Summer research project at the University of North Texas
In summer 2011, Dr. Costilla Reyes developed a research project on autonomous robots using proximity and vision sensing at the University of North Texas (UNT) in Denton, Texas, as a summer undergraduate research intern. He took part while completing his bachelor’s degree at the Universidad Autónoma del Estado de México, around the time he began work as a radio-access-network integration engineer at Ericsson. He returned to UNT in 2012 for graduate study.
How the robot sensed and avoided objects
The project presented an approach for navigating a rover autonomously with three kinds of sensing. Stereo vision infers depth from the small differences between two camera views, much as two eyes do. A laser range finder measures distance to surfaces along a scan line. Infrared proximity sensors detect nearby obstacles. A three-dimensional terrain-modeling technique turned these inputs into a depth map of the ground ahead, and because the size and position of an object changes which sensor sees it best, the robot could use different sensors to steer around different obstacles.
The method was developed in MATLAB and tested on a Pekee II mobile robot, using the programming interfaces supplied by its manufacturer, Wany Robotics. The algorithm was optimized so that the robot’s movements depended on the type of object its sensors and camera detected. The Pekee II proved a good platform for navigation, and its interfaces were sufficient to build an effective algorithm.
Later robotics work
Dr. Costilla Reyes’s master’s research at UNT continued in robotics and perception for autonomous systems, including the Indoor SLAM project, in which a depth camera supplied a three-dimensional map of a building hallway. In 2013 he also gave a live demonstration of autonomous robots using vision and proximity sensors as a science and engineering exhibitor at the Perot Museum of Nature and Science in Dallas.