|Thursday, March 28, 2013|
Students Design and Test Movement Algorithms for Mobile Robots
Photos by Chuck Thomas
Modeling and simulation engineering students recently took a hands-on approach to one of their courses at Old Dominion University.
Under the supervision of Yiannis Papelis, course instructor and research professor at ODU's Virginia Modeling, Analysis and Simulation Center, students in the senior-level Design and Modeling of Autonomous Robotic Systems course were charged with developing a more sophisticated movement algorithm for a mobile robot - and then testing that program by having the robot use sensors to follow a particular route.
During an assessment session on Thursday, March 21, the 10 students in the class spread out around the floor of the E.V. Williams Engineering and Computational Sciences Building lecture hall. They tested their wheeled robots on the course, which was laid out on the floor using high-contrast tape.
Studying the movements closely, sometimes even lying on the floor next to the robot, the students tested the algorithm they developed to see if it worked at the prototype level (see photo below).
Andy Ren, graduate assistant for the Modeling of Autonomous Robotic Systems course, said in addition to evaluation, the hands-on exercise allowed students in the class to test their own algorithm, to see if it sent accurate commands to the mobile robot to follow the simple movement directions over the course.
"It's a very good exercise for these students," Ren said at the start of the session. "We've done the laboratory work, now these students will each show how well these robots follow the lines on the floor."
The mobile robots needed to perform three tasks: follow the loop of tape laid out on the floor; detect intersections and dead ends in the tape course; and search the road network until finding an object located at one of the dead ends. As an added challenge, students received extra credit if they demonstrated that their robot could learn the tape course and traverse it faster the second time around.
The assignment underlined the importance of good software engineering practices, because the students were given complete freedom on how to develop their own algorithm architecture and level of sophistication. It also demonstrated the value of a good model, as students could use a virtual environment for the majority of their time-consuming testing, only making final calibration changes to the physical robot.
ODU is the only school in the country to offer a complete advanced education in modeling, simulation and visualization engineering (MSVE). It introduced the first undergraduate program in MSVE in the United States in 2010. The first class of four undergraduate M&S engineers will receive their bachelor's degrees this spring.