German
Collision avoidance sub-project
The basic idea
Mutual consideration between all road users is essential in any traffic situation. In public traffic areas, pedestrians, cyclists, cars, trucks and public transport vehicles inevitably encounter each other and potentially dangerous conflict situations can arise.
As a result of the ongoing transport transition, the proportion of vulnerable road users is increasing. At the same time, electrification is increasing the speeds at which these groups participate in road traffic. As a result, the risk of collisions with potentially serious consequences for vulnerable road users is increasing.
- Results
-
Data In Motion has developed a conflict detection algorithm that uses the data provided by the Informative Traffic Signal System (ILSA) and collected
Bild
data distributed via the SensiNact data broker to compare the traffic light circuit diagram with road users detected by the optical sensors in order to indicate a potential collision risk in real time. At the Felsenkeller traffic junction, for example, the sensors were calibrated to monitor a cycle and pedestrian path that is difficult to see for turning car traffic and leads to the traffic light junction. If cyclists approach the junction at a critical speed, the algorithm generates a warning message that adapts in real time - thanks to the low 5G latencies - and generates a warning signal depending on the traffic light circuit diagram. The functionality has been tested under real-life conditions and has reached a level of maturity that allows it to be rolled out to other traffic junctions. Future potential applications include, for example, integrated warning messages sent to road users via 5G networks (e.g. in the driver assistance system of a car or autonomous vehicle).
- Solution approach
-
As part of this sub-project, a model will be created that will be able to predict the future trajectories of people in the vicinity of conflict areas based on live movement data. This prediction serves as a starting point for a system that recognizes conflict situations in advance, derives a potential danger and assesses the risk of a collision. Based on this assessment, the system decides on measures to warn road users. To achieve this, information about traffic light systems (LSA) is included in addition to the movement data - in future, historical movement data will also be included.
The system is to be tested as a proof of concept at the following intersections in Jena:
- Schillerstraße / Teichgraben / Leutragraben
- Knebelstraße / Am Volksbad
- Kahlaische Straße / An der Brauerei
5G offers several important advantages for this time-critical and safety-relevant task. Low latency, high reliability and high transmission speeds should be mentioned here. This saves valuable milliseconds during data transmission. The use of 5G technology can also support precise positioning in areas where satellite-based positioning is unreliable due to building shadowing.
- The challenge
-
Detecting a potential conflict forms the basis for future accident avoidance applications. The question arises as to how the detected potential conflicts can be made available to road users in a meaningful and reliable way without distracting them from the traffic situation. It is important that the system gains user acceptance through its high reliability in order to ultimately achieve a safety advantage.
- The team
-
The sub-project is being realized by Data In Motion Consulting GmbH and the Technical University (TU) Dresden in close cooperation with Kommunalservice Jena and Jenaer Nahverkehr.