At FOSS4G 2019 in Bucharest, our team presented the talk “From paper to pods: Revolutionised fibre planning process at Deutsche Telekom AG with FOSS4G components.” The presentation was given by Torsten Drey, with Emmanuel Belo, Marc Jansen, and Markus Neteler as co-authors. I was present in the room to support the discussion and help answer questions around the project and its implementation.
The talk addressed a concrete industrial challenge: how Deutsche Telekom AG could accelerate fibre rollout in Germany by transforming a planning process that was still strongly dependent on manual work, fragmented data, and paper-based handovers. Germany needed to move faster on broadband and FTTH deployment, but construction can only start once planning is complete. That meant the IT systems, geodata workflows, planning tools, and automation pipelines had to be ready well before the rollout targets.
Our work focused on building a spatial data infrastructure able to support this transformation. Fibre planning requires a large amount of geospatial data, including 2D and 3D information. Existing fixed-network data had to be corrected and conflated, while new sources such as terrestrial LiDAR, panoramic imagery, aerial orthophotos, cadastral data, pedestrian data, and OpenStreetMap data were brought into the process.
A key use case was identifying potential trench routes. The cost of laying fibre depends heavily on surface type: grass or open land is cheaper than asphalt or concrete. By classifying surfaces and generating topological line networks, the system could help planners identify more cost-efficient routes for optical fibre cables.
The architecture combined open source geospatial components with cloud-native deployment practices. Components such as geOrchestra, SHOGun, and Actinia were used to provide web mapping, data services, processing services, and user interfaces. The platform also included 2D WebGIS tools and linked 3D/panorama views, allowing planners to inspect street-level imagery and support decisions such as cabinet placement.
Behind the geospatial applications, the project relied on automated infrastructure and delivery pipelines. Using Deutsche Telekom’s cloud environment, OpenStack APIs, OpenShift, GitLab, infrastructure as code, CI/CD, container images, automated testing, Grafana, and Kibana, the team created a scalable and maintainable deployment model across development, integration, staging, and production environments.
One important lesson from the project was that fibre rollout is not only a construction problem. It is also a data, software, automation, and collaboration problem. The project brought together Deutsche Telekom teams and external open source experts, working in agile cycles with Scrum and SAFe.
For me, being part of this work and present during the FOSS4G session was a strong example of how open source geospatial software can support large-scale industrial goals when combined with solid engineering practices and multidisciplinary collaboration.