The second OptiPEx consortium meeting was held at Johannes Kepler University, Linz on May between May 7th and 8th, bringing together representatives from eight European countries. Over the course of two days, consortium members engaged in in-depth discussions to assess the current status of the project and to define the next steps toward achieving its goals.
At the end of the the second day, consortium members, joined by a group of university students, participated in a hands-on data collection activity at the Linz tram depot located in “Remise Kleinmünchen”. The activity aimed to capture live video footage from the tram’s internal CCTV system while participants were on board, simulating realistic passenger scenarios.
To ensure no disruption to public transportation services, the tram operated in a closed-loop route within the depot premises. Participants received detailed instructions indicating when to board the tram, which door to use, and specific actions to perform inside; such as sitting or standing in predefined positions. After each loop, the tram stopped, and participants were instructed to follow a new set of directives to vary the recorded scenarios.
Low-resolution video recordings from 13 CCTV cameras installed throughout the tram were collected during the activity. These recordings are currently undergoing a meticulous labelling process to identify and annotate the precise location and behaviour of passengers throughout the ride. The resulting dataset will be used to train and fine-tune AI models designed to estimate tram occupancy levels using only visual input from the existing CCTV infrastructure.
Crucially, these AI models are intended to run on the trams themselves on embedded hardware (edge computing), ensuring that passenger data is processed without transmitting sensitive information externally. This approach is essential to maintain user privacy and ensure full compliance with GDPR regulations while enabling smarter, more efficient public transport management.
The data collection event was a key milestone for the OptiPEx project, demonstrating the collaborative spirit of the consortium and the potential of AI-powered solutions for real-world mobility challenges.
