Roboception offers innovative 3D vision hardware and software solutions that enable any robotic system to reliably perceive its environment in real-time. Robot vision is the key to a more flexible use of robotic systems: If a robot is able to perceive its environment and workspace reliably, its tasks, workpieces or requirements can be changed and adapted rapidly.
The rc_reason suite of software modules is an advanced AppliedAI tool designed for robotic applications. As described in a previous post, the rc_reason CADMatch module offers the possibility to import a template generated from the object‘s CAD data into the system, in order to efficiently handle different objects autonomously. One of the major challenges in flexible production is to maintain reliability and robustness in different settings. Every collision in a productive cell leads to an interruption of production processes and requires staff to troubleshoot. Therefore, it is of great interest for a stable production environment to keep the error rate caused by collisions as low as possible. This can be done with the help of intelligent robot vision that checks for potential collisions during approach when calculating grasp points.
Roboception’s original collision check feature in the rc_reason suite avoided collisions of the gripper with the bin and detected objects. Recently, a new feature has been added, and the software now also checks for collisions between the gripper and the point cloud. CADMatch will thus only provide grasps on objects that are not in collision with the point cloud, resulting in more reachable grasps even without using an external motion planner. Objects that have been detected below others are automatically excluded from the list of accessible grasps when the object above has not been detected.
The point cloud mesh used for collision checking can also be visualized in CADMatch’s 3D result visualization. As for all collision objects, collisions with the watertight version of the point cloud are checked at the pre-grasp pose, at the grasp pose, and at all points on the linear motion between the two poses. This allows for advanced motion planning to make the application more stable and reliable.
In the H2020 ODIN project, the new Collision Check feature will be applied and further developed to make applications more robust. This will further improve the production process and provide the highest possible accuracy in the innovative use challenges that are being solved in the project.