. This project organisation is the core based on in-line performance. Well working solutions of the centre’s innovation model, meaning that the developed in Pillar 3 will provide Pillar 4 with essential partner groups together will consider business cases quality data on an industrial scale. and innovation opportunities associated with the research. A Scientific Advisory Board (SAB) has been appointed for DigiFoods, consisting of researchers with A thorough process led by the leader team has defined competencies in the fields of research in the centre. ten research projects, which will initiate the process An important task of the SAB in the years ahead will SFI projects allocated in research pillars according to the figure of realizing the research that was proposed in the SFI be to advice during development of the annual plans. on the previous page. For all projects except “Opportunities” application. The projects address the outlined goals and “Robust”, PhD or postdoc students will be affiliated
Pillar 1 Pillar 3 Pillar 4 Fish OPPORTUNITIES Meat RAMAN HYPERSPEC
Vegetables ROBUST Pillar 2 Dairy
Bio-processing
Novel sensors and robotics Novel strategies for in-line success Novel solutions for process and designed for in-line use value chain optimization
2. Research Plan and Strategy DigiFoods Annual Report 2020 11 Article To contents
How robots find the best food by Georg Mathisen for NMBU
Robots that pick the finest berries, and sensors that find the best and healthiest raw produce. New technology will provide us with better food at cheaper prices.
Diseased plants result in bad food. Pål’s job is all about robots. He wants more. The sugar content, for exam- Robots are now learning how to to see how the robots can make ple. The colour of strawberries is usu- keep plants healthy, ensuring they the new sensors better, and he also ally the best indicator of when they get the right nutrients and making wants to find out how the sensors are ready to be picked. However, sure that fruits and vegetables are can get the robots to do a better this isn’t always the case. A robot picked when they are at their best. job. This will result in robots that can that measures the sugar content “In time, robots will probably be detect disease and determine when of strawberries can see much more «DigiFoods, will develop doing as many of the tasks as pos- it is best to harvest the food. than we can. sible that need to be done out in the smart sensor solutions fields”, says Pål Johan From. “For example, we make autono- Other examples: Are plants getting mous robots that operate out in the enough nutrients, or do they require to check the quality Food robots fields. We can equip them with a lot more fertiliser? Or do they need to He is a professor at the Faculty of of different sensors to measure as be sprayed, perhaps? of food both on farms Science and Technology at NMBU. much as possible and find out more He leads the work in Pillar 2 Robot about what is happening in the “Currently, we are working on robots and in factories.» and sensor integration. fields”, he explains. that will pick strawberries and treat plant diseases. Once robots are out Not only will it improve the quality of Sugar and fertiliser doing that job, we need to attach as food, it will also make it cheaper and The scientists have been working many sensors as possible and mea- ensure that less is wasted. DigiFoods a lot with strawberries. The robots sure as much as possible so that we will develop smart sensor solutions can find out when the berries are can operate farms in the most effi- to check the quality of food both on ripe and should be picked, detect cient manner”, says Pål. farms and in factories. disease and see if any of the berries are damaged. However, they are now going to start measuring a lot
2. Research Plan and Strategy DigiFoods Annual Report 2020 12 Article To contents
“The majority of time and resources “The technology is certainly going are spent on autonomy. In other to be cheaper than paid labour. For words, getting the robot to operate a high-cost country like Norway, by itself. We use machine learning it will be a great advantage”, says and artificial intelligence to make Pål. And: Agriculture is struggling to robots understand what to do and get enough people to do these jobs, how to move”, he explains. especially now that the corona virus has closed the borders during cer- Pål’s job is all about robots. He states: “In time, robots will probably be doing as many of the Healthier and better tain periods. tasks as possible that need to be done out in the fields”. The purpose of DigiFoods is exactly what it sounds like: To digitalise • Photo/cc: Norsk Idar Jøsang, Dag Landbruk Norwegian food production. Better information enables the better utili- sation of raw produce. It also ensures that food, and therefore people, become healthier. When we know exactly what affects quality, all the way from farm and fjord to fork, it is then possible to learn from it, produce better food and make the whole process more sustainable.
70 to 75% of the price of a food item is the cost of producing the raw produce. Therefore, digitalisa- tion can make good food cheaper. Furthermore, the robots and sensors can help us to throw away less food because the quality improves and the food has a longer shelf life.
Cheaper than people Finally: Isn’t it going to be expensive making robots that are going to cul- tivate relatively cheap berries, pota- toes and vegetables? Not at all.
2. Research Plan and Strategy DigiFoods Annual Report 2020 13 To contents
3. Organization
Organizational structure, and cooperation between the centre’s partners
Board DigiFoods has a decentralized organizational structure. It’s headquarter is at Nofima, Campus Ås. The food industry is by nature decentralized, and Nofima, Innovation Leader group Scientific NMBU and SINTEF have successfully worked together advisory board advisory board Centre leader, pillar with industry and research partners, independent Industry users leaders, communication External of their locations.
Centre administration Research Pillar 1 Education
Nofima SINTEF NMBU
Research Pillar 2
NMBU
Research Pillar 3
Nofima
Research Pillar 4
Nofima
3. Organization DigiFoods Annual Report 2020 14 To contents
The DigiFoods Board oversees that obligations In addition, Mona Gravningen Rygh, the contact person • Pillar 1 Novel sensor systems and application are fulfilled, and decide on financial, partnership for DigiFoods at the Research Council of Norway, development (Lead: SINTEF) and IPR matters, as well as ratifying annual research will be an observer at the Board meetings. • Pillar 2 Robot and sensor integration (Lead: NMBU) plans made by the leader group. In 2020, the Board • Pillar 3 Integrated in-line sensing solutions met for one digital meeting. The Board consists The centre scientific work is organised through close (Lead: Nofima) of the following elected members: collaboration between four Pillars: • Pillar 4 Utilization of large-scale quality assessment (Lead: Nofima)
Furthermore, NMBU leads the recruitment The DigiFoods Board and education process in DigiFoods.
Board Member The leader group manages and leads DigiFoods, such Chairperson and representing as ensuring strategic planning and running of projects, of the DigiFoods Board Board Member the host institution Board Member recruitment of qualified personnel, providing a good working environment, accounting, dissemination and reporting.
The leader group consists of: • Jens Petter Wold (Nofima) – Center Director, overall scientific and administrative leader Anne-Cathrine Whist, Anne Cathrine Eva Veiseth-Kent, Julie Steffensen • Marion O’Farrel (SINTEF) – Scientific Manager TINE Gjærde, NMBU Nofima on behalf of Ingvild Dalen of Pillar 1 (maternity leave), • Pål Johan From (NMBU) – Scientific Manager Lerøy Norway Seafood of Pillar 2 • Nils Kristian Afseth (Nofima) – Scientific Manager Board Member Board Member Board Member of Pillar 3 • Ingrid Måge (Nofima) – Scientific Manager of Pillar 4 • Kristian Hovde Liland (NMBU) – Manager Recruitment and Education • Stine Thøring Juul-Dam (Nofima) – Centre Coordinator • Wenche Aale Hægermark (Nofima) – Communication Leader Mari-Ann Akerjord, Mats Carlin, SINTEF Odd Arne Kristengård, • Anne Risbråthe (Nofima) – DigiFoods Controller Prediktor Maritech
3. Organization DigiFoods Annual Report 2020 15 To contents
The Centre Director and Pillar Leaders, from left Pål Johan From, Jens Petter Wold, Nils Kristian Afseth, Kristian Liland, Ingrid Måge og Marion O’Farrel. An external Scientific Advisory Board is appointed • Photo/cc: Wenche Aale Hægermark, Nofima and will annually review results and research plans and participate at the annual centre meetings to assist in ensuring scientific quality and industrial and societal relevance. The members are: • Prof. Søren Balling Engelsen, Dept Food Science, Univ. Copenhagen • Prof. Bjarne Kjær Ersbøll, Dept. Applied Mathematics and Computer Science, Technical Univ. of Denmark • Ole Alvseike, Head of division Animalia, Norway • Onno de Noord, Advanced Data Analysis Consultancy, Amsterdam.
The centre has also appointed an Innovation Advisory Group with representatives recruited from user companies. The members will oversee, evaluate and advice on how innovation processes are promoted and incorporated in the research activities, including knowledge transfer, learning and innovation arenas, as well as industry involvement and business case development. The members are: • Silje Ottestad, NEO • Marije Oostindjer, Norilia • Atle Rettedal, Robot Norge • Roy Martin Hansen, Lerøy Norway Seafoods
DigiFoods will be organized to facilitate excellent collaboration between three groups of partners: R&D institutions, food companies and sensor, robotics and digital platform companies. The user partners will be Frequent meetings are organized at Board level (each with a news feed where centre participants can post involved in the planning of experiments, execution and six months), Centre level (annual meetings), leader e.g. news, links to documents, research plans, results, discussion of results. Research will be conducted in group (every third week), and thematic or project pictures and videos. In addition to a formal news the end-user process lines and require that scientists, level (as required). In addition to physical and digital channel, the SharePoint will also act as a social media, engineers and user partner personnel are involved. meetings, DigiFoods has an internal SharePoint site thus contributing to build the DigiFoods team spirit.
