Francis Powlesland & Elena Kotljarova - 1 year ago
On the straight and narrow path - How to get cars to drive themselves using reinforcement learning and trajectory optimization
IBM provides several offerings helping the industry to make autonomous driving become reality and this "Race Track" demo is one perfect example of it. This demo mimics a real-world project done in collaboration with Thyssen Krupp Presta (TKP), a world leader for steering systems. Together we developed an approach were the vehicle learns over time to optimize itself towards a defined goal. It's very often the case in the tech world that we hear about self-driving cars. Just look at Tesla for instance. But, as the saying goes - "all roads lead to Rome"; there are self-driving technologies that approach the problem in novel and creative ways. Unlike current autonomous vehicles, we enable the car to learn "on-the-drive." In this talk we'll discuss some of the more in-depth concepts around reinforcement learning and how we apply this tothe concept of trajectory optimization. We will also give you the opportunity to see not just a demo, but a sneak peek into a real life application.