Ix3 pilots achieve emission reduction
10 February 2023
During the AI4Cities project, the Ix3 system was implemented in four different locations in the capital city areas of Finland and France (Kalasatama, Suutarila, Saint-Maur-de-Fossés and Meudon) with very varying base infrastructure and technical configuration. The live pilots conducted during the last phase of the project in Meudon, Paris region, France and Helsinki, Finland showed more than 2% emission reduction from the baseline, which consisted of the average annualised reduction potential of 55.509 stops by vehicles and 360 hours of irrational waiting time. Previous phases in other intersections showed an even higher emission reduction potential of up to 7%.
Live sensing in an intersection in Kalasatama in Helsinki was carried out during the first phase of the project. The annualised reduction potential was 162.944 stops by cars, 17.822 stops by heavy vehicles and 654 hours of irrational waiting time. Together these form approximately 4% of the emissions. During the second phase we built the prototype in a cluster of three adjacent intersections in Suutarila, Helsinki. The processing was done in the cloud with dedicated cameras. The emission reduction potential was 1% when the green-wave functionality remained active, and up to 8% when it was disabled.
During the second phase we also carried out a live sensing in an intersection in Saint-Maur-de-Fossés in Paris region, France. The implementation used edge processing on a device delivered by MarshallAI and a pre-existing traffic camera. The emission reduction potential was measured as 3% of the current emissions.
In Meudon, during the third phase of the project, optimisation was feasible only outside of rush hours. During such times, vehicle stopping was reduced by 8%, which resulted in approximately 6% less emissions. Additionally, driver convenience was increased by 43%, as median vehicle waiting times were 4 seconds shorter outside of rush hours. The overall impact, including times when the solution was not applied, consisted of 1.26% less stopping by vehicles, leading to approx. 1% less emissions. Cumulatively this accounts for 1 month less waiting per year. However, while the vehicles waited less, pedestrians could not jaywalk as often, which resulted in an 3% increase in cumulative waiting times.
MarshallAI and Dynniq Finland Oy designed the Ix3 (Intelligent Interconnected Intersections) system in order to reduce emissions from road traffic and to optimise traffic flows using a broad set of AI tools. Ix3 was built to make traffic flows in cities more fluent by combining state-of-the-art machine vision and traffic management systems to form the traffic management solution for tomorrow. The system gathers and processes dynamic traffic information in real-time, making traffic in cities more efficient by reducing emissions, congestion and idle time as well as increasing safety.
Image (Cap Digital)