# Robotic Arm Control with a Reinforcement Learning Algorithm implemented with Dataflow Graphs _by Gendron Gaël (IETR-VAADER) - 2019.08.20_ ###### tags: `VAADER` `Seminar` ![](https://media.giphy.com/media/lj11uKKMhr4Os/giphy.gif) ## Abstract Robotic devices can be controlled using Machine Learning, and more specifically Reinforcement Learning (RL). This method is very effective for solving complex positionning problems but it requires a lot of time to compute. PiSDF is a model of computation based on Dataflow Graphs allowing more parallelism in applications, it could be used to schedule optimally computations in order to reduce this time. During my internship, I implemented RL on a robot arm with and without Dataflow Graphs and analysed the gain in learning time. ## [Slides](https://mycore.core-cloud.net/index.php/s/19qE7HdmT34RHIJ)