Service robots are being built with more specialized functions to support humans with very specific tasks. Researchers have been working on algorithms for a robot that can help people who train for marathons.The robot would be able to carry belongings and water for the runner and also provide a map with a running path. The robot has to be able to move at a human running speed, while avoiding obstacles in an unpredictable outdoor setting.
A marathon service robot would need to be able to detect the running human using a laser range finder. Since most robots are usually programmed to operate indoors at a slower speed, there are challenges to teaching the robot how to avoid other moving obstacles while still tracking the runner outside at a faster pace.
Researchers have created a human detection algorithm and an obstacle avoidance algorithm to help sidestep any potential snags in functionality. For the robot to sense the human’s location, it collects data from the laser range finder, which is placed on the robot at torso height and scans the environment. To avoid obstacles, a weighted radius was developed to measure the relative velocity between the runner and obstacles. Through experiments, the researchers were able to test the algorithms and show that they could potentially be used in the creation of a marathoner service robot. Training for a marathon using a robot may just help improve an athlete’s results to help to reach a personal record.
Written by IEEE on September 11, 2017
For More Information
This paper presents a human detection algorithm and an obstacle avoidance algorithm for a marathoner service robot (MSR) that provides [...]