Improving computation capabilities of robots with FPGAs

Robots usually process information from several sensors and have to produce results as fast as possible but this is not always possible due to the capabilities of Central Processing Units (CPUs) that work sequentially. Therefore, we use Field Programmable Gate Arrays (FPGAs) to implement processing algorithms as their internal parallel architecture allows to perform multiple computations concurrently.
We show different robotic applications and two use cases with FPGAs. The first one focuses on a Non-linear Model Predictive Controller (NMPC) for drone. The second one is an object avoidance application on a skid-steer mobile robot.