AI for supporting people with disabilities

• 25 April 2024
• 2 minutes

The REXASI-PRO (REliable & eXplAinable Swarm Intelligence for People with Reduced mObility) project aims to use safe and reliable artificial intelligence to develop drone-assisted, self-driving wheelchairs to support people with motor and sensory disabilities.

Thanks to progress in artificial intelligence and robotics, innovative solutions for assisting people with reduced mobility are emerging. However, one of the remaining challenges is to propose more technological, safe and reliable products and devices within a manufacturing framework that is more and more concerned with environmental sustainability and the safety of workers and end users.

The REXASI-PRO project, in which the Trustworthy Autonomous Systems (TAS) research group of the Dalle Molle Institute for Artificial Intelligence (IDSIA USI-SUPSI) is participating, aims to develop a new framework based on the paradigms of safe and responsible artificial intelligence (Trustworthy Artificial Intelligence) and collective, self-organizing intelligence (Swarm Intelligence) that will enable collaboration between heterogeneous robots to support seamless door-to-door navigation for people with reduced mobility. In other words, autonomous wheelchairs and drones will be able to coordinate to avoid collisions, obstacles, and other hazards much more effectively than individual action.

Recently, the research team simulated some decision support scenarios for safe crossing of roads in the presence of other vehicles. The experiment took place in the IDSIA USI-SUPSI Robotics Laboratory and allowed for the evaluation of the safety benefits of multi-agent integration of redundant and heterogeneous sensors, including those based on computer vision, via information fusion. Information from the different detectors was then combined through appropriate algorithms to achieve a much higher level of accuracy and decision reliability than allowed by individual detectors.

"The trial described is part of the broader research project, whose results will enable greater autonomy and safety for people with motor and sensory disabilities, constituting an important element of social inclusion and sustainability in future smart-cities," comments Prof. Francesco Flammini, head of the research group.

Il progetto, coordinato da Spindox Labs con la partecipazione del Consiglio Nazionale delle Ricerche (CNR), il centro di ricerca tedesco sull'intelligenza artificiale (DFKI), il King's College di Londra e l'Università di Siviglia, è stato selezionato dalla comunità europea nell'ambito del programma Horizon Europe ed è finanziato nazionalmente dalla Segreteria di Stato per la formazione, la ricerca e l'innovazione (SEFRI).

The project, coordinated by Spindox Labs with the participation of the National Research Council (NRC), the German Artificial Intelligence Research Center (DFKI), King's College London and the University of Seville, was selected by the European community under the Horizon Europe program and is nationally funded by the State Secretariat for Education, Research and Innovation (SERI).

For further information:

• Project website

• Scientific article Safe Road-Crossing by Autonomous Wheelchairs: a Novel Dataset and its Experimental Evaluation