The world of Ai is influencing every aspect of our lives from powerful chatbots to walking robots. Love it or hate it Ai is here to stay and its capabilities are endless. In this article, we will narrow down the focus to the way Ai is being used in traffic safety in particular focused on Ai-powered edge cameras and their roles.
Cameras can be found mounted on most street intersections, street lights or stand-alone poles, more and more of these cameras are switching on Ai analysis to provide cities with far more data than they ever had. We all assume that our cities have a good understanding of the traffic conditions, congestion points or most dangerous intersections. However, that isn’t always the case.
One of the biggest benefits Ai cameras are providing is a constant analysis of activity, which is now providing a baseline data set for cities to evaluate their assumptions. Without this critical data, cities can only assume the impact traffic has on their citizens and in particular their safety.
Before we jump into the benefits of Ai cameras on traffic safety let’s first understand what edge Ai camera detection means. Simply put, edge detection is Ai analysis done on the camera, on the edge of the communication chain. This means the video and images are analysed on the camera and not sent back to a cloud destination. Naturally, this improves public privacy because only the camera can effectively see the video and images it is analysing and then send the data to the cloud or to the end user’s software. This also has an impact on the cost of analysis when you are only sending zeros and ones rather than large video files every second.
The Ai can be very specific in its analysis process, it will apply the Ai to the field of view set, without any distractions for 24 hours straight, if required. This means the amount of data can be enormous, but incredibly powerful for understanding all aspects of traffic activity.
Where can the traffic data be useful to cities attempting to improve their traffic safety, here are a few examples and use cases starting with vehicle management and then looking at pedestrian safety:
Speed limit enforcement: simply placing cameras that can detect speed and relay data to the enforcement agencies can improve driver behaviour. In Victoria, Australia the simple reduction of the speed limit by 10km in urban environments resulted in a 25-49% reduction in pedestrian injuries.
Crosswalk monitoring: Cameras mounted at crosswalks or intersections can monitor pedestrian behaviour, providing valuable data that is used to identify dangerous crosswalks or intersections, near-miss situations and repetitive pedestrian behaviour resulting in accidents. This is valuable data for town planning and incident reduction.
Pedestrian volume monitoring: AI computer vision edge cameras can assist traffic management teams in identifying high pedestrian traffic regions and providing ideas for traffic flow management and enhancements by analysing pedestrian traffic patterns. This naturally contributes to vehicle traffic flow improvements as well as providing data to better manage pedestrian movements.
Pedestrian counting: When it comes to making improvements cameras can also be used to count the number of pedestrians crossing at a particular intersection, which can be helpful in determining the need for crosswalk improvements, such as adding pedestrian signals, crosswalk markings, or street lighting. Although this appears to be a basic function it is critical in prioritising city planning, without data the severity of the situation will never be identified.
Once a city has continuous reliable data it can make decisions to implement measures that change the behaviour of the drivers and further reduce pedestrian fatality or incidents.
Both vehicles and pedestrian gain from the implementation of edge Ai cameras, the data alone has already led to a reduction in curbside incidents and as more data is collected cities will have a better understanding of the streets, intersections and crossings leading to management processes and planning.