Transformer-based localization improves robot positioning on construction sites
Transformer-based Monte Carlo Localization in Construction Meshes
Robotics
Summary
Robots working on construction sites need to know exactly where they are within a building to do their jobs safely. The authors created a new method that uses laser scans combined with a type of AI model called a Transformer to help robots figure out their position even when buildings look very similar inside. They trained their system only using computer-generated data but tested it successfully on real buildings, showing it works well and quickly. This approach helps robots avoid getting confused when places look alike and can recover from mistakes during location tracking.
What this means in practice
- •For construction robotics teams: Use this localization method to improve building-scale robot positioning by accurately relating laser scans to construction site meshes, even in low-texture and repetitive layouts.
- •For indoor mapping technology developers: Develop enhanced indoor positioning systems by integrating transformer-based observation models trained purely on synthetic data for robust global localization.
Authors
Linus Kramer, William Talbot, Olga Vysotska, Marco Hutter
Abstract
To be able to perform inspection or digitization tasks, mobile robots on construction sites must be able to localize themselves reliably with respect to a global reference frame that is shared with a building map. Similar room layouts and low-texture surfaces pose a challenge for existing LiDAR- and vision-based localization methods. We approach this problem with a LiDAR-based global relocalization system that estimates the robot's pose relative to a building mesh and combines a PointNet++ encoder with a place recognition decoder, whose outputs serve as a learned observation model within a Monte Carlo Localization (MCL) framework. The pipeline is trained exclusively on synthetic LiDAR scans obtained by simulating the robot's sensors inside the building mesh. Our approach is robust in ambiguous environments due to an uncertainty-aware decoder that scales positional likelihoods and a resampling strategy that injects model hypotheses into the particle set, enabling recovery from potential particle depletion. Evaluations on real-world datasets show that our method outperforms both diffusion-based and ScanContext++ baselines while maintaining fast inference (18 ms per call), demonstrating the practicality of synthetic-data training for mesh-referenced global localization in construction robotics.