At Bar-Ilan, robots box, play soccer – and maybe soon, drive cars

For over a decade, robots developed by Bar-Ilan University students, led by robot tech pioneer Prof. Eli Kolberg, have represented Israel in contests around the world – playing soccer, boxing, and performing other autonomous activities. Now, that technology will be used to make drivers safer with robot-like capabilities in cars warning drivers of dangers ahead.

And eventually, said Kolberg, that technology will be used to make self-driving cars that behave better on the road.

Similar to the MobilEye safety system, which uses cameras and sensors to warn drivers if they are getting too close to vehicles ahead of them, the Bar-Ilan system will use its robot tech – based on cameras, a computer, gyro and accelerometer sensors, and independent motion capabilities – to warn drivers of EVE (Electric Vehicle Evolution) cars when they get too close to the vehicles ahead of them.

Is there a red light ahead?  A green light?  A pedestrian? A yield sign? EVE vehicles will be equipped with the technology to recognize all this, and alert drivers of what’s ahead.

EVE, a US-Israeli firm, makes customized electric vehicles for high-end customers, loaded up with, for example, touchscreens and information screens integrated with the dashboard and the entertainment system.

The Bar-Ilan tech – which is being marketed by Bar-Ilan University’s Research and Development Company (BIRAD) – will not only warn EVE drivers about dangers ahead on the road, but also use sensors to provide them with real-time information about the health and functioning of every aspect of their cars. Alerts draw attention to component warnings and failures, together with additional maintenance recommendations. Battery Management System (BMS) recommendations, including resource availability that will enable maximum integration of battery charging with the driver’s personal linked schedule, are also set to be incorporated, and the data will be integrated with real-time traffic reports and other relevant information, EVE said.

The driving detection algorithm was develop by Bar-Ilan University Prof. Gal Kaminka, while Kolberg and his team are concentrating on image processing technology and the artificial intelligence functionality developed in Kolberg’s robot lab. Eventually, Kolberg said, both those technologies will be combined to enable EVE to produce autonomous vehicles, similar to Google’s driverless cars.

Such vehicles, Kolberg said, could be joined together in a sort of ad hoc network that would promote better driving habits. It’s similar, he said, to the cooperation algorithms built into soccer-playing robots his lab produces, in which the robots “decide” (based on what their cameras and sensors observe) which robot will go for the ball and kick it, which one will block an opponent, which one will wait for its teammate to pass the ball to it and which tactics to choose according to the play state.

Those algorithms (which are based on location, opportunity, and other factors) can be adapted in order to enable cars to pass each other safely and within a reasonable time interval, to identify traffic signs and act accordingly, obey traffic rules while performing turns, drive in roundabouts, and more.

As part of EVE’s efforts in this direction it has joined the US Department of Transportation ITS Connected Vehicle Project.

“Eventually our device will be able to recognize surroundings and environments and take action, even activate the brakes if necessary,” said EVE Chairman and Chief Strategy Officer Henry Kauftheil. “Autonomous cars and smart cars that can assist with driving, reduce traffic accidents, avoid congestion and save lives are just a short distance away. The automobile as we know it has been ripe for disruption for decades. Everyone thought BetterPlace was the answer, but the execution had problems. EVE has the right execution strategy, and with Bar-Ilan University’s technology, the sky is the limit.”