When it comes to safety, cities across the globe share similar concerns. Urban mobility is rapidly evolving with shifting usage and new vehicles. About half of the road accidents occur in intersections, many of them ending up with fatalities. In this context, improving the citizen’s security in crossroads is an important goal for municipalities, especially in a context where different forms of mobility are developing (cars, pedestrians but also bikes, electric scooters, etc.).
The use of video detection, combined with A.I. analysis, and early warning systems can have a critical role in improving safety. It can allow both to better identify risk situation to plan long term transformations of road infrastructures and shorter-term solutions. This can take the form of early warning systems for road users and for emergency services.
However, the deployment and use of this technology raise several challenges. First, low latency and rapid treatment of the collected information is essential. Uploading collected images to cloud infrastructure that have high processing power but are distant may prove impractical. And furthermore, having unedited pictures of street interactions transit through the internet raise important privacy questions.
Kentyou promotes solutions based on Edge computing and Artificial intelligence to better manage urban road intersections.
Edge computing allow to perform an initial analysis of the captured video images as close as possible to the intersection. This not only reduce the amount of data to transmit to cloud-based infrastructure (thus reducing latency). But it also protects the privacy of road users as only anonymized data are transmitted.
The Eclipse sensiNact platform allows for the deployment of data analytics and artificial intelligence close to the edge of the network. The modular nature of the platform allows for optimised deployment, integrating only the required component, and thus reducing the required computing power. This allows for deployment on devices with limited computing capabilities.
Furthermore, the use of the sensiNact platform allows to correlate the video analytics with other sources of information such as sound recognition or other IoT devices feeds. The interoperability of the sensiNact platform enables to diversify sources of information and thus increase overall understanding.
The Kentyou Eye software suite is used to display the status of the intersections on a convenient dashboard. And Kentyou team is working on data analytics to help cities transform the data into actionable information. Our objective is to build a complete decisions support solution that can help city to both respond in real-time to specific risks and to plan ahead important transformation of their road infrastructure.
The digital twin approach of Kentyou Eye and its ability to store, analyse and display historic data allows to visualise the status of the intersection over different period of time. This allows for longer term perspective and analytics of the road infrastructure. It can also be used to built models and forecasts of the future status of the infrastructure and of how modifications can impact traffic conditions.
The integration of intersection monitoring in the Kentyou Eye mobility hypervisor also allows for cross analytics with other mobility data. This can allow to visualise impact on traffic and emissions, but also alternative solutions such as soft mobility or public transport. It provides municipalities with a complete perspective that allow to better plan future solutions.
These developments of road safety solutions have been initiated and made possible by the participation of Kentyou to several collaborative research projects.
The DECENTER project, a EU-Korea funded collaboration project enabled an early development of a proof of concept and its deployment in two intersections in the city of Trento (Italy).