Researchers at the Georgia Institute of Technology have developed a new algorithm that can control autonomous vehicles as it maneuvers at the edge of its handling capacity.

This new method is intended to enhance the safety of driverless cars even in hazardous driving conditions.

The project was sponsored by US Army Research Office.

"By merging statistical physics with control theory, and utilising leading-edge computation, we can create a new perspective, a new framework, for control of autonomous systems."

This new technology was jointly assessed by researchers of Georgia Tech’s Daniel Guggenheim School of Aerospace Engineering (AE) and the School of Interactive Computing (IC).

The fully autonomous car underwent various tests, such as sliding, jumping one-fifth-scale, as well as racing at 90 mph.

Researchers used advanced algorithms and onboard computing, along with installed sensing devices, to boost vehicular stability and performance.

Georgia Institute of Technology AE professor Panagiotis Tsiotras said: "An autonomous vehicle should be able to handle any condition, not just drive on the highway under normal conditions.

"One of our principal goals is to infuse some of the expert techniques of human drivers into the brains of these autonomous vehicles."

Georgia Institute of Technology AE assistant professor Evangelos Theodorou said: "Aggressive driving in a robotic vehicle, maneuvering at the edge, is a unique control problem involving a highly complex system.

"However, by merging statistical physics with control theory, and utilising leading-edge computation, we can create a new perspective, a new framework, for control of autonomous systems."

The researchers made use of a stochastic trajectory-optimisation capability, based on a path-integral approach, to create their MPPI control algorithm.

By using statistical methods, the team integrated large amounts of handling-related information, along with data on the dynamics of the vehicular system, to compute the most stable trajectories from several possibilities.

As the vehicle carries high-power graphics processing unit (GPU), 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 every moment.

This MPPI approach merges both the planning and execution of optimised handling decisions into a single phase.