Bremen-Borgfeld is a typical and pleasant suburb. Surrounded by farmland, it has residential areas, schools, shops and restaurants. The ideal place to live, away from the hustle and bustle of the city – or to put a self-driving car to the test.
Compared to driving on a motorway, navigating around a residential street in Borgfeld is a challenge. "Rural roads often lack clear boundaries. Sometimes there's a pavement, sometimes there are trees and bushes, sometimes there are parked cars, cycle lanes or gravel. It's very difficult for a computer to recognise all this and tell the difference," says Dr Mitja Echim.
The research fellow at Bremen University's Center for Industrial Mathematics (ZeTeM) is an expert on self-driving cars. Over the past two years, for example, he has been involved in the AO Car project, in which a self-driving VW was sent on hundreds of test drives. This experimental vehicle is also key to the follow-up project, 'OPA3L – Optimally Assisted, Highly Automated, Autonomous and Cooperative Vehicle Navigation and Localisation'. It was launched in early 2019 and will run for four years, with a total budget of €5.3 million.
Closing the mobility gap
OPA3L conducts research into cooperative, networked and autonomous driving in rural suburbs, which is still uncharted territory even for the large car manufacturers. "The drive assistance systems developed by Tesla, Mercedes and Google are already very good at navigating along motorways, but it's a very different story on rural roads. We believe there's still a lot of research to be done in this area," says Echim.
The researchers are anticipating high demand for self-driving vehicles in rural areas. Ride-hail services, for example, could shuttle people from their front doors to local public transport and back again. "We believe that the advent of self-driving cars will see car ownership fall to a lower level than today. Anything else would make little sense, as it would hardly relieve transport issues and reduce traffic jams. The future belongs to ride-hail services – just like a taxi, but without the driver," he adds.
This is what the researchers call 'vehicle on demand', a service that will open up a whole new dimension of mobility for commuters, the elderly and people with disabilities.
Heaven and earth, close together
The project team, which has more than 29 members, is taking a different route than the research departments at the large car manufacturers. The algorithms they are using to control the car in Bremen were originally developed for space exploration. "Robots and space probes have been able to autonomously navigate in space for some time now. We have taken the technology behind autonomous space vehicles and applied it to road traffic," says Echim, who manages the OPA3L project.
The goal of the research is to create uniform software environments and artificial intelligence that can be further developed for a wide range of applications. This is in contrast to the car manufacturers' research departments, which are developing algorithms for their own purposes only.