Smart traffic lights – Testing of AI-driven traffic management system enters second round
In an earlier post Prof. Dr.-Ing. Anton Kummert explained the potential of AI-supported analysis software for smart traffic flow control and used the practical application of LiDAR sensor systems in Solingen to illustrate the practical opportunities of smart traffic technology for cities. Now we present the first results of the pilot project and further deployments within the framework of bergisch.smart_mobility.
In March 2021, testing of a LiDAR sensor system with AI-based data analysis software for use in intelligent traffic flow measurement and control began in Solingen as part of the KI.NRW flagship project bergisch.smart_mobility. The abbreviation LiDAR stands for Light Detection and Ranging. This is a method for optical distance and speed measurement using laser beams that are invisible and harmless to humans.
Seven months full of new discoveries and further developments
LiDAR sensors have been collecting traffic data at the intersection of Hildener Straße and Untere Benrather Straße in Solingen-Ohligs for just under seven months. Thanks to the analysis software developed by the Munich-based tech company LiangDao, which was specially adapted for use in a smart city during the trial, the collected data can be visualized and evaluated. In addition to the Technischen Betrieben Solingen, the Bergische Struktur- und Wirtschaftsförderungsgesellschaft and the Bergische Universität Wuppertal are also involved in the evaluation and are promoting the further development of the measurement methodology and analysis software. Regular system updates, raw data dumps, evaluations as well as a sensor exchange last carried out in September ensure constant new findings and improvements of the system.
Innovation development hand in hand: New synergy for further testing in Wuppertal
The successful installation and use of the software and hardware for traffic counting in Solingen has paved the way for further testing: The road traffic engineering department of the city of Wuppertal is adopting the findings and will also test the LiDAR system – this time, however, not for traffic counting, but for AI-based control of a traffic light system. The plan is to link the sensor technology and software with the control unit of a traffic light system – this will be done using a hardware component developed in-house especially for this purpose, a prototype of which is being worked on by Wuppertal Traffic Engineering together with LiangDao. The contactless demand request made possible in this way will initially benefit cyclists and pedestrians, but can basically be used for all road users.
Artificial intelligence could take over traffic control as early as 2022 on a test basis
If the prototype development for traffic light control continues as planned, a traffic light system could be set up in Wuppertal as early as the first half of 2022 in which the pushbuttons are switched off and the LiDAR system independently takes over traffic light control with its artificial intelligence. In the future, for example, the times of the red and green phases for cars, bicycles and pedestrians could be adapted to the respective traffic volume, depending on the time of day. This would permanently ease traffic flows and efficiently relieve particularly busy traffic junctions in the city.