Many thanks to Lazaros Gymnopoulos project Quality Assurance and Risk Manager, representative of the Coordinator, for taking the time to meet with us to discuss the development of the PROTEIN project. Lazaros, why did you decide to develop this project?
The Visual Computing Laboratory (VCL) of the Centre for Research and Technology Hellas (CERTH) / Information Technologies Institute (ITI) is now very experienced in applying Computer Vision and Machine Learning research results in various application areas such as Security, the Internet of Things (IOT), E-learning and Gamification, Media and Content Convergence, 4D Reconstruction and Tele-immersion, and Migration.
For some time now we have observed that several of our algorithms and ICT research in general are perfectly applicable in personalized nutrition areas including Food Intake Assessment, AI for Diet Personalisation/Adaptation, Pre- / Post-meal Analysis, Dietary Behavioural Training through Gamification.
At the same time, the Institute of Applied Biosciences (INAB) at CERTH, aims to bridge the gap between bioscientific research excellence and applications in order to meet the evolving needs of the end-user. INAB is involved in the development and use of biotechnological applications taking full advantage of the current advances in the field with emphasis on functional genomics and molecular engineering in the field of Agri-bio-technology and Health.
Therefore, we decided to extend our research endeavours in the field of Personalised Nutrition by developing PROTEIN.
What do you and the team aim to achieve through the project?
The main aim of the PROTEIN Project is to design and develop a computer platform that will engage people to adopt a healthy, pleasurable, nutritional and sustainable diet by providing nutrition and physical exercise advice and support within the context of their everyday life. Advice and support will be provided mainly in the form of weekly dietary and physical activity plans. Other goodies will include personalized notifications, recipes, shopping tips, fun and educational mini-games. These will be adapted to the specific needs of the user and will be driven by the users’ personal preferences, physical and physiological characteristics, and health status. Thus, one could speak of a PeRsOnalized nutriTion for hEalthy living (PROTEIN) system.
The PROTEIN computer platform will be accessible to consumers primarily through a mobile smartphone application; online access (i.e via a webpage) may also be considered at a later stage. The system will focus on three use cases: personalized nutrition in the physical or electronic food store, in restaurant settings, whether physical or via online delivery), and at home. While the system will be geared to the general public, significant effort will be undertaken to create versions of the platform adjusted to be helpful for specific user groups: older adults, people attending to weight issues or living with obesity, people wanting to improve diet quality and who have nutritional deficiencies, adults with type II diabetes, patients with cardiovascular disease, and athletes
What can you tell us about this interesting and diverse multidisciplinary consortium?
The PROTEIN consortium is composed of 20 partners from 11 European countries: the United Kingdom, Ireland, Spain, Portugal, The Netherlands, Belgium Germany, Luxembourg, Italy, Serbia, and Greece. All partners have high innovation capabilities related to the objectives of the PROTEIN project. Partners groups include: 4 industrial partners (INTRA, ARTICA, OCADO, SLB), which will contribute significant industrial ‘know-how,’ covering the food retail, sports, and ICT industries; the elicitation of requirements, the design of the overall system architecture and the construction of robust exploitation plans and business models. We also have nine research partners, including 2 research centres: CERTH, and BIOS; 7 universities: AUTH, VUB, UoS, CHARITE, ATEITH, KUL and FMH-UL, which will contribute the required state-of-the-art knowledge and the technological expertise in the areas of nutrition & dietetics, biology & genetics, health & medicine, IoT and IoHT sensors, behavioural analysis, and law & research ethics. Our five SME partners will either introduce novel tools, sensors and services providing support to the overall project deployment and integration (DW, PLUX, VGYM) or will validate and enrich project solutions towards realizing efficient personalised nutrition as well as facilitating the early introduction of developed project solutions in relevant markets (FLUV, AFC), and 2 organisations (EASO, PE), which will play a vital role in engaging civil society and user groups, evaluating the system against identified needs, and augmenting project impact through their strong existing communication channels.
Tell us about the most innovative aspects of the project.
PROTEIN proposes a radically novel AI-based ecosystem that collects heterogeneous data related to consumer’s choices and food intake in various real-life environments (through their interaction with mobile devices and novel sensing technologies) and provides personalized nutrition and physical activity support, as well as early-warnings related to eating deficiencies, suboptimal eating patterns and non-communicable diseases. By taking advantage of the recent advances in ICT technologies as well as in direct-to-consumer genetic testing, blood and gut microbiome analysis, PROTEIN collects big data related to consumer’s choices/food intake from various real-life environments in order to discover new knowledge related to customers’ dietary behaviours. This new knowledge coupled with existing experts’ knowledge and customers’ profiles is used for the continuous improvement of the personalization capabilities of the system. In other words, PROTEIN is a completely dynamic ecosystem providing personalized or tailored dietary feedback advice, and therefore has the potential to help individuals to make positive sustainable changes to their dietary behaviours.
In your view, Lazaros, what are going to be the most exciting challenges?
PROTEIN aims to use cutting-edge technologies, such as artificial intelligence, machine learning, big-data analytics, wearable biosensors, smart mobile devices, Cloud computing, direct-to-consumer genetic testing, blood and gut microbiome analysis, gamification and games to promote healthy dietary behaviours. To this end, the most exciting challenges of the project are:
- development of novel AI algorithms for dietary plan personalization and adaptation
- modelling experts’ knowledge by the multidisciplinary project team of nutritionists, clinical nutritionists, biologists, medical experts, physical activity experts and engineers
- the development of machine learning algorithms for the analysis of heterogeneous data, including genetic data
- development of novel food intake assessment algorithms
- the analysis of eating behaviour during meals for early behavioural training
- the use of novel sensing technologies for bowel sound and volatile organic compound analysis (pre- and post- meal analysis)
- the development of the PROTEIN Game Suite for dietary habit adaptation and promotion of sustainable dietary choices and
- validation of the PROTEIN system in a variety of real operational environments with a large number of healthy individuals and different groups of patients.
Impressive! What can we expect to see in the first 6 months, and the first year of the project?
PROTEIN is a 3.5-year research project with a major milestone at month 18 (1.5 year), when the first prototype of the PROTEIN computer platform will be ready. During the first year of the project, though you won’t see impressive outcomes, this may nonetheless be the most crucial period of the project. A lot of hard work will be conducted to lay the foundations of the system to be developed: user scenarios and requirements will be finalised, followed by development of the system architecture and specification, a framework for user data/privacy protection, both ethics and societal acceptance will be developed (TARESS framework) which will be used during the entire project, and, finally, various internal project plans and tools will be developed, covering project aspects such as management, internal communications, advisory boards, system integration and evaluation.
From a user perspective, a concrete outcome of the project during the first months will be the project website which will launch at the end of February 2019 with material and information related to the project. Stay tuned!