The objective of the research project was to create a vision-based system to monitor a robot’s movements in real-time. The system provides a mechanism to monitor the control system’s function and check that there are no errors in the system. Video feed captured by a camera (Microsoft Kinect sensor) is compared to how the robot is supposed to function at a particular instant of time to determine any discrepancies using image comparison. These discrepancies are detected using threshold values. For example, if a measured relationship value is below the threshold, one can determine that the robot is not in the position it is supposed to be in. In addition to this, a message can be sent to the system manager or the robot system can be shutdown to prevent any damage to its surroundings. Read More

• Engineering