Hyundai will seine autonomen Autos mit menschlicher Intuition ausstatten. Klingt spannend, schauen wir uns an!
Keine News mehr verpassen! Jetzt den Newsletter abonnieren .
Hyundai wants to teach its autonomous cars to think like humans, investing in AI specialists Perceptive Automata to develop new software for self-driving vehicles that will mean your car is as good at spotting unpredictable hazards as you are.
Hyundai’s Cradle divison – a department set aside for innovation, not unlike BMW’s i division – made the decision to partner with the American startup after Perceptive Automata developed new software that is able to predict the often unpredictable intentions of pedestrians, cyclists and other road users.
“One of the biggest hurdles facing autonomous vehicles is the inability to interpret the critical visual cues about human behaviour that human drivers can effortlessly process,” says Hyundai Cradle vice-president, John Suh. “Perceptive Automata is giving the AV industry the tools to deploy autonomous vehicles that understand more like humans, creating a safer and smoother driving experience.”
Considered one of the stumbling blocks to our autonomous future, the erratic behaviour of human road users is much harder to predict than that of other self-driving cars. The latter, it is presumed, would be talking to each other, but humans remain the unpredictable off-line element.
Perceptive Automata tackles the issue by creating software that allows an autonomous vehicle to think like a human, installing “human intuition” that allows for what it calls “rapid judgements” about what pedestrians, cyclists and other motorists might do next.
It works by processing data that shows how people and cars interact on the road, which is then used to teach artificial intelligence software to interpret human behaviour in the same way another person would. The result, says Perceptive Automata, is an autonomous car smart enough to detect when a pedestrian might suddenly cross the road, or even if that person knows there is a car on the street.
One example given is that, if a pedestrian sees a car coming but waves the vehicle past, an autonomous car would still wait until either the person crossed the road or stepped back from the edge of the street. But with Perceptive Automata’s software, the car would recognise it had been waved on and continue driving.