Engineering and Architecture
Domenec Puig Valls
Computer Science and Mathematics of Security
Automatic difficulty level generator for physical rehabilitation exercises by entertainment
Application of computer science in rehabilitation by entertainment has been the subject of many studies including analysis of a specific physical ability along with recommendations on its improvement. But none has so far delivered an intelligent levelling strategy designer by which the individuals can reach the goals step by step with a sustainable enthusiasm nor did they studied how the resulting data could help the subjects to improve their social inclusion. The specific target group of this project is kids suffering from cerebral palsy and the practical applications of such a platform would cover a wide range of issues, from therapeutic solutions to mere physical and social entertainment which not only helps subjects but also their parents and caregivers.
The modular core of the platform must be founded so strong that it would be open to inclusion and exclusion of new input devices such as cameras or devices to record physiological signals. Another novelty, is the inclusion of compiled data derived from humanoid robots that, in special cases, interact better with kids. So performing physical activities at the presence of such robots brings relevant data through embedded sensors of the robot, specifically as reactions to replanted emotional and physical expression of the robot.
The recommender system should also take a user-collaborative approach along with subject’s profile, to make better recommendations from the driven data of users across the globe which must be stored in a database accessible over the web. Hence a web platform is needed by which the acquired data could be stored anonymously and securely which are greatly important due to the focus of the project on disabled kids. The web platform opens the gate to improvements in social inclusion, which entails implementation of concepts such as cosine similarity and social networking factors such as Betweenness Centrality, Eigenvector Centrality, Directed Graphs and PageRank.
Additionally, assessment of emotional status of the subjects provides the platform an asset to better determine the appropriate levels difficulties. Moreover, a pre-analysis module is needed to visualize the results of analytics such as improvements made over the course of time in the targeted physical ability, emotional state and social inclusion. In parallel, a predictive analysis module to predict the time the subject reaches his/her next level according to the intensity by which s/he follows his goals, the data derived from his profile or in comparison to other users is needed. Predicting the index of social inclusion and emotional state over the course of time in future is also a necessity.
Developing a recommender system for the levelling strategy with aforementioned specifications.
Developing a web platform to both record user data and deliver analytics.
Predicting physical, emotional and social improvement of the individuals.
Assessing emotional state and social inclusion of the individuals.
37.5 hours a week
|This project has received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No. 713679|