Traffic lights

Contactless green request - how does it work?

One or more cameras record the flow of traffic in the crossing area. An on-site computer system specially trained in object recognition analyzes the data and automatically detects waiting pedestrians. This information is transferred to the traffic lights and a clearance is requested.

Traffic analysis - what exactly is analyzed?

The system recognizes objects that usually occur in road traffic and assigns them to the appropriate categories. For example, it can distinguish between cars, bicycles, buses, pedestrians and others. Based on the number, speed and direction of movement of the classified objects, the system decides which direction of traffic should currently have priority.

Controlled release - what does this mean?

The release time is the time during which a traffic flow is shown green. By detecting objects and their direction of movement, the system can react to special situations. For example, an approaching line of cars can be waited for before the pedestrian is given clearance if this can reduce overall congestion times. In the same way, the green light for pedestrians can also be extended after school classes or people with restricted mobility have been detected.
By the way: Did you know that the pedestrian green is a start signal? If you have just entered the street and the signal changes to red, you still have enough time to cross as a normal pedestrian before vehicle traffic is cleared. This applies to every traffic light in the city.

Is image capture compatible with data protection and privacy?

Data protection and privacy are GDPR-compliant. This is ensured by the following measures, among others:

• Images are captured exclusively for the purpose of optimized signal control
• The images are discarded immediately after analysis on site
• There is no data storage, no data export and no data processing in a cloud
• Image sections that are not required for the above-mentioned purpose are permanently blanked out when the system is set up
• Individual personal characteristics are not processed
• Live streaming of the camera images in the sense of a webcam is technically impossible
  

Thermal imaging cameras and conventional camera technology are used for recording.

Will all traffic lights in Jena be switched intelligently?

The new technology will be used gradually as part of a funding program and will initially be limited to pedestrian traffic lights. We will then evaluate the technology. If its use proves successful, it could conceivably be extended to complex junctions.

What goals are being pursued?

You have probably already noticed that when you are at a pedestrian crossing and request the green light, you wait a long time even though there is hardly any traffic on the road. Or you are driving in a large group of vehicles and have to stop again and again, which can also harm the environment.
The aim of this new control system is not only to distribute the limited release windows available in a "cycle" more fairly, but also to adapt them dynamically to the traffic. This means that the clearance times can be dynamically extended if necessary, for example when a large crowd is approaching or when very slow pedestrians are crossing the road. In future, the traffic lights will also be able to react more quickly to a lower traffic density, reducing the waiting time for people crossing the road.

Is the project funded?

Yes, the project is being funded in accordance with the Federal Ministry of Digital and Transport's Digitalization of Municipal Transport Systems Directive. The technology required for this was put out to public tender and will be retrofitted to selected traffic lights in the city of Jena from May until the end of 2024.

What happens if the technology goes on strike?

When introducing new technology, unforeseeable challenges can arise, especially in the initial phase. Be it that setting parameters need to be readjusted or detection difficulties need to be resolved. In such a case, the traffic light "falls back" to its familiar switching pattern without intelligence. For this reason, the familiar "yellow pushbuttons" will continue to receive the green requests.

Can traffic chaos occur if every traffic light decides autonomously?

1. Several protective mechanisms work together to prevent this:
   Object recognition is a separate process from the traffic light. The traffic light first processes the current phase and only switches the desired image at the next possible point in the program sequence. If object recognition is disrupted, the traffic light continues to run normally.
2. The traffic lights in the urban area are connected to each other and to the traffic computer via a dedicated network. They are permanently monitored there, allowing faults to be detected at an early stage and downtimes to be minimized. Higher-level coordination can also take place.
3. A redundant computer system on site constantly checks and logs impermissible signal states. If such a fault is detected, the system shuts down before the signal status is visualized. Green for two conflicting traffic flows are technically ruled out as a result.
4. The logs are legally binding.

What are the specific advantages of intelligent traffic lights?

• Increased acceptance thanks to adaptively controlled traffic light phases that are based on real traffic flows. In practice, the system is designed to handle traffic on the main route and pedestrians crossing in batches during periods of heavy traffic, whereas during periods of low traffic, the focus is on ensuring a passage with as little delay as possible, even if the main direction is permanently green.
• Waiting times at traffic lights can be reduced by automatically registering pedestrians with a recognizable desire to cross
• Extending the clearance time for people with mobility impairments based on their position and speed
• The city hopes that the introduction of the system will reduce traffic congestion and accidents by optimizing traffic flows across the board.