Research Vehicle Edgar on the Move in Berlin Autonomous driving: Testing Methodology Trialed Under Real Traffic Conditions

Related Vendors

FAULHABER Drive Systems

SIMTOS (Seoul International Machine Tool Show)

The autonomous research vehicle Edgar, developed by TU Munich (Germany), demonstrated in Berlin (Germany) that it operates safely and reliably under real and complex traffic conditions—during a drive from the Federal Ministry of Transport to Berlin's Radialsystem. In the process, TÜV tested a new testing methodology for autonomous vehicles for the first time.

As part of the mobility conference TÜV MobiCon, which took place on June 10 and 11, 2026, in Berlin (Germany), Dr. Dirk Stenkamp, President of the TÜV Association, and Ute Bonde, Senator for Mobility, Transport, Climate Protection, and the Environment of the State of Berlin, presented the results of the real-world test drive. In the German capital, Edgar completed a test drive under Level 4 conditions. "The Berlin real-world test drive demonstrates that autonomous driving is technically possible even in complex urban traffic," emphasized Dr. Dirk Stenkamp, President of the TÜV Association, and stated: "At the same time, it becomes clear: the safety of autonomous mobility is not solely determined by vehicle technology, but by how reliably a vehicle detects its surroundings, interprets traffic situations, and responds to various road and traffic conditions. These are precisely the aspects such real-world test drives can evaluate."

Operational Area Approval for Autonomous Vehicles

With the research vehicle, they tested how operational area approval for autonomous vehicles can be obtained in real traffic. An important step for the commercialization of such vehicles.

The approval of Level 4 vehicles in Germany takes place in two steps. First, the vehicle undergoes a technical inspection and type approval. Subsequently, the specific operational area must be officially authorized. For this, independent experts assess, among other things, during real-world test drives how the vehicle handles complex driving tasks and reacts to real traffic situations. "Autonomous vehicles must not only adhere to traffic rules but also understand the specific characteristics of their respective operational area," explains Stenkamp. "Traffic light sequences, signals, or traffic management can vary from city to city. Systems must therefore be tested and evaluated for their specific area of deployment."

Assessment Guide

The Berlin real-world test drive served not only the approval of the research vehicle Edgar but also the testing and further development of an independent testing methodology for autonomous mobility. The vehicle completed the route through Berlin three times. Three TÜV inspection teams accompanied the drives and independently evaluated the driving behavior based on a newly developed methodology. The driving tasks considered were based on the assessment guidelines published in the Verkehrsblatt.

The preliminary results are positive: the real-world test drive confirmed the practicality of the developed testing concept. The measurement technology used proved to be field-tested and could be integrated into inspections with a reasonable effort. The insights gained form a basis for further development of objective and harmonized evaluation criteria for autonomous vehicles. This has established an important building block for the safe market rollout of autonomous vehicles. Edgar navigated Berlin’s city traffic within the tested operational area safely and reliably. However, the vehicle still lacks key prerequisites for a regular Level 4 operational permit, particularly a technical supervision system capable of monitoring and intervening if necessary.

Access to Safety-Relevant Vehicle Data Necessary

Bringing autonomous mobility to the road is not just a question of technology. Clear regulatory frameworks for testing, monitoring, and approval of autonomous systems are also necessary.

A prerequisite for the independent testing of autonomous vehicles is comprehensive access to safety-relevant vehicle data. Inspection bodies and authorities must have non-discriminatory access to the data of autonomous vehicles to effectively evaluate, monitor, and control systems in operation. The TÜV Association therefore calls for standardized and interoperable interfaces as well as binding European regulations.

Furthermore, the inspection association advocates for a harmonized legal framework across Europe for autonomous vehicles and learning AI systems. AI in safety-critical systems must be mandatorily tested and certified—both during type approval and throughout the vehicle's operational lifecycle. Additionally, the TÜV Association calls for the introduction of a digital vehicle record, where software versions, safety-relevant updates, and inspection data are documented over the entire lifecycle of a vehicle.