The project “5G and artificial intelligence-based public transport optimisation system”, which AAI Labs is implementing together with Vilnius University, UAB “INNTEC”, Vilnius City Municipality and SĮ “Susisiekimo paslaugos” (JUDU), is gaining momentum. The project is funded under the European Union’s economic recovery and resilience instrument NextGenerationEU.
Sensors and artificial intelligence forecasts being installed in the city of Vilnius will make it possible to monitor passenger flows in real time, anticipate loads for the coming hours and, if necessary, dispatch additional vehicles in advance.
Aistis Raudys, CEO of AAI Labs, said that this project represents a fundamentally new step for Vilnius in how a city can manage public transport using artificial intelligence and real-time data, and at the same time a great proof that solutions of this calibre can be developed right here in Lithuania.
At this stage of the project, the main technical preparatory work has already been carried out and the first solutions have been tested under real city conditions together with Vilnius City Municipality and JUDU. Systematic data collection for training artificial intelligence models has begun. Information is being gathered on traffic, congestion, road conditions, public transport movement and passenger flows, which will be used to train the artificial intelligence.
Two main technological solutions have already been tested in the project:
- A fixed 5G camera, mounted at a specific location (e.g., at an intersection). It can monitor traffic, calculate transport flows, help identify congestion and record dangerous situations.
- A mobile 5G camera, installed in a public transport vehicle, e.g., a bus. It makes it possible to assess road surface condition (potholes, unevenness, obstacles) and monitor traffic as the vehicle moves through the city.
Two additional solutions were also tested:
- Monitoring passenger flows based on wireless signals, i.e., attempting to assess boarding and alighting based on detected devices. Practical tests showed that this method is not yet sufficiently accurate due to technical limitations and privacy requirements.
- Monitoring the technical condition of vehicles, whereby data such as engine or other equipment parameters are transmitted via 5G. Limited access to this data was encountered here, so this solution is not yet fully developed at this time.
In summary, at this stage of the project we already have:
- a verified 5G technical base in the city, and data collected for artificial intelligence models,
- working demonstrators (a fixed and a mobile camera),
- clear initial use cases (congestion monitoring, road surface assessment)
- and identified risks.
Going forward, the project plans to focus on the technologies that can realistically be applied in the city – primarily road surface monitoring and traffic load and congestion analysis using fixed and mobile cameras – and to bring these solutions to fully functioning demonstration versions. Final conclusions and recommendations for the city on practical application will be prepared later, together with the project’s final report following the EP1–EP4 stages.