Transformer-based Monte Carlo Localization in Construction Meshes

2026-09-25

AI Summary

This study addresses the challenge of robot localization in construction sites, where weak textures and repetitive layouts degrade conventional methods. We propose a Monte Carlo Localization (MCL) system trained exclusively on synthetic LiDAR data. Methodologically, the framework employs a PointNet++ encoder coupled with an uncertainty-aware Transformer decoder to achieve global relocalization. Furthermore, we introduce an uncertainty-scaled likelihood function and a hypothesis-injection resampling strategy to effectively mitigate particle depletion. Experimental results demonstrate that the proposed system outperforms existing baselines in real-world scenarios while maintaining an inference latency of only 18 milliseconds. These findings validate both the feasibility and robustness of deploying models trained purely on synthetic data for practical robotic localization tasks in challenging environments.

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.

Problem

• Monte Carlo Localization
• construction robotics
• global relocalization
• LiDAR
• building mesh

Innovation

• Monte Carlo Localization
• PointNet++
• Synthetic Data Training
• Uncertainty-aware Decoder
• Global Relocalization