The European PAL (Personal Assistant for a healthy Lifestyle) project (ref. H2020-PHC-643783) shall develop the behavior of a robot and its virtual counterpart to act as a pal for children with diabetes type 1, building on a previous European project (ALIZ-e, aliz-e.org) in which TNO was involved.
Diabetes has a major impact on the lives of children with diabetes and their families. Children with diabetes have problems keeping their blood sugar levels stable. Not only food, but metabolism, exercise, hormones, and emotions as well affect blood sugar levels. It is therefore very hard to maintain good blood sugar levels during puberty. There are growth spurts that alter the hormones and the first infatuations arise. In addition the consequences of stress have a bigger impact on children with diabetes, it can cause their blood sugar level to rise or fall. During a test for example the concentration can diminish and this in itself can cause the stress to increase.
For parents, this means anxiety and difficulty to hand over responsibility for the disease to their child because incorrect dealing with the disease can have serious consequences. Direct consequences are a hypo or hyper (too low or too high blood sugar) and in the long term, for instance, damage to organs.
In the PAL project we teach the child to deal with their diabetes independently, in a good manner. At the same time we teach parents to trust their child. We develop a pal who can support the child in the form of a robot or its virtual counterpart on a screen. With this pal the child will do various activities, such as keeping a diary and play a quiz, to achieve certain goals. Professional caregivers can set these goals together with the child and keep track of how the health of the child develops. They can also change the content of the activities to suit the child. Parents can view information about the health and goals.
The figure below gives an idea of our plans. The robot will be used in the hospital and during camps, where children can do 1-on-1 and group activities with the robot. In other situations like at home and at school, they can do some of the same and other activities with a virtual character. During the 4-year project there will be three design and evaluation cycles. Every cycle the target group will get broader, the duration longer and the effect measurements increasingly focused on behavioral change.
The necessary knowledge on children (with diabetes), dialog systems, emotion recognition in text, behavior interventions, adapting the interaction based on knowledge of the user and his/her behavior, design of interfaces and exploitation of the results will be supplied by the different partners (MeanderMC Amersfoort, ZGV Ede, DVN, Ospedale San Raffaele Milaan, Sostegno70, TNO, FCSR, DFKI, TU Delft, Mixel, Produxi en Imperial College London).
If you would like to know more about the project please contact Rosemarijn Looije.