Wireless sensor network-based navigation for human-aware guidance robot

While most mobile robots possess the ability to navigate to and from given locations, for a robot to truly perform guidance, human awareness is also needed. Our guidance robot system possesses both, using a wireless sensor network for navigation and a vision-based tracking system for human awareness. We use the network to create a virtual directional field for the robot. As the network is ad-hoc and scalable, it is ideal for quick installation in existing environments. The vision system is used to estimate the relative direction and distance from the robot to the following human. Using these values, the system ensures that the human does not wander away from the robot during guidance. We setup a simulated hallway environment in order to perform an experiment to test the efficacy of our system in a real world application. The robot was able to successfully navigate to the goal and chase the human to reestablish guidance if the human wandered. The two requirements for performing true guidance.