Speaker
Description
Maintaining high-precision survey and alignment software at accelerator facilities is challenging due to small teams, regular staff turnover, and incoming engineers with limited domain or programming experience. At CERN, years of fragmented development had produced a heterogeneous survey software landscape, with several applications abandoned after their original developers left. The creation of a dedicated Survey Software section in 2019 established long-term ownership, combining experienced surveying engineers with IT graduates on two-to-three-year rotations.
The team modernised and rationalised its software ecosystem by retiring obsolete applications, consolidating functionality around shared and validated computational libraries, and replacing legacy programming languages with a reduced, maintainable technology stack centred on C++, Python, and modern frameworks. Moreover, standardised data interfaces ensure interoperability between tools, and with CERN’s survey database through a dedicated API. Quality assurance practices—including version control, continuous integration, testing, and documentation—were introduced to improve reliability, reproducibility, and user confidence. SurveyPad, a unified delivery platform, provides access to desktop applications, scripts, and web services built on this common foundation.
The benefits have been substantial. The shared architecture has lowered the barrier to contribution, enabling students and non-specialist engineers to develop and deploy more than ten production tools, including applications for publishing survey results to CERN’s Geographic Information System (GIS) and automated measurement quality control. Open workflows, comprehensive documentation, and structured onboarding have strengthened transparency, facilitated knowledge transfer, and improved long-term maintainability.
Our experience shows that sustainable survey software depends not only on technical coherence - a reduced stack, standardised interfaces, shared libraries - but also on organisational discipline: clear ownership, documentation, and knowledge.
