The project focused on developing a low‑cost, mobile‑phone‑based traffic analyzer that counts and classifies passing vehicles using acoustic signal processing. The core algorithm was implemented in a cloud‑hosted application and validated against a reference measurement system. In the validation campaign, the system correctly identified 834 cars, 63 trucks, and 8 motorcycles, yielding a total of 905 vehicles, which matched the real‑time counts recorded by the reference equipment. No significant difference was found between analyses based on spectrograms or third‑level frequency bands, indicating that either approach can be used for vehicle detection. While the system can classify vehicle types, speed estimation could not be evaluated due to missing reference speed data. Planned extensions such as unique vehicle signatures or sensor‑network traffic flow analysis were not realized within the 12‑month timeframe. Additional environmental data (rain, wind) were deemed unnecessary for the application’s performance.

Software development followed an agile methodology at Tigerbytes. Regular bug‑report reviews and performance tuning sessions were held, with daily meetings between product designers and developers to align on progress and system optimisation. The application’s data transfer and security were continuously refined, leveraging a cloud workspace for data exchange and tools such as Confluence, Git, Jira, and Miro for documentation, version control, task management, and collaborative design. The algorithm’s noise‑analysis capability achieved an accuracy of approximately 0.5 dB for event detection of average traffic noise levels, though it does not yet satisfy the stringent requirements of the German Federal Office for Traffic Safety (BASt) for traffic detection. Future work aims to evolve the prototype into a commercial product that can be integrated into the dBEL platform as a digital service, with potential applications in smart‑city and e‑health contexts.

Collaboration involved two German companies: Wölfel Engineering GmbH + Co.KG and Tigerbytes GmbH. Wölfel supplied hardware components and reference measurement equipment, while Tigerbytes provided the software stack. The project was planned for 12 months and executed in close coordination, with weekly Microsoft Teams meetings among project leaders and developers. Data sharing occurred primarily through the shared cloud workspace and email. The project received external funding, as indicated by references to “Förderdung,” though the specific grantor is not named in the report. The final report was submitted to the Technical Information Library (TIB) in Hannover, and the results were presented at several conferences, including the dBEL_trafficanalyzer kick‑off in 2022, a closing event in 2023, Couchbase Connect 2023, the German Acoustic Conference (DAGA) 2024, and the AI & Big Data Expo 2024. Wölfel Engineering also plans to disseminate findings through seminars and webinars. The project’s digital infrastructure—a cloud workspace and data processing pipeline for standard mobile phones—positions it to reach a broad user base quickly, given the widespread use of such devices. The experience and lessons learned will be shared through case studies and presentations, supporting further innovation in data‑driven traffic monitoring solutions.