Chalmers researcher Ola Benderius said the traditional way of developing vehicles is to constantly base progress on earlier vehicle models and to add new functions gradually. But he believes this method might not work when developing the autonomous vehicles of the future.
“Traditionally,” he noted, “the aim has been to try to separate and differentiate all conceivable problems and tackle them using dedicated functions, which means that the system must cover a large number of scenarios. You can cover a large number of different cases, but sooner or later the unexpected occurs, and that’s when an accident could happen.”
Instead, the team has chosen to regard a self driving vehicle as a completely new type of vehicle that is more like a biological organism than a technical system.
“Biological systems are the best autonomous systems we know of,” Benderius claimed. “A biological system absorbs information from its surroundings via its senses and reacts directly and safely.”
Information captured from sensors and cameras is converted into a format that resembles the way in which humans and animals interpret the world. This, the team claims, enables the truck to adapt to unexpected situations in its basic design.
Instead of just one large program with dedicated functions for all conceivable situations, the team is working on small and general behavioural blocks that aim to make the truck react to various stimuli, just like an animal. The truck is programmed to constantly keep all stimuli within reasonable levels and will learn to do this as efficiently as possible. This makes for a flexible framework which is good at managing sudden and new dangers, Benderius contended.
“We are trying to design a system that adapts to whatever happens, without pointing to specific situations – and this is something that even the simplest animals can usually do better than existing vehicle solutions.”
The software, called OpenDLV, is being developed as open source code and is freely available on the internet. In this way, Benderius and his group hope that other researchers will join the project by running and developing the software in their own vehicles.