3. Organization DigiFoods Annual Report 2020 16 To contents • Photo/cc: Eriksen, Sune Norilia
Bioco, a joint venture between Norilia/Nortura and Felleskjøpet Agri, uses enzymatic hydrolysis to refine poultry rest raw materials. This is an interesting industrial case for the Digifoods SFI, with the goal to improve the process even further 17 To contents
Partners
Per April 1st 2021, DigiFoods has 27 partners, where 6 are R&D partners and 21 are user partners.
R&D Partners
Nofima is one of Europe’s largest University of Lincoln has established academic and industry partners. Our key personnel institutes for applied research within an international reputation for the contributing in DigiFoods will be Prof. Simon Pearson the fields of fisheries, aquaculture quality of its research and teaching. and Tracey Watson. and food. Nofima’s vision is “Sustainable Food for All”, Two of the University’s leading research while our objective is to actively contribute to solve centres will participate in DigiFoods, Ulm University (UULM) the large social challenges such as increased food namely the Lincoln Centre for Autonomous Systems is among the leading ‘golden security, better food safety and health, reduced food Research (L-CAS) and the Lincoln Institute for Agri- age’ universities ranked waste and reduced environmental and climate foot food Technology (LIAT). L-CAS specialises in systems #3 in Germany, and #18 in the world (Times Higher prints. We have excellent knowledge in food science integration, bringing together technologies to Education Ranking). The Institute of Analytical and and are recognized for our research on applied bio- tackle challenging real-world applications in food Bioanalytical Chemistry (IABC) operates the Elemental spectroscopy, rapid spectroscopic measurements manufacturing and agriculture, security, assistive care, Analysis Center and the Focused Ion Beam Center of food quality, for multivariate data analysis and and intelligent transportation. LIAT’s mission is to UULM for Micro/Nano Fabrication/Characterization. consumer science over the last 30 years. Nofima is the develop new technological solutions for the business IABC provides ample laboratory infrastructure (1,000+ host institution of DigiFoods and will contribute with of producing food through agriculture at all stages m2), and operates analytical instrumentation including peak expertise in applied spectroscopy (Raman, NIR, of food production including cultivation, harvest, FTIR. IABC has been leading several national and fluorescence, FTIR and hyperspectral imaging), process processing and packaging. The undertaken research international projects specifically for developing analytical technology, data analysis, consumer science is strongly applied, with many links to the local, national advanced vibrational spectroscopic sensing concepts and food science. Nofima will also provide an extensive and global agri-food industry. Our main contribution for industrial, medical, environmental, and food safety state-of-the-art lab for spectroscopic analysis, food pilot to DigiFoods will be with our world leading expertise applications. In DigiFoods, we will provide our expertise plants and food technology labs. Our key personnel within agricultural robotics. The University will in food quality and safety monitoring/sensing techno contributing will be DigiFoods Centre Director Jens also welcome students, PhD scholars, faculty, and logies, sensing networks, and data mining via advanced Petter Wold, Pillar 3 Lead Nils Kristian Afseth, Pillar 4 practitioners from industry to spend time in Lincoln analytical techniques and strategies developed at IABC Lead Ingrid Måge, Senior Scientist Karsten Heia with the objective to strengthen collaboration within ensuring food safety and public health. Especially, and Scientist Lars Erik Solberg. A group of about 16 the centre. We expect that DigiFoods will enable IABC-UULM will develop miniaturized mid-infrared scientists and technicians will also take part in the continued collaboration in agricultural robotics and sensing platforms based on thin-film semiconductor research. new collaboration in food automation, both with and diamond waveguides for analyzing relevant food
3. Organization DigiFoods Annual Report 2020 18 To contents
constituents and pathogens. We anticipate that this SINTEF Smart Sensor Systems Uni. Valencia The Universitat collaborative effort will result in the submission of (SINTEF) has been developing Politècnica de València (UPV) joint publications and the development of further in-line sensor systems for industry, is the only Technical University collaborative research projects. Our key personnel including the food industry, for more than 30 years, in Spain in the top 500 world’s most prestigious contributing in DigiFoods will be Prof. Boris Mizaikoff. resulting in many process-applied publications and universities based on the Academic Ranking of World patents of international relevance. SINTEF has specific Universities 2018. It is particularly relevant in the NMBU’s mission is to contribute to the competence in designing optical measurements areas of Engineering and Sciences and a national well-being of the planet. Our interdisciplinary systems, based on e.g. spectroscopy, x-ray or cameras leader in patent license income and start-up creation. research and study programmes generate and data analysis. A core part of the research involves Multivariate Statistical Engineering Research group innovations in food, health, environmental designing and building robust optical measurement was established with the aim of offering the scientific protection, climate and sustainable use of natural prototypes based on novel measurement concepts, community and the business & technological resources. As a University, NMBU aims to educate moving as quickly as possible from the lab to the field, enterprises a working environment in which to develop outstanding candidates, perform high-quality research and gaining a fuller understanding of the industrial research, development and innovation (RDI) in the area that produces new perspectives, and create innovation. measurement environment. In DigiFoods, SINTEF of multivariate statistical techniques for quality Two research groups from the Faculty of Science Smart Sensor Systems will contribute by designing & productivity improvement. The group is active in Data and Technology at NMBU take part in DigiFoods: The and building new sensor prototypes that are designed Analytics, Six Sigma, Process Analytical Technology Biospectroscopy and Data Modeling group (BioSpec for measurement in the industrial process and adapting (PAT), Multivariate Image Analysis (MIA), Process group), led by Professor Achim Kohler, and the Robotics existing scientific instrumentation to industrial sites Chemometrics and Statistical Methods for Knowledge group, led by Professor Pål Johan From. In DigiFoods, for process characterisation measurements. SINTEF Discovery. Our experience working with industry the BioSpec group will contribute with the development will work closely with the PhD students in Digifoods so and research-based innovation can be very valuable and application of novel handheld and portable that they have a greater understanding of the theory for Digifoods. On the other hand, getting involved infrared devices for quality measurements of food. behind the sensor prototypes, and make modifications in DigiFoods will provide us an excellent opportunity The Robotics group will contribute with competence as required. Our key personnel contributing in DigiFoods to be exposed to the needs of the high-tech food in robotics, in particular agricultural robotics, and will will be Pillar 1 Lead Marion O’Farrell, Senior Researchers industry opening new research lines to get involved. develop robots for automation in food processing and Jon Tschudi, Kari Anne Hestnes Bakke and Trine Kirkhus, UPV will provide joint supervision of at least one PhD sampling. The Robotics group and the Biospec group and Researcher Anders Hansen and Tim Dunker. student on data analytics and real-time process control will collaborate on the development of a robot for & optimization. Our key personnel contributing automated sample handling for infrared and Raman in DigiFoods will be Professor Alberto J. Ferrer-Riquelme. spectroscopy. The two research groups will supervise several PhD students and postdoctoral researchers in the course of the DigiFoods project. Associate Professor Kristian Hovde Liland from the department of Data Science will be responsible for the education of master students.
