What is the OCARIoT project?
Long term behavioural disturbances and interventions in eating and physical activity (and as consequence an energy imbalance between calories consumed and expended) are the primary cause of obesity; therefore, obesity is largely preventable by making healthier eating choices and exercising regularly1. Childhood is an important period for forming healthy behaviours in order to reduce obesity inequities. This also implies the need to involve a wide range of stakeholders, from family counselling, behaviour modification, physical activity training and nutrition and dietetics and, where necessary, to medication [Aiello et al. 2015]. Proving that point, training of children up to 12 years of age to eat and move in a non-pathological way has been demonstrated effective in obesity and changing behaviours in randomized control trials [Christos et al. 2014] [Sola et al. 2012]. Behavioural treatment is an approach used to help individuals develop a set of skills to achieve a healthier weight. It is more than helping people to decide what to change; it is helping them identify how to change [Foster et al. 2005].
Having proven the benefits of such interventions in patients under supervised conditions, the logical next step is to apply the same kind of methodologies to the general population by combining our expertise in identifying and modifying behavioural patterns with current technical advancements in the fields of sensors, information analysis and personalised guidance systems. We therefore aim to develop a sophisticated, non-invasive, unobtrusive, personalised, IoT system that can detect and normalise the behaviours that put an individual at risk of developing obesity or eating disorders. So it is important to start educating children in order to promote healthy nutritional and lifestyle behaviours in the future. Schools provide opportunities to ensure that children understand the importance of good nutrition and physical activity, and how they can benefit from both. Teachers and health professionals are often involved as providers of health and nutrition education [Bemelmans et al. 2011]. ICT have been acknowledged as a significant enabler to accommodate the provision of Active and Healthy services for obesity and other conditions. Especially IoT-based solutions aspire to provide a strategic component in supporting the creation of an ecosystem able to dynamically answer and prevent the challenges faced by the health and social care systems. The IoT combined with medical experience has a significant potential to correct eating and physical activity behaviours in an earlier development stage and prevent the obesity evolution while decreasing the overall healthcare costs because enables [Sola et al. 2012]: (i) Real-time transmission of time sensitive data allows personalising the medical care. (ii) Empowerment and guidance to the young individuals to help them modify their eating and activity behaviour towards personalised pre-set goals. (iii) The intervention/treatment efficacy increases due to close monitoring and personalisation. (iv) Cost of medical and healthcare infrastructure are reduced due to unnecessary hospitalisation and decreasing demand on medical staff. The “always-connected” paradigm is becoming a way of life, and this could result in a positive transformation for health and social care systems, which are strangling to find new ways for preventive care and keep people active and independent for longer. As a result, various IoT attempts in Europe are being deployed for sensing, measuring, controlling smart connected objects, while various IoT platforms (e.g. universAAL), and solutions are emerging in the market aiming to support users. The constantly increasing availability of affordable, “smart”, personal devices (e.g. smartphones, wearable “quantified-self” artefacts, etc.), as well as other connected devices that surround us, situated from the home environment, to other outdoor smart environments (e.g. smart cities) provides unique opportunities to obtain input that can be exploitable to understand the individual user needs and, thus build personalized models, but also to facilitate intuitive and context-aware communication with medical experts an families. Nevertheless, while various promising services are being developed on top of (open) IoT platforms, these offer “snapshots” of support (e.g. diet guides, physical exercise guidance apps, etc.), which lack continuity across the entire spectrum of everyday life of the user, while also illustrating a great deal of heterogeneity concerning the way that they reach the end-user, requiring the user to adapt to their functionality, rather than the services to the user’s needs. For example, though there are many applications and tools, which try to promote healthy habits in children (e.g. serious games, exergames, etc.), they are not customizable according to the children’ needs and, the proposed obesity treatment is rarely fully successful in the longer term.
Considering all of this, OCARIoT will be based on three key steps:
- Finding information: It is important to collect all relevant information not only related to the children health condition, but also related to context environment conditions, clinical and medical knowledge and Open Data, which can impact on the well-being and care process of the child.
Challenge 1: Collect all relevant health, physiological and context information from different sources creating an IoT network in an easy and transparent way and then, enrich it semantically to be processed by intelligent algorithms. OCARIoT will measure physiological and behaviour data (through wearables such as activity wristband or smartwatches; and sensors deployed in the environment) in real life conditions by the adoption of sensors to be used and integrated in the daily living IoT environments of the involved beneficiaries. All done in a non-obtrusive and transparent way in order to provide a personalised coaching to children.
- Making decisions: On the basis of the pre-processed data, a personalised guidance and follow-up plan (known as coaching plan) must be defined for each child. That coaching plan will set realistic well-being and treatment goals by taking into account the user’s needs and preferences. At this step, the usage of validated intelligent algorithms can be very helpful: combined with the collected data and processed in real-time they can support the prediction of the obesity and well-being evolution of children.
Challenge 2: Context-aware intelligent algorithms to understand the information, make the decision, and dynamically personalise and adapt the obesity coaching to the child. OCARIoT will provide a pipeline of algorithms starting from raw measurements processing, feature extraction, indicators quantification and finally algorithms for risk assessment of obesity based on eating and activity behaviour patterns and personal profile data. OCARIoT will integrate autonomous learning capabilities to optimize future decisions.
- Taking action: This step is focused on the child and how he/she follows the recommended obesity coaching plan while improving his/her health-related behaviours (such as exercise or weight control).
Challenge 3: Empowerment and guidance to the children and young individuals to help them modify their eating and physical activity behaviour towards personalised present goals (habits determine lifelong preferences and health behaviours). For example, Participation in 150 minutes of moderate-intensive aerobic physical activity each week (or equivalent) is estimated to reduce the risk of ischaemic heart disease by approximately 30%, the risk of diabetes by 27%, and the risk of breast and colon cancer by 21–25%2.Thus, a lifestyle approach that encourages long-term behavioural changes is needed to tackle childhood overweight and obesity. For doing so, the project will provide a set of challenge games, game templates and apps for educating children about obesity, overweight and promoting healthy lifestyle behaviours (nutrition, physical activity and healthy habits) and reinforcing social inclusion of children with obesity and/or overweight. Besides, usability and accessibility will be also essential for guiding the interaction design.