Car ownership has been increasing over the last three decades in Sofia, Bulgaria and the city has one of the highest numbers of private vehicles registered per capita in Europe[1]. This has led to growing traffic flow, congestion and worsening air quality. To address these and other mobility challenges, the city’s Sustainable Urban Mobility Plan is focused on delivering an integrated transport system by promoting the use of sustainable transport modes, greening the city’s transport system and promoting its digitalisation. Sofia-based transport technology company Theoremus has been playing a significant role in the digital public transport transformation in Sofia.
Theoremus specialises in developing and implementing large, complex transport data systems that integrate mobility modes in order to drive the uptake of sustainable transport. Theoremus’ systems empower cities, and public transport authorities and operators to compare, evaluate and improve public transport networks and measure the impact of mobility policies and urban interventions.
Driving the digital transformation
In 2021, Theoremus partnered with a consortium of companies led by CNS, to deliver the Sofia Ticketing System, a modern account-based ticketing system using open-loop payment systems and Apple/Google wallet compatibility, processing 1.2 million journeys a day.
The following year, Theoremus and A1 teamed up to deliver a next-generation Automatic Vehicle Location (AVL) system in the city that uses GPS data from thousands of vehicles in real-time, alongside AI analytics and prediction capabilities, to provide estimates about the expected schedule and route deviation for over 1,200 overground public transport vehicles. Additionally, the company has worked to integrate public transport data to the National Data Access Point (NAP) in compliance with EU regulation together with global partner ESRI.
Currently, the company is working alongside UCL, Sofia Urban Mobility Centre, Romanian Smart City and Mobility Association (ARSC) and the Bucharest Transport Company to further develop and test its crowd-monitoring solution as part of the EIT Urban Mobility co-funded project BCMS.
Public transport systems are under increasing pressure
Urban transport systems across Europe are under increasing strain from rising passenger demand, irregular traffic conditions, and limited real-time operational control. These pressures often lead to overcrowding and bus bunching—two critical issues that significantly deteriorate the quality and reliability of public transport.
Overcrowding issues can mean commuters face long wait times, are denied boarding, or experience uncomfortable journeys due to uneven passenger distribution. Operators often lack a scalable solution for real-time passenger load monitoring and dynamic crowd management.
Until recently, passenger crowding insights relied on systems that transmitted high-volume CCTV data to remote servers for processing. However, such systems have proved difficult to scale with high-speed data transmission infrastructure, a prerequisite for this model, absent in many cities, limiting the reach and usability of the system.
Moreover, existing tools were designed for use by technical specialists, creating a barrier to widespread adoption among public transport operators with limited IT capabilities. A more decentralised, user-friendly, and cost-effective solution was needed to empower operators to manage crowding.
Tackling overcrowding with AI
As part of the BCMS project, Theoremus is testing an AI-driven passenger crowding measurement technology (CMT) – transforming it from a backend system into a real-time, edge-based solution. This shift involves relocating the image processing function from central servers to onboard devices—such as CCTV or ticketing units—installed directly on buses.
By processing data locally, this approach eliminates the need for high-bandwidth connections, making the solution viable for a wide range of urban environments and ensures data privacy compliance. Additionally, the project will deliver a custom operator-facing dashboard that visualises both vehicle crowding levels and locations on an intuitive map interface. Powered by UCL-developed models, the system also suggests real-time intervention events, helping dispatchers dynamically manage fleet distribution and alleviate congestion where needed. Through these interactive dashboards and visual tools, stakeholders are empowered to make more transparent, participatory, and evidence-based decisions.
This new crowd management technology solution aims to achieve:
- Reduced overcrowding and waiting times, leading to improved passenger satisfaction and fewer denied boardings.
- Improved service reliability and punctuality, by minimising bus bunching and better aligning vehicle deployment with real-time demand.
- Greater operational efficiency, through data-driven allocation of vehicles and optimised scheduling.
- Lower barriers to technology adoption, as edge processing removes the need for costly infrastructure upgrades and makes the system more accessible to small and medium-sized PTOs.
- Stronger public transport uptake, contributing to a modal shift away from private car use and supporting broader decarbonisation goals.
Scaling impact across Europe
As part of the BCMS project, the CMT solution will be tested on more than 20 buses across at least two lines in both Sofia and Bucharest, in close collaboration with local operators and project partners. The pilot phase is expected to validate the system’s usability, scalability, and operational impact in diverse network conditions.
The rollout of this next generation crowding management system promises to significantly contribute to the digital public transport transformation in Sofia as well as improve the attractiveness and efficiency of public bus services in other parts of Bulgaria, Romania and beyond.
[1] https://urban-mobility-observatory.transport.ec.europa.eu/resources/case-studies/sofias-sump-2019-2035-addressing-urban-mobility-challenges_en
