I. Introduction
Typically, motor rehabilitation in medical centres has to face a socio-economic burden and it is limited by high costs and/or geographical barriers. Tele-rehabilitation (TR) has recently emerged as an effective tool for providing assisted living, increasing clinical outcomes, positively enhancing patients' Quality of Life (QoL) and fostering their reintegration into society, also pushing down clinical costs. Cloud computing in combination with Edge Computing, Internet of Things (IoT), Big Data storage [1] and analytics [2] and Artificial Intelligence (AI) [3], [1] are the main enablers for TR. Edge rehabilitation devices can act as smart digital biomarkers sending quantifiable physiological and behavioural patients' data to the Hospital Cloud. Furthermore, TR allows assessment, intervention, consultation, and education. However, the market of rehabilitation is currently dominated by expensive vendor lock-in stand-alone robotic rehabilitation devices aimed at clinical centres. This paper presents TR as a Service (TRaaS), i.e., a research project founded by the Italian “Projects of Relevant National Interest” (PRIN) programme. It aims at overcoming the gap in the field of TR by creating a reference intelligent Cloud/Edge framework architecture and a standard data model for the development of different kinds of new de-hospitalized TR services. The strategic relevance of these aspects is emphasized at the national level by the PNR (Programma Nazionale per la Ricerca 2021–2027) and by the PNRR-Health (Piano Nazionale Ripresa e Resilienza - Sanita) of the Italian Government and, at EU level, by the Horizon Europe Work Programme 4 - Health of the EC and by a large number of research/innovation projects (eg. https://dih-hero.eu/, http://adapt-project.com/, etc.) active on the topic. Particular attention will be given to the design, development and testing of the TRaaS framework architecture specifically considering AI and preserving patient privacy. TRaaS will embody the Cloud/Edge continuum principle. On the patient side, the TraaS architecture will include an Edge layer, acting either at the patient's home or in self-service clinical centres offering robot-assisted rehabilitation devices equipped with sensors and actuators connected with the hospital Cloud. Furthermore, Machine Learning (ML) techniques will also be studied to perform patients' clinical data analytics at both Cloud and Edge layers for identifying and/or improving the effectiveness of the patient's motor, cognitive and/or sensory rehabilitation outcomes. A pilot will be arranged in the IRCCS Centro Neurolesi “Bonino Pulejo”, Italy, i.e. a clinical and research centre specialized in rehabilitation. The project will aim also at assessing technological, scientific, economic and social impacts. In the end, it will aim at stimulating both the scientific and industrial healthcare communities toward the development of new innovative secure smart TR services.