Reconfigurable Constraint-Based Reactive Framework for Assistive Robotics With Adaptable Levels of Autonomy

Abstract

In this work, we propose a constraint-based reactive framework that allows an easy customization of assistive robotic systems to comply with the particular needs of users with different levels of mobility. This framework enables the specification of modular assistive reactive behaviors which can be easily composed to allocate autonomy over robot-related or task-related variables. Such assistive behaviors can be systematically modulated online, which results in adaptable levels of autonomy. Modulation can be performed directly by the user through suitable human-machine interfaces, or can be performed automatically by the computer. Although our framework also shows potential for handling other types of tasks, we focus on assisting the user with reaching tasks. We experimentally validated the flexibility and applicability of our framework by presenting three assistive use cases, each targeting users with different level of mobility. Each of those cases was easily reconfigured to include various HMIs (e.g. SpaceMouse, eye tracker, etc), exploiting different assistive strategies. During the experiments, the level of autonomy could be automatically or manually adapted.