AI-powered camera-sensors and Intelligent Video Analytics Technology help Noord-Holland (NL) towards cleaner, smarter, and safer bicycle travel
Mobility is essential for a vital society. As manager of roads and a public transportation concessionaire, Dutch province of Noord-Holland is in charge of road safety, regional transportation, and accurate travel information. The Smart Mobility Implementation Plan was put out by the Provincial Executive to improve mobility by using new technologies. In this use case, we take you through the problem definition, our approach, and the results of ViNotion’s contribution to this test project for the Province of Noord-Holland in the context of bicycle welfare and safety.
Connectivity options such as built-in Wi-Fi or mobile phone apps allow an increasing number of road users to communicate with one another. Traffic lights, bridges, and traffic control centers could all communicate with one another. Intelligent Traffic Light Control (ITLC) is vital to modern road management. ITLCs send information back and forth between the traffic and traffic lights. Smart camera sensors and technology that uses Artificial Intelligence to analyze video can give very accurate and useful information about traffic (such as the type, distance, and speed). This is in contrast to traditional loops or push buttons, which have a hard time telling the difference between a single cyclist or pedestrian and a group of them. If we had full situational awareness and detailed information about the presence and behavior of cyclists and other road users at intersections, we could control and prioritize traffic flows in a way that is safe and efficient. With ViSense® AI Video Analytics and smart camera technology, ITLCs can recognize, forecast, and prioritize any traffic. Individual riders and cycling groups can be differentiated, and specific road users can be addressed and prioritized
Noord-Holland did not have reliable and versatile real-time cycling statistics, which led to inconvenient routing, undesirable cycling experiences and traffic accidents.
Conventional traffic control loops and pushbuttons sometimes miss bicyclists. This caused unnecessary interruptions and excessive waits at traffic lights in Noord-Holland. Too-short crossing times for bicycles fostered risky behavior and led to dangerous situations. In ViNotion, the province found a committed partner with a highly skilled, experienced team of experts and sophisticated technology, ready to respond quickly and deliver a solution that fits the Smart Mobility Implementation Program’s vision. They hired ViNotion to explore large-scale intelligent traffic control employing advanced video interpretation for live and retrospective analysis. The main question was: could ViNotion’s Intelligent Video Analytics technology help the Province of Noord-Holland create a safe, fast, and convenient cycling experience?
Living Labs across the province
Pilots and tests for the Smart Mobility Program were mostly done on provincial roads within the province. This led to the creation of several so-called Smart Mobility test sites, where things like communication between traffic and infrastructure, the provincial traffic control center, and smart traffic lights were tested in the real world. There were pilots for every mode of traffic: bicycles, cars, freight, public transportation, and shipping.
Improve the traffic light control system with smart camera sensors and AI video analytics.
ViNotion integrated Video AI technology with the pre-existing Traffic Light Control system to gain in-depth, accurate, real-time, and historical cycling data. We installed camera sensors with AI visual detection and analysis. This provided complete situational awareness and historical and real-time insights for dynamic, data-driven traffic light regulation.
The result is a more appealing travel experience for cyclists, free of annoying interruptions and long waits at traffic lights, and increased safety for cyclists owing to dynamic and intelligently managed intervals.
These are the steps that got us there:
Situational analysis and sensor planning
We started with the situational analysis. There was already an ITLC system in place. Loop detection has proven less effective over time, not only because nowadays bicycles have fewer metal parts, but also because it just doesn’t work well when cyclists ride close together. Noord-Holland tested cycling apps, smart bicycles, and ViNotion’s smart cameras to give them in-depth insight and complete situational awareness in real time for dynamic traffic light control and emergency intervention.
After choosing the site (Crossing of Cruquiusdijk-N201), we established the sensor plan. Smart cameras would be put on light poles that are already there, which would make installation faster, but the system should be self-contained. While examining the existing lighting poles, the camera sight lines were carefully considered to guarantee the optimum coverage of the bicycle lanes. The stopping line and approaching bicycles had to be visible at least 30 meters away. This distance would allow the traffic light control system to trigger green light intervals for cyclists.
The installation plan and implementation
Installation began when the sensor scheme was authorized. We created a stand-alone installation plan. With clearance to use the light poles’ night power, the equipment was fitted with batteries to assure 24/7 operation. An LTE modem ensured connectivity. With excellent planning, the installation took a few hours. After the ViSense® software was installed, it was set to send Cooperative Awareness Messages (CAM) to the Dutch UDAP platform. These messages included the GPS coordinates, speed, and direction of cyclists. We were prepared to go on to the next stage.
