Frontiers in Robotics and AI (Aug 2022)
Scalable and heterogenous mobile robot fleet-based task automation in crowded hospital environments—a field test
Abstract
In hospitals, trained medical staff are often, in addition to performing complex procedures, spending valuable time on secondary tasks such as transporting samples and medical equipment; or even guiding patients and visitors around the premises. If these non-medical tasks were automated by deploying mobile service robots, more time can be focused on treating patients or allowing well-deserved rest for the potentially overworked healthcare professionals. Automating such tasks requires a human-aware robotic mobility system that can among other things navigate the hallways of the hospital; predictively avoid collisions with humans and other dynamic obstacles; coordinate task distribution and area coverage within a fleet of robots and other IoT devices; and interact with the staff, patients and visitors in an intuitive way. This work presents the results, lessons-learned and the source code of deploying a heterogeneous mobile robot fleet at the Tartu University Hospital, performing object transportation tasks in areas of intense crowd movement and narrow hallways. The primary use-case is defined as transporting time-critical samples from an intensive care unit to the hospital lab. Our work builds upon Robotics Middleware Framework (RMF), an open source, actively growing and highly capable fleet management platform which is yet to reach full maturity. Thus this paper demonstrates and validates the real-world deployment of RMF in an hospital setting and describes the integration efforts.
Keywords
