Lola Gayle, STEAM Register
Self-driving robotic vehicles are made to cruise at safe speeds, and in near-perfect driving conditions. But what happens when road conditions are less than perfect, or worse yet, downright hazardous? Sure, you can cover an autonomous vehicle in sticky fly paper to catch pedestrians, but that still won’t make it any safer in certain conditions.
And what about autonomous rovers, such as those currently exploring Mars and the Moon? You certainly can’t judge the exact terrain or conditions they might be facing from moment to moment. You can only plan for possibilities.
To address these issues, Georgia Institute of Technology researchers recently visited the Georgia Tech Autonomous Racing Facility to test a new approach for keeping these vehicles under control as they maneuver at the edge of their handling limits.
The new approach uses advanced algorithms and onboard computing, in concert with installed sensing devices, to increase vehicular stability while maintaining performance, reports Rick Robinson for Georgia Tech.
See Also: This Robotic Hand Can Grasp Whatever It Wants To, All By Itself
The researchers, from Georgia Tech’s Daniel Guggenheim School of Aerospace Engineering (AE) and the School of Interactive Computing (IC), put two one-fifth-scale, fully autonomous auto-rally cars through their paces by racing, sliding, and jumping at the equivalent of 90 mph.
These rolling robots were made to utilize special electric motors to achieve the right balance between weight and power. They are equipped with a motherboard and quad-core processor, a potent GPU, and a battery. They also carry with them two forward-facing cameras, an inertial measurement unit, and a GPS receiver, along with sophisticated wheel-speed sensors. Additionally, the power, navigation, and computation equipment is housed in a rugged aluminum enclosure able to withstand violent rollovers. Each vehicle weighs about 48 pounds and is about three feet long.
According to Panagiotis Tsiotras, an expert on the mathematics behind rally-car racing control, autonomous vehicles need to be able to handle any road conditions, good or bad. “One of our principal goals is to infuse some of the expert techniques of human drivers into the brains of these autonomous vehicles,” Tsiotras said.
Traditional techniques are pretty static, using the same approach regardless of road conditions. The Georgia Tech method – known as Model Predictive Path Integral control (MPPI) – was developed specifically to address the non-linear dynamics involved in controlling a vehicle near its friction limits.
“Aggressive driving in a robotic vehicle — maneuvering at the edge — is a unique control problem involving a highly complex system,” said Evangelos Theodorou, an AE assistant professor who leads the project. “However, by merging statistical physics with control theory, and utilizing leading-edge computation, we can create a new perspective, a new framework, for control of autonomous systems.”
See Also: Researchers Are Training Computers To Train Dogs, No Humans Needed
The researchers used a stochastic trajectory-optimization capability, based on a path-integral approach, to create their MPPI control algorithm, Theodorou explained. Using statistical methods, the team integrated large amounts of handling-related information, together with data on the dynamics of the vehicular system, to compute the most stable trajectories from myriad possibilities.
Georgia Tech researchers are using an electric-powered autonomous vehicle to help driverless vehicles maintain control at the edge of their handling limits. Shown (l-r) are Georgia Tech students Sarah Selim, Brian Goldfain, Paul Drews, Grady Williams. Georgia Tech/Rob Felt
Processed by the high-power graphics processing unit (GPU) that the vehicle carries, the MPPI control algorithm continuously samples data coming from global positioning system (GPS) hardware, inertial motion sensors, and other sensors. The onboard hardware-software system performs real-time analysis of a vast number of possible trajectories and relays optimal handling decisions to the vehicle moment by moment, Robinson reports.
Basically, the MPPI approach combines both the planning and execution of optimized handling decisions into a single highly efficient phase.
A paper covering this research was presented at the recent International Conference on Robotics and Automation (ICRA), held May 16-21.
Click here to read more about their research.
See Also: Self-Parking Chair – A New Level of Lazy or Totally Cool?
Follow us on Facebook page and Twitter