The controlling of mobile robots has been and still is in the focus of researchers. Fuzzy rule-based controllers are extensively used to control robots. Path planners were created using different search methods finding the shortest trajectory between two points and thereby avoiding recorded obstacles.
The low degree of automation in hospitals promises huge potential for increasing logistical effects by the use of mobile robots. Especially the automated transport of a hospital bed, which manually requires two persons, would release time from transportation activities. The usage of a mobile robot joined to a hospital bed makes controlling considerably harder.
This document describes a project accomplished by me in cooperation with RobCab AB. During the project, a path planner was created and controlling strategies to move a mobile robot joined to a hospital bed between two locations was implemented. The controlling strategies are not only based on sensor readings from a laser mounted on the mobile robot and a list of way points provided by a path planner but incorporates also the position of the bed in the environment.
The robot adapts its behavior to ensure a safe movement considering the position of the bed which is not only determined by the kinematics of the robot-bed model but also by an external force in the form of a person following the transport (human intelligence). Development was done on the Stage simulator and is intended to be finally transferred to a real robot platform. Tests done in the simulator showed that the implemented methods are applicable to bring the bed to the desired goal taking into account the reaction of the person following the transport.
Source: Örebro University
Author: Erdis, Sahit