3. Organization DigiFoods Annual Report 2020 19 To contents
User Partners Food Companies
TINE SA is a cooperative that processes value for our unitholders. Nortura has a strong focus approaches, as well as for pilot and industrial testing. milk for dairy products and owns some on innovation and R&D and is involved in more than We will also contribute with our competence of Norway’s largest food brands. Around 30 research projects. In DigiFoods we will concentrate and know-how on enzymatic hydrolysis, products 10,500 farmers on 8,500 farms are the foundation for our work on our poultry and pork value chains using (raw material, hydrolysates, fats and sediments) TINE’s business operations. TINE has Norway as its main sensors and big data. We expect to optimize and markets (pet food, food and dietary supplements). market, but also subsidiaries internationally. TINE’s our production and processing lines and hope to get DigiFoods will be a great platform to develop strategic goal is to implement Integrated operations more value out of our raw material. By optimizing new knowledge and tools that will enable us to realize (IO) as our future operational standard within dairy processes and products we will achieve higher yield our ambition to add more value to our plus-products. production. For TINE, IO means the integration of and less food waste and thereby reduce the impact Our key personnel contributing in DigiFoods will be people, disciplines, organizations, work processes, on the environment. One main goal with participating Director Business Development Heidi Alvestrand information and communication technology to make in DigiFoods is to serve our customers and consumers and Chief Advisor Bioprocesses and Business Strategy smarter decisions. DigiFoods we will provide us with with even better products in the future. Our key Marije Oostindjer. the opportunity to develop and test technology personnel contributing in DigiFoods will be Research with deeper research requirements, but also Director Per Berg, Development Director Atle Løvland Lerøy Aurora is a world leading company higher potential beneficial outcomes, i.e. a deeper and relevant staff from our factories, quality and in salmon and trout farming and understanding of our raw material - the milk. Our key technical departments. slaughtering, as well as the manufacture personnel contributing in DigiFoods will be Director of products based on these raw materials for the R&D Anne-Cathrine Whist, Technology specialist in Norilia refines and sells rest raw materials consumer market. We have long experience with Cheese and Start Cultures Jorun Øyaas, Manager R&D (plus products), from the Nordic meat handling large amounts of fish, both in the fish farms, Cheese and Fat Products Kjetil Holstad and Technology and egg industry, thereby contributing through the slaughter process and in production Specialist Cheese Kjetil Jørgensen. to a more sustainable and profitable agriculture. of consumer products. Our overall strategy is to secure Our biorefinery Bioco uses enzymatic hydrolysis a sustainable economic future for fish farming Nortura is the largest brand supplier to refine poultry offcuts. There is a large potential for and production, both locally and worldwide. DigiFoods in Norway in the meat and egg refinement of other raw materials as well, and Norilia represents a unique opportunity to share knowledge business, and the company behind has the ambition to implement and industrialize viable and learn from other companies. The possibilities the Gilde and Prior brands. We are organized as a processes. This may include new lines using enzymatic for new knowledge and innovations seem very cooperative, owned by 17 700 Norwegian farmers that hydrolysis on different raw materials, such as bones promising and can be both of a generic nature (sector supply raw material to our customers and factories. and offal from pork, beef and lamb, feather or blood, independent) as well as specific for our business. Our Nortura slaughters, cuts, refines and develops meat or other processes such as fermentation. In DigiFoods, key personnel contributing in DigiFoods will be Factory and egg products that are sold to foodservices, retailers Norilia will offer our process line at Bioco for development Manager Tore Pedersen. and other food related industry with the aim of creating and use of new senor systems and optimization
3. Organization DigiFoods Annual Report 2020 20 To contents
Biomega was founded in 2000 quality, in addition to generic challenges related Lerøy Norway Seafoods is Lerøy’s on the premise of advancing to technology and data handling. Hoff wishes to make quality brand for sustainable whitefish innovative biotechnology to release use of in-line measurements (NIR) either at intake of the caught in the wild – and sourced the full nutritional and functional value of otherwise potatoes or during processing. The NIR measurements from the Arctic seas in the north. The very best raw underutilized side streams from the salmon industry. will hopefully give us useful information concerning ingredients are picked, processed and packaged, then Today, Biomega has a rich patent family of various process control which in turn, and in combination with distributed to markets worldwide. With a history of technologies, with the continuous enzymatic hydrolysis our participation in the projects ROBUST and MODEL, more than 140 years of fishing in these waters, it is safe process at its core. We continuously invest in innovation can help us develop a statistical process control (SPC). to say that our products are the result of developing through R&D to ensure best-in-class technology We also see great value in sharing knowledge and learn and preserving a proud craft. Our main activities are and respond to customers’ needs, including product from other food companies with similar challenges. Our within processing for filet products and ready-to-eat development, traceability, and sustainability. In our key personnel contributing in DigiFoods will be Process meals. Lerøy has high focus on improving the utilization Norwegian biorefinery we turn food-grade fresh salmon and Product Development Manager Ingvild Sveen. of our raw material and thereby reduce food waste raw materials into premium feed and food-grade and increase profitability as well as consumer satisfaction. ingredients. Sophisticated biorefining processes Lerøy Havfisk is a large trawler company Assessing key quality properties by advanced sensors ensure careful separation of nutritional components. in Norway. We have long experience will help achieving this, and by combining data from Biomega’s mission is to transform undervalued raw in handling large amounts of fish different sources – knowledge and improved processes material into premium food and petfood ingredients and facing quality challenges in whitefish can be obtained. In DigiFoods, we will contribute with through accelerated biorefining. In DigiFoods, production, with highly skilled personnel. Our strategy user expertise and production lines and we see this we will be an industrial test facility for new in-line for improved handling of fish is making it possible as a unique opportunity to discuss innovation ideas monitoring solutions, and our expectations is that along to sort fish into different quality grades. These are key and improvements for our quality development work, the DigiFoods lifespan new in-line process monitoring factors, as we see it, in order to secure a sustainable e.g. sensors that are easy to use, practical and cost equipment is devolved that could contribute to a more economic future for the fishing fleet and the land- efficient. Our key personnel contributing in DigiFoods stable production and end-product quality. Our key based seafood industry. DigiFoods represents a unique will be Quality Manager Rune Hansen, Project Manager personnel contributing in DigiFoods will be CTO opportunity to share knowledge and learn from other Julie Steffensen, Leader Technology Project and Andrew Dustan and Director of R&D Bjørn Liaset. companies. The knowledge and innovations Business Development Ingvild Dahlen and Quality Lead to be generated can be both of generic nature (sector Raw Materials Whitefish Roy Martin Martinsen. Hoff SA is Norway’s largest potato independent) as well as specific for our business. processing company, processing It is hard to see that all outlined innovations can 1/3 of Norway’s potato production. Hoff be established without this joint initiative. Our key is producing a range of different potato-based food personnel contributing in DigiFoods will be Operation products and food additives, such as e.g. pommes frites, Manager Odd Johan Fladmark. mashed potatoes, potato starch, potato glucose syrup and potato spirits. We believe that DigiFoods can help us solve specific challenges related to variations in potato
3. Organization DigiFoods Annual Report 2020 21 To contents
Sensor & Robotic
Prediktor Instruments develops NEO Norsk Elektro Optikk AS RobotNorge was established and delivers advanced sensors is a privately owned research company in 2003 as a private spin-off from and associated software for industrial within the field of electro optics. NEO’s ABB Robotics at Bryne. The applications. Our instruments are based on Near main commercial interest is within hyperspectral history goes back to the root of robotics in Norway, Infrared spectroscopy and designed for in-line imaging. Our line of hyperspectral cameras (HySpex) ie the development of the first paint robot in the 1960s. mounting for the purposes of continuous monitoring, is recognized as the most advanced and accurate Now, RobotNorge develops robotic solutions for future process control and optimization. Important customer hyperspectral instrumentation available in the market. production needs. New technology that advances segments are food, feed and dairy industries, with Through the SFI we want to develop new methods sensory, camera and AI is combined with traditional, common needs for controlling their processes, for applying our hyperspectral imaging technology industrial ABB robots. Our vision is to actively minimize the expenses and achieve optimal product to different food industry applications and to develop participate in creating a profitable and sustainable quality. In DigiFoods, Prediktor will actively take part integral customized solutions. We could also be production industry in Norway. Over the past two in the research activities, and we will also contribute interested in designing dedicated instruments for one years, RobotNorge has stepped up activities within with relevant equipment (sensors) for the field trials. or more of the food partners both within imaging and food handling and production. Recent developments We foresee that being part of a long-term research point spectroscopy. Our main contribution in the SFI will within sensor/vision technology, Al and robotics control, center together with the research organizations, food be testing the suitability of our instrumentation provide potential for a new range of advancements and companies and other technology providers will be of for measuring different food quality parameters. better solutions for the food industry. We believe that great value for our business development in general; We have our own camera lab and expertise within data DigiFoods has the potential to become an important through networking, increased understanding of the analysis. Rental of instrumentation for use by other enabling Centre and a catalyst for these developments customers’ needs and challenges, and opportunities partners will also be one of our main contributions. and foresee a Centre which can provide context, for collaborations and sharing of experiences with We expect that DigiFoods will allow us to gain network, shared experience, distribute research project a wide range of technology suppliers. Our key a better understanding of the need for spectroscopic results and give support to new initiatives. Our key personnel contributing in DigiFoods will be CPO information within the food industry and that this will personnel contributing in DigiFoods will be Chairman Mari-Ann Akerjord, CSO Terje Karstang and CSO help us identify new commercial opportunities within Atle Rettedal. Dag Martin Romslo. our field of expertise. Our key personnel contributing in DigiFoods will be Senior Research Scientist Silje Ottestad, Hyperspectral Applications Manager Julio Hernandez and CEO Trond Løke.
3. Organization DigiFoods Annual Report 2020 22 To contents
nanoplus focuses on the and beverage industry. Our key personnel contributing OptoPrecision GmbH is a small, development of customer specific in DigiFoods will be CEO Brian Marquardt, VP of Data yet leading company in research, optoelectronic devices for sensor Analysis Thomas Dearing and Senior Application development, and production applications and has significant experience with Scientist Bharat Mankani. of high quality optical sensing devices. complex coupled distributed feedback (DFB) laser Today, we address with our products applications in the diodes, but also the GaSb material system and Saga Robotics develops robots for the chemical and steel industry, security and observation associated challenges like water-free chip processing. agricultural domain. We have developed business and also in the pharmaceutical market. nanoplus will in particular contribute to Digifoods by the Thorvald platform which is a modular The strategic goal of OptoPrecison is to strengthen bringing in capabilities and related expertise in the field and completely autonomous robot that and expand its business via network actives with of ICL and QCL technology. DigiFoods will enable carries out a wide variety of agricultural tasks. The research institutes and complementary companies us to maintain a strategic position with respect modularity of the robot allows us to operate in open to new fields of applications based on the adaption to emerging technology and related market opportunities fields, greenhouses, and polytunnels where the robot of already available in-house solutions as well as the concerning infrared emitters in the food industry field, uses advanced sensor systems and machine learning joined development of new technologies. In DigiFoods, and to related investigations for future device to navigate autonomously in the field. A very specific we will contribute in terms of developing multi-purpose applications in biophotonics. Our key personnel outcome that we expect from DigiFoods is a close driver electronics for different infrared emitters (LEDs contributing in DigiFoods will be Johannes Koeth. collaboration with developers of sensors and tools or lasers) and detection electronics as well as the that have products or can develop new products that corresponding embedded software to operate these MarqMetrix offers a simple, they would like to put onto our robots to collect large circuit boards for the development of novel sensing stable and powerful Raman amounts of data that has not previously been available technologies proposed herein and for the realization spectroscopy platform built to farmers or researchers. We look forward to sharing of a dedicated analyser platform. DigiFoods provides for field and process applications at a performance level our knowledge and experience in the DigiFoods partner a partner network and an excellent basis and previously available only in costly lab instrumentation. network and see this as a good basis and opportunity opportunity to discuss and push innovative ideas We make affordable solutions that operate at scale to to discuss innovation ideas. We will also offer to the market. Our key personnel contributing monitor and control processes in real-time for efficiency an autonomous robot for field trials with sensors. in DigiFoods will be Markus Naegele. and quality optimization. Our fast and non-destructive Saga will work on integrating sensor systems on field sampling technology allows you to simply “touch” robots and to test these in the field. Our key personnel a sample to analyze gasses, liquids, solids and slurries. contributing in DigiFoods will be CCO Ellen Altenborg, MarqMetrix has years of experience using Raman CTO Lars Grimstad and CEO Pål Johan From. spectroscopy for analyzing lipids, collagen, and carotene concentrations in salmon fillets and cooking oil. We are excited about our participation in DigiFoods because it will enable close collaboration with food companies and third parties to innovate and broaden the applicability of Raman technology in the food
3. Organization DigiFoods Annual Report 2020 23 To contents
Digital Platforms, software and analytics
CAMO is a leader in industrial ldletechs AS was founded in order IBM is a leading global technology analytics and the preferred to stimulate the digitalization in the company engaged in 170 countries partner for industry leaders industry. We develop fundamentally and is transforming to become a global digitizing their value chain. With a world class industrial new tools for people-oriented, understandable artificial cloud platform and cognitive solutions company. For our analytics platform, we help companies optimize their intelligence. In DigiFoods we intend to stimulate clients, these tools and technologies help them improve processes, drive better product quality and efficiency to deeper understanding, creative innovations and more and work in smarter ways, improving production through innovative analytical solutions. Founded robust in-line implementations of modern multichannel and operations, and gaining competitive advantage. in 1984 by Norwegian scientists, Camo has applied quality monitoring instruments, as well as to supply We conduct research and development in the field of analytical science to process and product quality software for quality monitoring, deliver hyperspectral digitalization and blockchain technologies for the area problems for decades. The DigiFoods research centre software in the food production chain and simplify the of food production and distribution. In DigiFoods IBM will address the current knowledge and technology integration of multichannel sensor data from various will foucs on enabling centre innovations and allow our needs to achieve a successful digital transformation sources in the food production sector. DigiFoods will partners to interface and integrate with the IBM Food of the food industry. This is consistent with the strategies provide important market contacts and user feed- Trust platform, and thus enable a value add way to scale of our organization, where as part of our goal of bringing back for ldletechs and enable us to position us in the up innovations for a global market. In addition IBM can, insights from science-based industrial analytics into market. Our key personnel contributing in DigiFoods if needed, engage in the utilization of large scale quality daily operations, the food industry has been an will be CEO Andreas Wulvik and Project Manager Frank measures, where large data volumes are collected and important market for us for many years. Through Westad. analysed to produce actionable insight for end users. the DigiFoods partnership, we will gain valuable insight We will then engage with relevant skills and technology that will help us guide the development of our solutions and the IBM cloud platform, development tools, Al and so they best fit the needs of the food industry. Our key blockchain technologies can be used to develop and personnel contributing in DigiFoods will be Geir Rune test new, innovative technical concepts and solutions. Flåten, Leslie Euceda Wood and Lars Gidskehaug. Our key personnel contributing in DigiFoods will be Camo was acquired by Aspen Technology, Inc in Chief Technology Officer, IBM Norway Loek Vredenberg November 2020. and IBM FoodTrust Europe Espen Braathe.
3. Organization DigiFoods Annual Report 2020 24 To contents
Maritech Systems AS develops CGI was founded in 1976 and is among Intelecy is a software company and provides software and data the largest IT and business consulting creating software for advanced collection solutions to the services firms in the world. Operating analytics for process and production growing seafood industry. Today, Maritech is the market in hundreds of locations across the globe, CGI delivers industry. Intelecy has created software for analyzing leader for business solutions for the seafood industry, an end-to-end portfolio of capabilities, from strategic multivariate event driven and asynchronous timeseries with a market share of 80% of the traders, 50% of IT and business consulting to systems integration, data which is the characteristics in industrial sensor processing fisheries and 80% of food fish. 70% of fish managed IT and business process services and data. Intelecy also handle data ingestion from “factories”, exports from Norway goes through the Maritech intellectual property solutions. CGI contributes to cleaning data, labeling data and orchestration of how software systems. In addition to moving the existing the design and development of physical solutions sensor relates to assets or production lines. For Intelecy, software solutions from on-premise programs to a cloud- for both the agri- and aquacultural domain in close Digifoods will be a place to learn and share knowledge based software solution, we aim to develop novel collaboration with our customers and has a business with a larger community and be able to test how well software solutions for the seafood industry. We believe strategy focusing on FoodTech. The focus has been machine learning algorithms will perform against more that collaborations between the industry and research on business concepts and solutions for securing traditional approaches. Our key personnel contributing institutions are crucial for innovation. Partnership in sustainable food production and animal welfare, in DigiFoods will be COO Espen Davidsen and CEO DigiFoods will enable us to cooperate with partners that e.g. by using computer vision to capture and analyse Bertil Helseth. experience similar challenges in other food industries animal behaviour. CGI is an ambitious company and partners that have experience with the tools that that wants to drive innovation through emerging can be applied to help us develop new decision support technologies and new business ideas. Through the solutions for our customers and thereby increase DigiFoods partnership we intend to engage in both the value of our portfolio. Our key personnel contributing physical product design and prototyping, as well in DigiFoods will be CEO Odd Arne Kristengård, Data as IT/software solutions enabling us to put research into Scientist Helene Seyr, VP Data Science Oddvar Husby action through new innovations. Our key personnel and Director IoT Andre Lillebakk. contributing in DigiFoods will be Director Wilhelm Holmen, Consultant Simen Tofteberg and Junior Consultant Aslak E. Svensson.
3. Organization DigiFoods Annual Report 2020 25 Article To contents
“There is no doubt that DigiFoods will play an important Odd Arne Kristengård, CEO Maritech role in our innovation work in the years to come”
Article by Anne-May Johansen, Nofima
Light-based quality analysis of whole (Automated quality assessment of • Photo/cc: fish can revolutionise the Norwegian headed/gutted whitefish) has now fishing industry. The new technolo- led to Maritech becoming a key part- gy is the result of close collaboration ner in DigiFoods.. Lars Åke Andersen © Nofima. between scientists, technologists and the fishing industry. “We see our participation in active partnerships with key research Maritech Eye is the name of the new institutions as being very valuable. product family. Species and blood DigiFoods strengthens our work on identification of whitefish enable research and development, and is fish producers to find out whether a key driving force for innovation, they have bought top-quality fish learning and sharing of expertise even before they are gutted. between different industries and disciplines. This benefits both us, our Safer and more sustainable industry and Norway as a whole, food production and contributes to safer and more On Monday 23 November 2020, sustainable food production from a Maritech, Nofima and Norsk Elektro global perspective”, says Odd Arne Optics (NEO), in collaboration with Kristengård, CEO of Maritech. fishing industry companies Lerøy Norway Seafood and Lerøy Havfisk, were able to launch Maritech’s solution for hyperspectral quality analysis. The successful collabora- Minister of Fisheries and Seafood Odd Emil Ingebrigtsen joined Nofima scientist Karstein Heia for the official tion in the KVASS research project launch of the solution, which has been named Maritech Eye.
3. Organization DigiFoods Annual Report 2020 26 Article To contents
He points out that heavyweights an institute for applied research, Enthusiastic minister Futher on within the fields of food production with funding from the Norwegian Minister of Fisheries and Seafood Due to the fact that the Maritech and technology are involved, giving Seafood Research Fund, can con- Odd Emil Ingebrigtsen was responsi- Eye solution is compact and suitable the collaboration clout and making tribute to boosting the Norwegian ble for the official launch of Maritech for the industry, it provides a solid it a natural arena for Maritech to be business sector in terms of both Eye. He is thrilled with the collabora- beginning for developing new inno- a part of. technological developments and tive research that has taken place. vations in DigiFoods. innovations in the food industry, is a “We would also like to highlight big deal”, says Karsten Heia. “The project is a prime example of As Nofima has been a significant the exciting collaboration between what is possible when the business partner in developing the solution, green and blue industries, which Maritech is the largest global soft- sector and research communities and therefore knows it very well, new provides interesting perspectives on ware provider to the seafood indus- collaborate on technological inno- applications can be tested efficient- further opportunities across sectors. try, and has developed the machine vations. Better quality sorting is pos- ly, says Karsten Heia. There is no doubt that DigiFoods will which is now ready for commercial itive for both the profitability and play an important role in our inno- sale. Norsk Elektro Optikk (NEO) has reputation of the Norwegian sea- vation work in the years to come”, developed the hyperspectral cam- food industry”, says Ingebrigtsen. says Kristengård. era used for the quality assessment of whole fish. The industrial trial was
Big deal carried out at Lerøy Norway Seafood’s • Photo/cc: For decades, Nofima scientist Båtsfjord facility in Finnmark. Karsten Heia has been working on the development of the technol- “The collaboration has been extre Anunatak Elin Herjehagen, ogy that has resulted in Maritech mely constructive, very close, and is Eye. The first time he attempted to a big part of the reason why we are assess fish quality by using imaging where we are today”, says Odd Arne spectroscopy was in 1997. Back then, Kristengård. it took half an hour to measure a 10 x 10 cm skinless piece of fish. Now, “We would like to commend Nofima Maritech Eye can assess whole fish and NEO for being as on the ball as at industrial speeds. they have been, and we would also like to praise the initiative taken by “Being able to utilise imaging spec- Lerøy, and the willingness and cour- troscopy in an industrial setting is age of all the organisations to really a major milestone for me as a sci- go for it and do something they are entist, for Nofima as a research not used to”, he says. institute and for the Norwegian Senior Scientist at Nofima Karstein Heia together with Maritech’s KVASS project manager Nils Petter Farstad and fishing industry. The fact that we, IoT architect Jan Rune Herheim, have worked closely together on the development of the new technology.
3. Organization DigiFoods Annual Report 2020 27 To contents
4. Scientific activities and results
Pillar 1 Novel sensor systems and application development
In this Pillar, we will focus on the development of new sensor systems that will enable in-line measurement of food quality features that are not possible today.
We will explore solutions that are based on high- FTIR Handheld and portable IR resolution spectroscopy, imaging sensors and low- FTIR generates highly resolved, information-rich spectra. This project is tightly related to the FTIR project. powered spectral sensors. There are several industrial In this project, the focus will be on the determination The work in 2020 was mainly related to planning 2021, partners in DigiFoods that are at the forefront of of more complex food quality parameters in industrial and this was done in close collaboration with other developing in-line food measurement technology, and applications, such as specific fatty acids and proteins research partners and industry. IIn this part of the in 2021 we will start a range of research projects that in dairy processes, or measurements for controlling centre, we will use sensor technology that is under focus on developing new in-line applications using bioprocessing of residual raw materials. Since water development in two on-going ICT photonics projects, hyperspectral imaging, FTIR, Raman and IR. in aqueous solutions efficiently absorbs some of the for different applications. We will adapt them for food wavelengths in FTIR spectra, dry-film analysis quality measurements in Digifoods. We have planned All of these new applications will be explored in is particularly interesting for protein characterization, to set up the optical components in preparation partnership with both end-users in the food industry since multiple protein-related infrared absorbances for two prototypes in the spectroscopy lab at NMBU. and potential technology providers with the potential are “buried” when water is present in the sample. One of the prototypes will, at the end of the to realise results into a commercial product. The development, be the size of a mobile phone, and will research partners in these projects bring with them One of the first steps in this research project will be used for quality monitoring tests. The other prototype expertise in sensor design, applied spectroscopy, be to develop a portable FTIR system for dry-film will be larger and likelly be used for reference analysis. machine learning and food science, and will take measurements that can be brought close to industrial The monitoring prototype will be the first setup primary responsibility for supervising the PhD students. process lines, enabling industrially relevant to establish in the lab in 2021 and will be based on LED Another activity in Pillar 1 is the exploration of new measurements. SINTEF will take responsibility technology from partner nanoplus. It will be assembled opportunities, for brainstorming and discussing for developing this instrumentation. in collaboration with research partner Ulm University new ideas and feasibility studies. and partner OptoPrecision. The instrumentation will be By the end of 2021, the goal is to have collected sufficient tested for different quality scenarios that are relevant Pillar 1 is led by Senior Scientist Marion O’Farrell at FTIR data for the first PhD-paper related to how sample for food producers in the centre. The obtained data will SINTEF Digital. Key partners in this Pillar include Lerøy preparation affects FTIR analysis of protein-containing be further transferred to and integrated into platforms Aurora, Lerøy Seafoods, Lerøy Havfisk, Nortura, Norilia, liquids. This will ensure that we have a portable FTIR such as the block chain-based IBM Food Trust™ data Biomega, TINE, MarqMetrix, nanoplus, Prediktor and system that has been tested in the laboratory, making platform for tracing food contamination and improving Nofima. it ready for testing in industrial environments. food quality.
4. Scientific activities and results DigiFoods Annual Report 2020 28 To contents • Photo/cc: Sintef
RAMAN The project RAMAN will study how Raman spectroscopy can measure quality parameters such as fatty acid and OPPORTUNITIES protein composition in different foods. The focus of the This project will give the centre the opportunity project will be on novel sampling strategies and the use to incorporate new ideas and needs for rapid of state-of-the-art technology to reduce sampling time in-line quality measurements into the project, and make Raman suitable for process measurements. Sensor prototypes for measuring for instance chemical complexity or to modify their initial ideas based on what in food samples will be build and tested. Raman spectroscopy shows great promise for future we learn along the way. This project will give us the smart food sensor systems. Raman relies upon inelastic chance to investigate possible novel applications scattering of photons, known as Raman scattering. and cutting errors in salmon fillets will be tested. Fatty based on present and future ideas and needs. New The method is gaining increasing interest for its ability acid composition will be addressed together with the technology is being constantly developed in univer to capture subtle chemical distinctions in foods. Recent RAMAN project where focus from HYPERSPEC will be sities and research institutes, or is being made feasibility studies show how the techniques can be used on the applicability of hyperspectral imaging in the commercially available on the market, or it is falling to quantify complex food quality features such as fatty SWIR region as an alternative to Raman spectroscopy. to a more reasonable price range. It is important acids in muscle foods, sugars in apples, and mineral for the partners in DigiFoods to have an overview and bone contents in meat slurries. Based on hyperspectral imaging and point measure of these opportunities, so they can pivot and adapt ments a feasibility study will be conducted on detection accordingly, seizing new opportunities as they arise. In recent years, robust and more low-cost Raman of soft tissue (Jelly-fish) in Greenland halibut instrumentation has become available. The goal (Reinhardtius hippoglossoides). The same will be tested Nofima will take responsibility for developing new of our partner MarqMetrix is to develop and provide for whitefish. In both cases the fish will be measured application concepts from the end-user perspective affordable and practical in-line sensor solutions before cutting into fillets. In this work also protein, and SINTEF will be responsible for coordinating also to the food industry. fat and water will be addressed. tracking of new enabling measurement technology, e.g. through conference attendance, organising HYPERSPEC Automated decapitation and cleaning of round fish workshops and generating technology roadmaps. will continue on-going R&D on the industrial application is a widely used practice onboard fishing vessels. Tests NMBU will also partake in these activities. of hyperspectral imaging, with a long-term emphasis will be conducted to combining Maritech Eye and 3D on achieving robust measurements using lower-cost laser scanning for automatic identification of cutting In 2021, the goal is to deliver an overview of new cameras. The main focus in 2021 will be to improve the errors and improper cleaning of the belly cavity. application possibilities that arise, a summary modelling of interaction between light and sample of exploratory work on a selected technology tissue. Furthermore, to test a set of different applications Post Doc Samuel Ortega Sarmiento will work on of interest, and a roadmap of new spectroscopic based on Maritech Eye and point measurements. strategies for combining Magnetic Resonance Imaging technology that can be of interest. and other reference methods for robust industrial Combining hyperspectral imaging with 3D-laser applications of hyperspectral imaging, as well as measurements solutions for colour, fat, melanin improving physical modelling tissue/light interactions.
4. Scientific activities and results DigiFoods Annual Report 2020 29 To contents • Photo/cc: Kristoffer Skarsgård, Saga Robotics
Pillar 2 Robot and sensor integration
Robots and sensors are important in several different areas of the food industry. The rise of the agri-tech sector has shown a demand for robots and sensors to work closely together to increase the performance and accuracy of production both in outdoor and indoor systems. To contents
In this pillar, we will look at how robots can be used ROBOSENSE • Photo/cc: Kristoffer Skarsgård, Robotics Saga to enhance the performance of sensors by accurate The aim of this work is to enable in-line measurement positioning of sensors for optimal sample taking of heterogeneous foods by robotic control of smart and measurements. We will also look at how sensors sensors. We will develop new robotic techniques can be used to increase the performance of robots to optimize the performance of current and new sensors and improve decision making process and overall mounted on robotic arms. Robot arms can use sensory performance. We will develop fully autonomous robots input to accurately position the sensors to guarantee and automatic sample preparation and enable in-line optimal measurements of food that vary in size measurement of heterogeneous foods by robotic and shape. This will give a more reliable and robust control of smart sensors. set of data and can be used to build better models and analytics. We will develop new prototypes for use In this work, we will develop automatic samples under real process conditions, i.e., in processes that preparation for high-throughput spectral fingerprinting have a very high throughput and high requirements
of biological liquid samples by FTR and Raman, which for speed and reliability. The sensors to be used in this • Photo/cc: Kristoffer Skarsgård, Robotics Saga is closely related to the work done in the other pillars. work will be developed in close collaboration with The work will develop fully automated robotic systems the other pillars in DigiFoods. for separating phases such as microbial biomass from media and protein solutions. We will test the solutions MOBILESENSE in fully automated screening micro-cultivation systems The purpose of this work is to develop fully autonomous and different scale bioreactors. Machine learning robots for automatic collection of large-scale quality and advanced AI is used to optimise the sample data in the field. We will integrate a suite of sensors preparation and provide improved and more efficient on the Saga Robotic’s Thorvald mobile platform measurements. for exploration purposes in open fields. This will give large amounts of data over time and space and increase The research area is led by Prof Pål Johan From at our understanding of how to collect and analyse date NMBU and divided into two main projects. Key partners in the agricultural environment. As a part of this work, in this pillar include Saga Robotics, Robot Norge, TINE, we will develop autonomous robots for measuring Prediktor, Hoff, Norilia, BioMega, Nofima, SINTEF quality features in potatoes using NIR and investigate The Robot Thorvald is a fully autonomous robot. New robotic and University of Lincoln. how to use sensory input to predict yield and quality techniques will be developed to optimize the performance on strawberry farms. of current and new sensors.
4 Scientific activities and results DigiFoods Annual Report 2020 31 To contents
Pillar 3 Integrated in-line sensing solutions
When a food sensor has been developed in a controlled environment, there is still a long journey to industrial implementation. Several commercial food sensors have failed because they were not robust towards the inherent bio-variability encountered in the processes and products.
Strategies that address the practical and theoretical Many of the sensors proposed in DigiFoods will provide considerations for sensor implementation is clearly previously unavailable information from food processes. needed, both for the instruments that are already We will document and map quality variations along used commercially, but not least for techniques processes and over time with in-line sensors, evaluating with very limited industrial use, such as FTIR, Raman the potential for process improvements, real-time and fluorescence spectroscopy. process control and product differentiation.
In Pillar 3 we will develop and validate efficient Pillar 3 is lead by Senior Scientist Nils Kristian Afseth solutions and strategies for successful sensor at Nofima. Key partners in this Pillar include all food implementation in food production. Put in other partners, NEO, Camo, Prediktor, Marqmetrics, Idletechs, words: We will make the sensors actually work in the Sintef and NMBU. food companies. We will develop the appropriate tools “We will make the sensors for robust calibration of real-time industrial sensor ROBUST systems, meaning enabling the sensors to actually A major bottleneck for industrial sensor implementation actually work in the food provide the user with reliable quantitative outputs. is to get from the measured signal to reliable estimates of food quality attributes. A robust calibration model companies.” Food handling and processing is indeed complex, needs to handle chemical and physical sample and one sensor is not always enough to provide variations as well as harsh and changing surroundings. sufficient information on a particular process. Therefore, Spectroscopic sensor technology has many application we will also develop novel combinations of sensors areas in in-line food quality analysis. to provide more comprehensive and precise food quality information. Last, but not least, we will use Some applications are well established, and robust implemented sensors to explore and map variation calibration models can be purchased from instrument in food processes over time. vendors. New or less standardised applications require development of new calibration models, which can be a time consuming and expensive task. Also, it is
4 Scientific activities and results DigiFoods Annual Report 2020 32 To contents • Photo/cc: Jens Petter Wold, Nofima frequently necessary to maintain models over time for both instrumental, environmental and process reasons.
In ROBUST the main aim is to define strategies and methods for efficient and robust calibration and maintenance of in-line spectroscopic instruments. This will be based on collecting relevant calibration and process data from in-line processes at selected industry partners where all aspects relevant for robustness will be emphasized including positioning the sensor, sampling and considering environmental factors (such as temperature, light etc.). Building on this, further work will focus on the investigation and development of appropriate approaches for modelling, calibration and maintenance.
Finally, performance of models will be assessed over time in-line at industry partners with the objective of improving the understanding of industrial processes both in terms of distributions and dynamics, but also in terms of relationships between processing stages.
The implemented sensors will be used to explore and map variations in food processes over time.
4 Scientific activities and results DigiFoods Annual Report 2020 33 To contents
Pillar 4 Utilization of large-scale quality assessments
In this Pillar, we will develop data-driven solutions for process, product, and value chain optimisation. The solutions will be based on extensive food quality measurements, combined with other relevant data sources from farm, industry, and consumer.
What data scientists spend their time doing. There is a strong link between health and welfare and reduce food waste. We will investigate consumers’ of animals, fish and plants, and the resulting food quality. attitudes and willingness to pay for different quality Decision support for farmers involves for instance categories, and from that develop communication What data scientists optimised feeding, care, and time of harvest, as well and marketing strategies to target different consumer spend their time doing: as early detection of health and welfare threats. profiles. We will investigate how the growing focus We will combine food quality measurements with data on food waste may impact food choice with respect on environmental and husbandry factors to investigate to product quality. how they affect quality and health. This knowledge can be used in either long-term production planning Pillar 4 is led by Senior Scientist Ingrid Måge at Nofima. or in real-time decision support. Key partners include all food companies, Intelecy, CAMO, Ideltechs, CGI, Prediktor, Maritech, IBM, NMBU and UPV. In/on/at-line food quality measurements can be used to monitor, optimise, and control production processes. COMBINE We will develop solutions that transform the multitude Data preparation is a crucial and resource-demanding of measured and registered data in a production line part of any data science project, especially when into meaningful information needed to adjust and we need to combine data of different types and from stabilize the production or tailor-make specific different sources. Data preparation includes operations 3% Building training sets end-product quality categories. As in farming, such as cleaning, synchronising, aggregating, 4% Refining algorithms the information can be used in either long-term transforming, structuring, and validating data. 5% Other improvement work or real-time monitoring and 9% Mining for patterns optimisation. In COMBINE, the research partners and analytics 19% Collecting data sets companies will perform a systematic review of typical Well-documented and tailored food products data types and challenges in the food industry, along 60% Cleaning and organizing data can contribute to increased consumer satisfaction with available strategies and methods to solve these
4 Scientific activities and results DigiFoods Annual Report 2020 34 To contents • Photo/cc: JChris Liverani, Unsplash challenges. We will base the review on concrete cases from the DigiFoods food industry partners and available solutions from the analytics companies. The main aim of the project is to identify needs for further research and development in the data preparation pipeline.
MODEL All data-driven solutions require some form of data modelling. In DigiFoods, the models will typically relate quality attributes to controllable and uncontrollable factors from farming or processing. We will develop methodology for two types of models: path modelling “These models extract and real-time modelling. quantitative information Path models are intended to model causal relationships between sets of factors. These models extract quantitative about causal effects, information about causal effects, and thereby lead to increased insight that lay the foundation for and thereby lead to improvement work and decision support systems. Such models need to be transparent and interpretable increased insight that lay but will usually not be used for making predictions. the foundation for The second type of model is intended for real-time monitoring, control, or decision support. Such models improvement work and need to provide precise, reliable, and robust predictions, often at high speed. They also need to detect and decision support systems.” diagnose deviations from normality at an early stage and be formulated in such a way that they can be implemented in real time.
Two PhD students will be associated with the project, each addressing one of the modelling types. One or two By combining data from different sources and locations industry cases will be selected for each of the two appro in the value chain, we will enable efficient data analysis aches in 2021, and possibly more the following years. and lay the foundation for successful data science projects.
4 Scientific activities and results DigiFoods Annual Report 2020 35 To contents • Photo/cc : Kristoffer Skarsgård, Robotics Saga
During a workshop at a farm in Kent (United Kingdom), Saga Robotics and University of Lincoln are testing autonomic navigation of robots. From left Pål From (Saga Robotics), PhD. Grzegorz Cielniak (University of Lincoln), Jaime Pulido Fentanes (Saga Robotics) og Michael Hutchinson (Saga Robotics).
(Image to the left) From left Lars Grimstad, Michael Hutchinson and Halvard Grimstad (all Saga Robotics) are making the last fine-tunings before testing autonomic navigation. To contents
5. International collaboration
DigiFoods has established close collaboration with three excellent foreign research groups and three foreign high-tech technology providors who will be important for carrying out the planned research. They will take active part in the running projects and share supervision of PhD-students. Exchange of PhDs and post docs will be planned.
Foreign technology companies are also partners secondary structure changes in proteins. UUlm 5. nanoplus Gmbh, (Gerbrunn, Germany), is in DigiFoods, since they offer technology of interest will take part in the project IR and develop this represented by Johannes Koeth. They will contribute to the centre and Norwegian food industry: platform further for in-line measurement by bringing in capabilities and related expertise of protein, lipid composition in foods and dairy in the field of Quantum cascade laser (QCL) and 1. University of Lincoln (ULin), (UK), is represented and bioprocess control. Interband cascade laser (ICL) technology. nanoplus’ in DigiFoods by Professors Simon Pearson main task is to support in combining QCLs with and Gregorz Cielniak and their research group 3. The Polytechnic University of Valencia (UPV), waveguide technology developed by UUlm for at Lincoln Institute of Agri-food Technology. They (Spain), is represented by Professor Alberto J. in-line measurement of complex structures and will contribute with expertise in autonomous Ferrer-Riquelme, group leader of the Multivariate composition in food samples in-line, such as fatty and long-term navigation of agricultural robots, Statistical Engineering Group. The group is devoted acid composition. This will be explored in the IR sensor and implement integration and data to research, development and innovation activities project. gathering, management and analysis. The university in the area of multivariate statistical techniques has a research farm with more than ten of Saga for quality and productivity improvement 6. OptoPrecision GmbH, (Bremen, Germany), Robotics’ Thorvald robots that can be used and mega-database analysis. Professor Ferrer represented by Markus Naegele, is a leading for extensive testing in a realistic environment. will participate in the MODEL project and provide company in research, development, and production They will take active part in MOBILESENSE joint supervision of PhD students and on data of high-quality optical sensing devices and will and ROBOSENSE. analysis and real-time process control. contribute by developing laser-driver and detection electronics in conjunction with the corresponding 2. Ulm University (UUlm), (Germany), is represented 4. MarqMetrix, (Seattle, USA), provides modern, embedded software to realize a dedicated in DigiFoods by Professor Boris Mizaikoff, director easy to use Raman instruments for rapid material analyzer platform in Pillar 1 and the project IR. of the Institute of Analytical and Bioanalytical analysis and process measurements. They are Chemistry (IABC). UUlm has developed mini represented by Brian Marquardt, world leading aturized mid-infrared sensing platforms based in development of process Raman systems and on thin-film semiconductor, oxide/nitride, very interested in novel food applications based and diamond waveguides that have already on Raman. He will contribute with knowledge demonstrated their potential for analyzing e.g., and instrumentation in the project RAMAN.
5. International collaboration DigiFoods Annual Report 2020 37 To contents • Photo/cc: Håkon Sparre, NMBU PhD students at NMBU Robotics Group and Biospectroscopy collaborate on automated sampling. To contents
6. PhD recruitment and education
DigiFoods is planning to have a total of nine PhD fellowships and three post- doctoral fellowships associated with our research over the lifetime of the centre. The first PhD positions have been announced, and several candidates have been employed already.
These cover a large range of applications and At Nofima and SINTEF Bijay Kafle will build and test agricultural robots to gather data on yield estimates, instrumentations in the food industry. Their projects an FTIR protype system for analysis of dried liquid diseases, ripeness, etc. and develop algorithms for cover key areas from methodological and instrumental samples, combining development of new applications robotic arm controlled near-infrared sensors in food developments, optimal deployment and usage of sensors with industrial testing of the FTIR prototype. production lines. and analysis and understanding of sensor data. At Nofima in Tromsø post-doc Samuel Ortega This part is led by Kristian Hovde Liland. A connection At Nofima in Ås, Tiril Aurora Lintvedt, started her PhD Sarmiento will work on strategies for combining to the master programs in data science at NMBU has work on in-line Raman spectroscopy, aiming for Magnetic Resonance Imaging and other reference been established by offering relevant master thesis representative sampling and modelling of hetero- methods for robust industrial applications of hyper topics for students finishing their master education geneous foods. Christian Thorjussen will develop spectral imaging, improving physical modelling in 2022. The available projects include machine statistical path modelling approaches, aiming at better and light interactions. learning and calibration transfer for hyperspectral data understanding of factors and mechanisms causing and analysis of production data from food industry variation in food quality. Marco Cattaldo, enrolled At NMBU a PhD will develop a prototype hand-held and aquaculture. When the projects of DigiFoods have at Universitat Politècnica de València, will develop IR instrument for food quality applications. Antonio all started, potential for further relevant master thesis statistical methods for process and product optimisation Candea Leite will be employed as a post doctor to work topics, for students finishing in 2023 and beyond, is high. based on real-time measurements of food quality. on robot–sensor integration on autonomous mobile
Location Candidate Funding Project 2020 2021 2022 2023 2024 Nofima Tiril Aurora Lintvedt Nofima RAMAN
Nofima Christian Thorjussen Nofima MODEL
Nofima/UPV Marco Cattaldo RCN MODEL
Nofima/SINTEF Bijay Kafle RCN FTIR
PhD-students NMBU NN RCN IR
Nofima Samuel Ortega Sarmiento Nofima HYPERSPEC
Post- docs NMBU Antonio Candea Leite RCN MOBILESENSE
6. PhD recruitment and education DigiFoods Annual Report 2020 39 Article To contents Getting down to the nitty-gritty of food quality by Marion O´Farrell, Sintef and Nils Kristian Afseth, Nofima
Whether you are making yoghurt, meat products or fish oil, a food producer must know that the product has the quality it should have and that the label correctly describes the product.
Some food and feed producers make other components such as urea, free is based on quickly drying the liquid measurements using near-infrared fatty acids and the content of some sample before the analysis, so that light – light in the frequency spec- individual fatty acids. It has also been the liquid is reduced to a dry film. In trum just above what is visible to us shown, for example, that the fatty this way, the negative effect of water, humans – to measure the total fat acid concentration in milk will fluc- a very efficient absorber of infrared or protein content of the products. tuate slightly from day to day for an light, is eliminated from the analysis. This is how a sausage maker can say individual cow. It is known that being In addition, the components of inter- that a hotdog has 18% fat, or minced able to measure these lactation fluc- est are concentrated so that their meat can be sold as either 5% or 14% tuations regularly can enable bio- chemical information is more visi- “The laboratory based FTIR fat. These measurements can even logical modelling and reasonably ble. Based on our findings so far, we be integrated into the process line tailored healthcare approaches for believe three important elements systems used in the dairy to control or monitor the process in individual cows, which in turn could need to be addressed to advance real time. lead to greater productivity. in-line FTIR of liquid processes fur- industry are all based on ther: 1) capitalising on more readily However, in some instances this The laboratory based FTIR systems available compact and cost-effi- measurements performed information is simply not enough. used in the dairy industry are all cient FTIR components 2) achieve You may need to know, not only based on measurements performed automatic and repeatable in-line directly in the milk.” the fat content, but the fatty acid directly in the milk. In the last six preparation of dry films from liquid profile; not only the total amount years, Nofima, SINTEF and Prediktor samples 3) long-term trials to get a of protein, but the length and com- have been collaborating on liquid better understanding of the natural position of the protein chains. In the analysis research that aims to prog- biological variation in the samples. dairy industry, laboratory based FTIR ress beyond current in-line FTIR sys- systems are widely used to deter- tems that require complex pump In general, FTIR instrumentation mine major components such as fat, and tubing systems for measuring is expensive, complicated, and protein, lactose, and to some extent small liquid volumes. Our approach best suited for use in laboratories.
