Limited lanes, lights, signs, and an overwhelming number of vehicles rushing in from each direction—that’s the recipe for a perfect storm or an all-but-perfect intersection such as the infamous Meskel Square in Addis Ababa.
But other examples prove that too much of the good can be just as bad. Moscow’s Taganskaya Square, for instance, actually consists of two intersections with 11 roads coming in and out, resulting in a true driver’s nightmare.
Similarly, disorganised and complicated intersections haunt traffic controllers and planners at night. Taming the ever-shifting traffic chaos seems impossible. Seems.
Here, technology comes to the rescue with a tool that allows for reacting to traffic events faster than ever before - real-time traffic analytics. We’ll shed some light on this solution and discuss how it can revolutionise traffic monitoring, control, and management in a smart but cost-efficient way.
Safety, road condition, and traffic efficiency are in everyone’s best interest, whether it’s municipalities, traffic professionals, or road users.
But in reality, many obstacles stand in the way of optimising traffic:
Thankfully, video traffic analytics provides insights into the traffic situation, helping municipalities and traffic professionals find solutions to all of these problems. For many traffic challenges, retrospective analysis that considers historical recordings is the perfect answer. But for more time-sensitive issues, traffic professionals need a tool that would allow them to react faster and make decisions on the go, such as a real-time traffic management system.
Several problems arise with conventional real-time traffic analytics technologies, such as in-road magnetic loop sensors.
Other solutions are even more specific and only capable of recording particular information. Licence plate-reading cameras are limited to licence numbers, while radars are precise regarding speed but incapable of capturing vehicle type or appearance. So, to record all traffic data, municipalities and traffic surveyors have no choice but to weave together a patchwork of solutions, which is both ineffective and costly. Or do they?
When paired with trained AI algorithms, video analytics solutions such as GoodVision Live Traffic can replace all traditional real-time traffic data tools.
Live Traffic uses live camera footage to provide traffic data from a selected road section in real-time. One of its advantages is the fact that it can use the existing CCTV equipment that, in many cases, is already placed at crucial locations such as busy intersections, which significantly reduces setup costs.
The camera is paired with an EDGE computing device deployed near the camera that uses our proprietary AI-based detectors to process the recordings on the spot. Each processing unit can process up to 4 traffic streams, depending on the device model.
The third component is the internet connection. Although the captured traffic data is processed in the computing device, it’s immediately available in the cloud, on the Video Insights analytics platform.
The combination of all three components makes up for a real-time traffic solution that’s fast, reliable, and records granular event data instantly.
Each AI detector is configured individually according to the scene and customer needs. Many factors influence the computing power needed to capture and process traffic streams: resolution, framerate, codec, or number of objects on the scene. Computing power affects the price of the deployment, meaning that we can assign less computing power for simple scenes that don’t require detailed data capture to reduce costs.
For instance, if detecting smaller objects, such as pedestrians or distant vehicles, isn’t necessary at a given location, we can deploy a standard detector that requires less computing power. The same applies to single-use roads such as highways. Conversely, for complex intersections where all traffic needs to be recorded, we can deploy a more powerful detector. Computing power can be adjusted remotely in a matter of minutes, allowing for flexibility when additional resources are needed.
Live Traffic can identify 8 object types, with the option to define custom classes.
Live Traffic can also be connected to third-party traffic control systems and send automated alerts when it detects specific road events. You can use event presets (illegal stop, speed limit violation, wrong-way driving etc.) or set up your event triggers, including virtual loops. Another example is reporting violations like line crossing, ready with the exact speed of the vehicle.
Live Traffic is an extension of our Video Insights analytics platform, meaning that all real-time traffic data extracted from the streams is instantly uploaded and available there. This includes object classification, events, volume counts, saturation flows, and more. Based on the data collected from live streams, users can generate custom traffic reports such as O-D matrix, movement counts, or vehicle lists.
The platform also offers tools made specifically for real-time traffic analysis. After defining the road scenarios users want to track, they can create analytical widgets for each event that will be automatically updated with the data extracted from live streams. By representing traffic records visually, these widgets aid traffic managers in generating insights and making more informed decisions. All widgets are accessible from an intuitive Live Dashboard for an easier overview of the traffic situation.
You can create any number of widgets for every scenario to represent data in multiple ways.
Additionally, users can view traffic streams live in the GoodVision app as long as their device is in the same network as the processing unit.
Low deployment costs coupled with the wealth of data provided by video-based real-time traffic analytics make Live Traffic video analytics a powerful tool for any organisation concerned with transportation. What’s more, its customizability makes it suitable for diverse purposes.
To warrant the highest possible safety and efficiency of transportation networks, municipalities, traffic managers, and other mobility-related organisations need as accurate data as fast as possible.
Our article only scratches the surface of Live Traffic and the opportunities it can create. Get in touch with us, and let us tell you about the full extent of real-time traffic analytics for your organisation.