Linking the Intelligent Traffic Light Control system with the UDAP platform
ViNotion worked extensively with the Province of Noord-Holland after installation to evaluate and validate CAM communications. We tested and validated message transmission to the ITLC system for a month to make sure it worked well. Since ViNotion sensors are used to connect to the ITLC in real time and to stream the live data on bicycles, the sensor was tested by picking a few points in time to compare the number of cyclists counted by ViNotion with the number of cyclists on the road at that time. CAM testing could commence with a detection accuracy of 99% during the day and 94% at night. CAM messages were not displayed in the ITLC system during the initial round of testing. The components of CAM were compared to the ETSI standard and other messages received by the ITLC system using ViNotion. The supplier of intelligent traffic infrastructure supported testing and reporting. Messages were received correctly upon altering CAM’s time format. The second stage was completed on-site, employing a special app for road managers capable of receiving live CAM and transmitting its position to compare the messages within the traffic control application. The testing phase had gone smoothly; now it was time for final validation.
The bicycle applications, smart/connected bicycles, and ViNotion sensors are 3 separate sources that were compared in a final benchmark. Employees from the Province of Noord-Holland drove across the intersection, using one cyclist app or smart bicycle at a time. In the traffic control system, every CAM could be confirmed. ViNotion handled matching messages from the apps and ViSense® for each specific cycling route. Verifying that each message could be matched and comparing how accurate the GPS was, allowed us to move on to the next stage.
Real-time live data into UDAP platform
After validating all system components, Noord-Holland optimized traffic installation protocols to extend the green light time for bicycles and prioritize green lights for cyclists. With real-time live data in the UDAP platform, the messages got to the ITLC fast enough for cyclists to get longer green times, allowing them to keep going without stopping at the intersection.
Results of the smart mobility test project
The test compared 3 data sources: cameras, cycling apps, and connected bicycles. As a result of our approach, ViNotion, commissioned by the Province of Noord-Holland, demonstrated that:
- Timely bicycle detection and accurate localization for dynamic traffic handling are achievable.
- A guaranteed real-time link between Intelligent Traffic Control systems and the UDAP platform is feasible.
- Data provided by ViSense® is reliable, most versatile, and of the highest accuracy.
- Compared to the other data sources, ViSense® is the only source that captures all bicycles and has the most trustworthy GPS localization.
The impact of Smart Mobility applications on provincial mobility challenges
The test fits within the Smart Mobility Implementation Program’s pillars and contributes to a number of provincial responsibilities. It brings new information and knowledge that is shared with other authorities, knowledge institutions, and private parties on a regional, national, and European scale. The conclusions below provide some examples of the consequences for livability, safety, and accessibility.
Conclusions: Impact on traffic flow, traffic safety, and livability
The contributions of smart mobility to greater livability are promising. It is technologically feasible to prioritize (groups of) bicycles at traffic lights to increase the cycling experience and encourage bicycle use. For this purpose, we would need to equip traffic light control systems with intelligence (ITLC). ITLCs make traffic flow more smoothly and cut down on the number of times people must stop at traffic lights. This not only enhances accessibility, but it also stimulates environmentally friendly transportation.
In Noord-Holland, 16 of the 52 planned ITLCs were installed by 2019. By 2025, the province’s 253 traffic light control systems are projected to be intelligent. Paying attention to healthy mobility innovations and promoting efficient use of energy and space contribute greatly to livability.
Prioritizing bicycles
Being unable to traverse junctions swiftly and easily frustrates cyclists. Intelligent Video Analytics helps Intelligent Traffic Control systems detect bicyclists and groups of cyclists. With this information, we can grant cyclists precedence at traffic lights to encourage cycling.
Applications modality Bicycle
- Prioritization of groups of cyclists at ITLCs
- Priority on cycling through routes
- Bicycle detection for caution warnings
- Safety analyses in relation to vehicle automation
The test showed how useful ViNotion’s device-independent sensors and sophisticated video analytics are to improving Intelligent Traffic Light Control.
The results of the tests and the benchmarks were acknowledged positively. The test found that the low-speed heading needed to be rectified and that a stable connection with UDAP was required. With that, a trigger signaling disruption of connectivity is added. With these two changes, ViNotion now ensures continuity and advocates permanent deployment.
Want to learn more about smart mobility and intelligent traffic light control?
- Click here to watch the project video
- Click here to go to the Progress Report Implementation Program Smart Mobility
- Continue to Intelligent Traffic Light Control here
- Read the article: Not going anywhere soon? Cutting-edge Video AI Analytics Helps Reduce Traffic Jams
- Or check the ViDigest Media Library for Intelligent Traffic Control News and more here
Do you have any information gaps on traffic at intersections? Want to improve the flow of traffic and safety at crossing points? Need to cut down on noise and emissions from traffic?
Get in touch; we’ll talk about your mission and your situation—and explore whether our expertise and technology can help you.
