Once again, the cutting-edge world of artificial intelligence is transforming jobs traditionally handled by humans, and this time, the spotlight is on Toyota. The company has introduced an exciting innovation: two autonomous Supra sports cars capable of drifting in perfect harmony. This development showcases AI’s potential to manage complex tasks beyond just repetitive motions. Surprisingly, there is a safety angle to this technological feat, providing insights into what Toyota’s AI-driven systems might offer in the future.
The Toyota Research Institute (TRI), in collaboration with a team from Stanford Engineering, has spent seven years pioneering AI technology for automotive use. Initially, they succeeded in getting a single car to drift autonomously. Now, they’ve achieved a more complex task: two Supras drifting together without any human control, either inside or remotely. These tests took place at Thunderhill Raceway Park in Willows, California, using specially modified Toyota GR Supra cars designed to meet Formula Drift specifications.
Toyota’s team focused on creating stable control systems for the lead car, ensuring it could maintain its drift effectively.
Meanwhile, the Stanford team developed sophisticated AI models for the chase car, allowing it to follow the lead car closely without colliding. Both vehicles communicate via a dedicated WiFi network, exchanging real-time data on positioning and planned trajectories.
The vehicles rely on sensors to manage steering, throttle, and braking, conducting about 50 scans per second. They use a method called Nonlinear Model Predictive Control (NMPC), which translates driving objectives into mathematical constraints. The lead car aims to drift along a set path while respecting limits like maximum steering angle. The chase car’s task is to mimic the lead car’s movements, adjusting instantaneously.
This technology hints at a safer driving future for everyone. According to Stanford’s Chris Gerdes, “The track conditions can change dramatically over a few minutes when the sun goes down. The AI we developed for this project learns from every trip we have taken to the track to handle this variation. The physics of drifting are actually similar to what a car might experience on snow or ice. What we have learned from this autonomous drifting project has already led to new techniques for controlling automated vehicles safely on ice.”
Previously, autonomous drifting was limited to single vehicles. The groundbreaking aspect here is training the chase car to respond dynamically to the lead car’s behavior, mimicking real-world scenarios. In such situations, vehicle safety systems might need to intervene to prevent skidding into other cars, pedestrians, or obstacles. This approach goes beyond following simple routes or using basic braking systems. It involves real-time reaction to the environment using multiple vehicle controls.
While this innovation may seem like pure entertainment, it marks significant progress toward a future where AI plays a vital role in enhancing driving safety.