6. PhD recruitment and education DigiFoods Annual Report 2020 40 Article To contents
However, FTIR instruments often a prototype that can be adapted to hydrolysis of salmon by-products. developing methodologies for pro- provide highly resolved data, mean- the product line of the individual Since this research, two important gressing to at-line measurements on ing they often generate more data manufacturer. This was tested, with things have happened; there has the compact instrument. An import- than required for sufficient accuracy. Nofima and Prediktor, on relevant been a rapid development in the ant aspect of the FTIR project is This can give lead the way for simpli- industrial samples from bioprocess- field of miniaturized FTIR instru- developing a suitable sample prepa- fying the instruments, for example ing of by-products where enzymes ments (µFTIR), which are smaller and ration method that can be conduct- by limiting the number of wave- are used to recover proteins from cheaper than traditional FTIR instru- ed at-line, and eventually achieving lengths, or using cheaper detectors, by-products. The approach was suc- ments, and there has been growth the goal of in-line measurements. or more compact components. For cessfully demonstrated at-line for in the development of standard FTIR The automated sampling research example, SINTEF developed a low- monitoring the hydrolysis process at instrumentation with smaller foot- will be done in collaboration with er-cost FTIR solution that worked Biomega, a bioprocessing plant that prints and lower weight (less than 2 robotics experts at the Norwegian without the expensive laser and built has industrialised enzymatic protein kg), which could be compatible for University of Life Sciences (NMBU). use on the farm or integrating into process lines. The second project, led by the BioSpec group at NMBU, is close- It is, therefore, excellent timing that ly connect to the FTIR project and DigiFoods has started. Two of the will focus on using novel mid infra- projects in the centre focus on the red (MIR) light emitting diode (LED) industrial potential of more complex sources that are being developed in food measurements that require two on-going EU projects under the measurement in the mid infrared ICT photonics programme. The goal range. One of the projects will prog- is to investigate developing proto- ress from the research described types based on these components above, where a compact FTIR sys- that are simpler, can potentially be tem will be developed based on low-cost, even handheld, and test state-of-the-art components and them on selected applications. For will be adapted so it can be brought these analyses, measurements in close to the process line in industry, both liquid and dried state are pos- enabling a PhD student to do indus- sible. The overlap between these two trial research that is relevant to indus- projects is very interesting, where try. SINTEF will take responsibility the compact FTIR prototype will be for developing the instrumentation compared with the simpler hand- TINE and ensuring good robust measure- held IR prototype for the complex ments. Nofima will work closely with liquid measurement applications. the PhD student on using existing, Photo/cc: • Photo/cc: lab-based FTIR instrumentation and FTIR is a potential tool for monitoring of raw milk quality variations in dairy industry.
6. PhD recruitment and education DigiFoods Annual Report 2020 41 To contents
7. Communication and dissemination
Several relevant trade magazines acknowledged the grant of SFI DigiFoods
7. Communication and dissemination DigiFoods Annual Report 2020 42 To contents
In DigiFoods the purpose of the communication is to present inventions and know-how from DigiFoods research as well as network development and knowledge exchange.
Our communication principles: Our priority target groups:
Our communication will be active and targeted, to make sure that Food and bioindustry, Proactive Industry our research results reach our technology companies target groups.
We will be open about how Scientific Open Scientist and students and why we conduct our research. community
We will be honest and accountable The Public, including funding bodies Honest Public in our communication. and policymakers.
The official opening of the center was in Dec 2020. The majority of the dissemination activities in 2020 are linked to the grant in June. In addition, some of our industrial partners have acknowledged and promoted their participation.
7. Communication and dissemination DigiFoods Annual Report 2020 43 Article To contents
The devil is in the details Wenche Aale Hægermark, Nofima
“It was my interest in quantum physics that led to my fascination with optics and spectroscopic measurements,” says Tiril Aurora Lintvedt.
A minced chicken sample is being prepared in an elongated sample container, in which it will be sent down a conveyor belt at different speeds.
She is now a research fellow at soil or cowshed - to the table. Raman means more applications. It was • Photo/cc: Tiril Aurora Lintvedt, Nofima DigiFoods and investigating how spectroscopy is eminently suitable previously used in industries with Raman spectroscopy can be used for undertaking highly detailed expensive raw materials, e.g. phar- in the food industry for acquiring molecular measurements. Thus, it is maceuticals. We are now investi- detailed, accurate measurements possible to measure not just the dis- gating how Raman measurements of the structure and quality of raw tribution of fat, protein and water in of food raw materials can help to materials. In addition to using her raw materials, but also fat and pro- increase raw material utilisation, knowledge of physics for calculating tein types and quantities. ensure uniform product quality or how to obtain information from the develop differentiated products.” light spectrum of raw materials, she “Technological developments are is constantly learning about food on our side, enabling us to be Measures fatty acids composition chemistry. increasingly faster and more pre- in salmon and bone in chicken cise. Raman spectroscopy initially by-products Food quality control throughout consisted of one focused laser that One of the quality targets for farmed the value chain measured a 1 mm spot on a sample. salmon is its marine fatty acid con- There are several different ways of Now there are Raman probes that tent: the higher the percentage undertaking spectroscopic mea- can measure 6 mm spots, and more of omega-3 fatty acids, the better. surements suitable for various raw importantly, we can scan many Chicken by-products should contain materials and measurements. At spots on a sample, thus making as little bone as possible. Tiril is inves- DigiFoods, several of these meth- measurements more representative tigating how Raman can be used for ods will be used for developing real- of a larger area,” explains Tiril. measuring the fatty acid composi- time measurements on a large scale. tion of salmon and bone residue in The aim is to enable food producers Supervisor and head of the DigiFoods chicken by-products. to control raw materials and food Centre Jens Petter Wold agrees. quality, all the way from the ocean, “Cheaper, better Raman technology
7. Communication and dissemination DigiFoods Annual Report 2020 44 Article To contents
the speed we want them to mea- More responsible production sure. It is harder to measure whole DigiFoods is aiming to develop pieces or fillets than mixtures con- robots with inbuilt smart optical taining an even distribution of fat or sensors. It should be possible to bone,” says Tiril. make these sensors measure the most suitable spots, at the right Functional measurements speed and frequency – and provide for the industry answers about exactly what needs to She will now start testing to see what be identified. is needed to ensure that the mea- surements work just as well when Tiril is keen to be part of the solu- whole salmon fillets are transported tion and help to achieve responsible on the conveyor belt. The aim is to production. develop strategies that will enable measurements to be made at the “If manufacturers are to succeed, same speed as that which applies they need correct, understandable to fillets moving along the process- information that they know how to ing line. The measurements must use. We need to make the system
Photo/cc: Tiril Aurora Lintvedt, Nofima Lintvedt, Aurora Tiril • Photo/cc: also be equally precise for each fillet, sustainable. Our aim is optimal All of the 30 centimetres of sample will be measured when they pass the Raman probe on the belt. even when they follow each other usage of all raw materials. I think closely. it’s right to focus on the industry. They already have many automat- She is experimenting to see how fast Raman probe has been scanning Another challenge Tiril is starting to ic processes, so why not make them samples on the conveyor belt can samples and capturing enough solve is to find the best positions for smarter,” she asks rhetorically. move to ensure that Raman mea- details for measuring fatty acid com- taking measurements on salmon surements can capture the neces- position. Sometimes the aim has fillets in order to obtain the informa- “DigiFoods sary details in light spectra and pro- been to investigate how fast chicken tion required. is aiming vide exact predictions. The results by-products can be conveyed while to develop are promising. still undertaking suitably precise “Our research to date suggests that robots measurements. the belly provides the best informa- with inbuilt Many speed trials have been con- tion, simply because there is more smart optical ducted. At times the conveyor belt “So far I have mainly tested minced fat there. But we need to undertake sensors.” in Nofima’s meat hall has been run- raw materials and have used the more measurements and analyses ning constantly. Sometimes cups results to calibrate the instruments before we know exactly how we get of minced salmon have been mov- so that they measure exactly what the most representative measure- ing along the conveyor while the we want them to measure, and at ments,” she says.
7. Communication and dissemination DigiFoods Annual Report 2020 45 To contents
Click to play the video
DigiFoods Annual Report 2020 46 To contents
Smart sensors – sustainable foods To contents
SFI Digital Food Quality (short named DigiFoods) is a centre for research-based innovation (SFI) with the purpose of developing smart sensor solutions for food quality assessment directly in the processing lines, throughout the food value chains.
digifoods.no