SAMBA

Semi-autonomous marking and construction progress assistant

Duration: April 2026 – March 2028

ZIM – Central Innovation Programme for small and medium-sized enterprises (SMEs)

 

The aim of SAMBA is to create an integrated assistance system that, for the first time, combines autonomous site layout planning with construction progress monitoring. A collaborative system comprising a drone and a mobile walking robot collects data: The drone provides a semantic preliminary analysis of 2D aerial images for targeted robot deployment; the robot, equipped with hybrid sensor and actuator technology, traverses the construction site, sets BIM-compliant markers and records progress (camera, 3D scanner) for semi-automated plan verification. The data is linked to the digital construction plan using AI; visual language models enable precise segmentation, localisation and quantity measurement of the 3D scene, make deviations immediately visible and generate a digital twin in near real time. The system operates semi-autonomously: core tasks run automatically, whilst specialists monitor, trigger markings and validate AI results. AR/VR visualisation and an open interface and standards framework enable collaborative remote meetings, illustrate and archive agreements, distribute plans to devices, and ensure scalability, openness, future-proofing and long-term usability.

SAMBA project consortium

  • Hochschule für Angewandte Wissenschaften Hamburg
  • Fraunhofer Heinrich-Hertz-Institut
  • Flypix AI GmbH
  • LandPlan OS GmbH