The TrafficGo traffic optimization project, whose testing had begun in Turin in 2019 thanks to the collaboration between Torino City Lab and Huawei Research Center in Munich, has at its center an ITS (Intelligent Transportation System) called TrafficGo, based on Artificial Intelligence algorithms in the field of machine learning mechanisms at the service of urban traffic management.

The Piedmontese capital has been chosen by Huawei as a continental reference hub for testing the use of adaptive traffic optimization systems: the objective of the experimentation was to validate Huawei's solution with the supervision of 5T, a subsidiary agency of the City of Turin that manages the Mobility Centre of Turin in a simulated environment, given Huawei's desire to experiment and validate artificial intelligence processes in protected situations in order to test their safety and reliability. This choice was also perfectly in line with what are the objectives of the EU in terms of development of an artificial intelligence "that can be trusted", the so-called Trustworthy AI.

The phases of the experimentation

1. Definition of the simulation scenario: reconstruction of traffic flows of a typical week in Turin, based on data collected along the road network by 5T for the development of traffic control measures and existing traffic light plans.

2. Breakdown into two different scenarios based on sensor measurements and data in possession.

3. Evaluation of the effectiveness of traffic optimization algorithms applied to the two scenarios compared.

4. Intervention of 5T experts for the analysis and revision of the results obtained regarding the average waiting time at traffic lights and the amount of CO2 produced by the simulated traffic in the two scenarios.

Thanks to this experiment, Huawei has been able to increase its knowledge about the creation and modeling of simulation scenarios, exploring all the complexities involved in providing the machine learning mechanism with a model as close as possible to reality. Since it is an adaptive technology exploiting artificial intelligence, this solution can be applied to many other European urban realities, thanks to the possibility of developing tailored traffic models to be provided to the algorithm for reworking and adapting to new scenarios. Huawei's solution exploiting the computing power of cloud computing, combined with AI, will offer a very wide range of predictive solutions useful for traffic management in the coming years, revolutionizing the management of flows in the cities most affected by this problem and its ecological implications.