MarshallAI and Dynniq Finland Oy have designed the Intelligent Interconnected Intersections system to reduce emissions and optimise traffic flows using a broad set of AI tools. The system gathers and processes dynamic traffic information in real-time in a way that’s never been done before. It makes traffic in cities more efficient by reducing emissions, congestion and wasted time with increasing safety. Intelligent Interconnected Intersections is part of a 3-year EU-funded project aimed at exhilarating carbon neutrality in European cities.
The Intelligent Interconnected Intersections system consists of three main parts: camera sensors, MarshallAI machine vision platform and integration to the Dynniq traffic management systems and controllers.
Traffic cameras monitor and understand precise traffic flows and classify them as different objects, such as vehicles and pedestrians. The AI functionality also monitors their speed, direction, routes, dwell times and interactions with other objects.
Exceptionally detailed and accurate real-time traffic data is collected and can be used for traffic planning and easily provided to third parties when needed. It can also be used in cooperation with external systems, like dynamic LED-signs that direct traffic on different routes when congestion is detected. The logic can be altered to best suit the current traffic situation without human interaction or reducing traffic safety.
Intelligent Interconnected Intersections reduces emissions by altering the traffic management logic and removing unnecessary idling in traffic lights. This is achieved by using vision-based information of real traffic situations and prioritising traffic flows. When the traffic light management system understands the volumes of vehicles and their types, traffic can be managed in an optimal way to reduce CO2.
The solution can, for example, be programmed to let heavy traffic and public transport through more fluently or alternatively promote green values and prioritise cyclists. Connecting multiple adjacent intersections together allows broader understanding of traffic and enables pre-emptive actions, such as ensuring a green wave for a certain vehicle type. The possibilities are broad and can be planned together with the partner cities and their specific requirements.
As optimal traffic management reduces unneeded idling and stopping, it simultaneously makes traffic more pleasant for all traffic users. The advanced machine learning features promote traffic safety by detecting traffic violations, accidents and even close-calls. This means hazards can be identified before accidents occur. Fluent traffic (and good understanding of it) means less emissions, increased safety and a happier public.