The public procurement procedure for the construction of a demonstration sensor network for a one-kilometer section of the M1-M7 motorway has been completed as part of the Eureka Central System project. The implementation of the system will start soon and is expected to be completed by autumn, according to Dr. András Rövid, the project’s technical manager, who is leading the consortium on behalf of BME. The project involves six Hungarian and six Austrian partners led by BME.

Once the sensor infrastructure is in place, the sensor data is processed to create a real-time digital twin model of the static and dynamic elements of the road section. This will add position, orientation and vehicle size data of the vehicles travelling along the section to the existing high-fidelity three-dimensional model. The use of such a digital twin model will allow the evaluation of a number of Advanced Driver Assistance System (ADAS) functions in a public environment, as well as the testing of services that rely on the real-time twin model.

The planned system will include multiple sensor stations at several points along a one-kilometer stretch, which will be mounted on existing support elements. Each station will consist of conventional cameras with different angles of view, thermal cameras, lidar and radar sensors, and will be installed on support structures overhanging the road or spanning the entire road. Services that will improve road safety and reduce the number of accidents will be developed and then deployed.

The digital twin model will be used to test and validate various driver assistance functions, perception algorithms and to tune traffic management models. In the longer term, it will allow vehicles to receive driving assistance data that is specific to them, even if they do not have their own sensors. The data will be accessed via a device capable of receiving them from the digital twin. The use and management of personal data such as number plates will not be part of the operation.

The system on longer term will extend the sensing range of self-driving cars, solving, for example, detection problems due to blind spots. The vehicle gets a much more comprehensive, wide-coverage view of events in its surroundings, as well as the position and movement of surrounding vehicles.

The development of self-driving conditions implies a more widespread deployment and use of similar equipment on the roads. Adequate data flow speeds are essential for its operation, as traffic requires real-time data, so minimizing latency is a priority. Data is sent from the sensors to the data center via an optical network and from there to the vehicles via a 5G network. These technologies provide the expected data speeds and short response time.

The establishment of the pilot road section and the related tests and developments will facilitate the implementation of Level 4 self-driving technology by the end of the decade. This will mean that if we want to travel from the capital to Lake Balaton, we can leave the driving to our car on the motorway while we relax or work.

The video below shows the system output on simulated data.