The Kineo team’s latest demo of reactive path-planning generated a lot of interest at the recent International Robot Exhibition (iREX) in Tokyo (see part 1 of this blog series for more information about the event).
Kineo has been researching the application of its KineoWorks component for real-time obstacle avoidance in robotics. This would have big implications for improving safety in human-robot collaboration.
KineoWorks computes collision-free motion for industrial robots that are used in automated manufacturing and measuring processes. Robot motion is often computed offline for the purpose of simulation and programming. However, KineoWorks perfomance is capable of computing robot motion in real-time, where the robot avoids collisions by reacting to live changes in the scanned operating environment.
For our demo, we used an optical scanner to continuously update a point cloud model of a workspace in our booth. This enabled us to display a point cloud that dynamically updated in response to the movements of a visitor at the workspace. At the same time, we simulated the operation of an industrial robot in the workspace. The visitor could interact in real-time with the virtual robot, which would adopt new trajectories to avoid collisions with our visitor – see the video below.