SPATRA Road Workstream: Field Test at Batrovci and Validation Workshop at NELT
The SPATRA project organized a two-day event in Serbia to test and validate its Road Workstream solutions. The first day featured a large-scale field test at the Batrovci border crossing, one of the busiest freight gateways in the Western Balkans. The second day continued with a validation workshop at NELT’s offices in Belgrade, where the evidence collected in the field was consolidated and reviewed. The event provided an opportunity to assess SPATRA technologies in real conditions, bringing together partners to evaluate the system’s ability to monitor traffic flows, predict waiting times, and support more efficient border operations.
Field Test at Batrovci
The Batrovci crossing was chosen because it represents a realistic challenge for freight operators. Long queues and limited parking capacity make border management complex, and these were exactly the conditions that the SPATRA system set out to measure and analyze.
The test involved a NELT truck departing from Dobanovci and traveling toward Batrovci. The truck was fitted with Galileo-based tracking devices that recorded its exact position along the journey. An observer in the cabin carefully maintained a manual log of coordinates, stops, and timing. This record acted as a ground-truth reference for evaluating the accuracy of the integrated system. At the same time, the SPATRA dashboard tracked the truck in real time, with screenshots taken at key moments to create an evidence trail for later validation.
Alongside the truck journey, drones were deployed to carry out two targeted missions. The first mission focused on scanning the parking lot at the border. Drone footage and computer vision algorithms produced counts of available and occupied spaces, which were checked against manual observations on the ground. The second mission concentrated on the length of the truck queue, with drones capturing both video footage and data points that were compared against coordinates of the last vehicle in line. This created a clear dataset to test how well the algorithms matched real conditions.
By the end of the first day, the team had created a broad and detailed collection of evidence. This included:
• Ground-truth logs from the NELT truck.
• Dashboard screenshots and positional data.
• Drone recordings and detection outputs.
• Extensive photos and videos documenting the operations.
This evidence formed the backbone of the validation effort and ensured that all perspectives, from the driver and system operator to the drone team and field observers, were represented.
Validation Workshop at NELT
On the second day, partners met at NELT’s offices to bring together the material gathered during the field test. The session began with a review of the activities at Batrovci and then moved into structured data consolidation. All datasets, including truck logs, dashboard records, drone detections, and visual evidence, were merged into a single repository. This allowed the team to carry out cross-comparisons and check for consistency.
The workshop marked the start of D4.2 Integration and Initial Validation. The first step was to confirm that data from different sources aligned correctly. Automated detections were compared against manual records, and any discrepancies were documented. Technical issues experienced during the test were discussed, along with potential improvements for future iterations.
The second part of the workshop looked ahead to the D4.3 Results and Analysis report. Partners discussed how the evidence would be mapped to project indicators and how to align with the rail workstream. Particular attention was given to validation templates, ensuring consistency across use cases, and to the project’s key performance indicators.
Some of the most important considerations included:
• Confirming that system outputs met technical KPIs defined in earlier stages of the project.
• Ensuring that the evidence collected would support project-level KPIs, such as the required number of validation reports and the end-user approval threshold.
• Defining how the integration and validation findings would be presented in upcoming deliverables.
The day concluded with agreement on next steps, including drafting the initial validation report and preparing the final structure for the comprehensive analysis.
Moving Forward
The two-day event demonstrated the value of combining different technologies into one operational system. The integration of Galileo tracking, drone-based detections, and satellite imagery into a single dashboard proved effective in providing real-time insights into border conditions. By fusing multiple data streams, the system was able to overcome the limitations of single-source monitoring and offer a more reliable picture of traffic flows.
The SPATRA Road Workstream will now continue with the preparation of validation reports. The immediate priority is completing D4.2 on integration and initial validation, followed by D4.3 on final results and analysis.
By building on the results of the Batrovci test and the workshop at NELT, SPATRA is advancing toward its goal of supporting predictable and efficient cross-border logistics operations.