Click it and Unblock the Notifications
Autonomous Tandem Drift By Toyota And Stanford Sets New Safety Standards
Today, Toyota Research Institute (TRI) and Stanford Engineering revealed a groundbreaking achievement in driving research: autonomously drifting two cars in tandem. This collaboration, spanning nearly seven years, aims to enhance driving safety. The experiments focus on automating the motorsports maneuver known as "drifting," where a driver controls a vehicle’s direction after losing traction by spinning the rear tires—a skill useful for recovering from slides on snow or ice.
By introducing a second car drifting in tandem, the teams have better simulated dynamic conditions where vehicles must quickly respond to other cars, pedestrians, and cyclists. Avinash Balachandran, vice president of TRI’s Human Interactive Driving division, stated, "Our researchers came together with one goal in mind – how to make driving safer." He added that using AI to drift two cars autonomously is the most complex motorsports maneuver and has significant implications for future automobile safety systems.
Chris Gerdes, professor of mechanical engineering and co-director of the Center for Automotive Research at Stanford (CARS), explained that the physics of drifting are 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," he said.
The experiments took place at Thunderhill Raceway Park in Willows, California, using two modified GR Supras. Algorithms for the lead car were developed at TRI, while Stanford engineers created those for the chase car. TRI focused on developing robust control mechanisms for the lead car to ensure safe and repeatable runs. Meanwhile, Stanford Engineering developed AI models and algorithms enabling the chase car to adapt dynamically to the lead car's motion without colliding.
Both vehicles were modified by GReddy and Toyota Racing Development (TRD) with changes to their suspension, engine, transmission, and safety systems like roll cages and fire suppression. Although slightly different from each other, they were built to Formula Drift competition specifications to collect data with expert drivers in a controlled environment. Equipped with computers and sensors, these cars can control their steering, throttle, and brakes while sensing their motion such as position and velocity.
AI Integration
The vehicles share a dedicated WiFi network allowing real-time communication about their relative positions and planned trajectories. To achieve autonomous tandem drifting, they continually plan their steering, throttle, brake commands, and intended trajectory using Nonlinear Model Predictive Control (NMPC). Each vehicle starts with objectives represented mathematically as rules or constraints it must follow.
The lead vehicle aims to sustain a drift along a desired path within physical laws and hardware limits like maximum steering angle. The chase vehicle's goal is to drift alongside the lead vehicle while avoiding collisions proactively. Each vehicle solves an optimization problem up to 50 times per second to decide the best commands while responding to changing conditions.
Implications for Driver Safety
This technology holds promise for assisting drivers in reacting correctly during sudden dynamic situations that often lead to accidents. Car crashes cause over 40,000 fatalities annually in the US and about 1.35 million worldwide. Many incidents result from loss of vehicle control in unexpected scenarios.
Balachandran emphasized that when a car begins to skid or slide, drivers rely solely on their skills to avoid collisions with other vehicles or obstacles. "An average driver struggles to manage these extreme circumstances," he noted. This new technology can intervene precisely when needed to safeguard drivers by managing loss of control like an expert drifter would.
Gerdes added that achieving what has never been done before demonstrates potential advancements in making cars safer: "If we can do this, just imagine what we can do to make cars safer."
The AI developed for this project learns from every trip taken on the track handling variations such as changing track conditions when the sun goes down. This continuous learning process enhances its ability to manage different scenarios effectively.
Toyota Research Institute (TRI) focuses on amplifying human ability through research aimed at making lives safer and more sustainable. Led by Dr Gill Pratt since its establishment in 2016 TRI operates offices in Los Altos California Cambridge Massachusetts For more information about TRI visit http://tri.global
Continue with Google