Applying software variability methods to design adaptive assistance robots

Abstract

We believe that assistance and service robots are a major technological trend that could lead to a spread similar to smartphones today. Thus, it could be normal to have such a robot at home to support people during their daily life. The main challenge with today's developments in this segment is a lack of features and a lack of adaptivity. Most service robots are specialised systems (vacuum cleaners) or do not consider the user's needs and preferences when fulfilling their tasks. Our goal is to cope with these challenges and propose a solution by applying principles of software variability engineering to create an adaptive and accessible robot that can be extended and re-configured to fit the user's needs better. Therefore, we modified feature and variability modelling known from product lines and software ecosystems and developed an approach to configure the robot based on user requirements (needs and preferences).