Every year, millions of road accidents occur worldwide, and the vast majority of them – 94% according to statistics – are caused by human error. But what if vehicles could anticipate danger and react faster than the most experienced driver? This is the question being addressed by the Systems and Control Theory (SCL) Research Laboratory at HUN-REN SZTAKI, which is developing new solutions in collaboration with world-leading technology and engineering companies.
Self-driving cars that really listen to you
Imagine a car that not only follows the rules, but is able to anticipate a pedestrian coming out from behind a parked bus and change its route in time to avoid trouble. Although self-driving vehicles are becoming more advanced, adapting to real traffic situations – and the unpredictability of human behaviour – remains a major challenge.
SZTAKI SCL has been at the forefront of solving such problems for decades. The lab develops artificial intelligence-enabled control systems that enable autonomous vehicles to not only react, but also think ahead – learning and adapting to complex environments.
Vehicles learn on a virtual test track
At the heart of the development is a special test environment called AI MotionLab, where researchers test systems not only in simulation, but also in augmented and mixed reality. They create situations that would be too dangerous to try in real life – for example, a child suddenly stepping out onto the road in front of a car.
Here, researchers can model virtual pedestrians, cyclists and even extreme weather conditions that vehicles – even miniature real cars – interact with. This allows developers to test the behaviour of autonomous systems in a safe and cost-effective way.
Faster learning, less risk
By simulating rare but critical accident situations, vehicles can gain experience that would be difficult to gain in real life. This not only speeds up the learning process, but also helps to make self-driving systems more reliable before they are actually put on the market.
SZTAKI’s research not only improves the safety of cars, but also the speed of development. No wonder more and more domestic and international technology companies are turning to them for cooperation.
More than cars – the transport of the future is taking shape
SCL’s activities go beyond self-driving cars. The lab is also involved in the development of safety-critical controls for intelligent transport systems, vehicle networking, and applications in aerospace, rail and industrial energy.
“Our algorithms not only improve the performance of individual vehicles, but also provide the opportunity to redesign entire transport systems – making them safer, more efficient and more sustainable,” said Professor Péter Gáspár, head of the lab.
Although self-driving vehicles are not yet dominating the roads, research at SZTAKI is already contributing to making future transport smarter and safer.
Read the full article HERE to read.
