Answers are grounded in this paper and cite their sources.
Demonstrates a new simulator improves crack filling accuracy in pavement maintenance, suggesting enhanced safety for road users.
Automating pavement crack repair is vital for reducing worker exposure to unsafe working environments and increasing repair consistency, underpinning preventative maintenance to prolong the lifespan of roads. To ensure the safety of road users, particularly (motor-) cyclists, one key challenge is consistently achieving a smooth surface finish where the repaired crack surface is flush with the surrounding surface. Current literature falls short of an effective solution. In this paper, we propose a crack filling simulator using position-based fluids (PBFs), aiming at developing data-driven accurate filling capabilities. The proposed simulator features a random crack geometry generation algorithm, a novel quantitative repair accuracy metric and a benchmark library of state-of-the-art flush-fill process models, all integrated with a PBF simulation environment. We demonstrate the simulator usefulness through training a data-driven feedback-controlled flush-fill process using Bayesian optimisation, implementing it in the real world and comparing it with existing processes in simulation. The findings show that the proposed simulator can effectively support optimising crack filling processes with submillimeter-level average simulation errors in the repaired surface finish, which can be a pre-training step for repair robots. We believe that the simulation-informed crack filling approach will pave the way for autonomous pavement maintenance and other future infrastructure applications.
No takes yet. Share an insight, caveat, or question.
Schaefer et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: