Road occupancy prediction of traffic participants

Althoff, Matthias; Heß, Daniel; Gambert, Florian · 2013 · OpenAlex-citations

DOI: 10.1109/itsc.2013.6728217

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the challenge of predicting road occupancy for traffic participants to enable safe, automated collision avoidance systems. The authors argue that existing methods, such as Monte Carlo simulations or finite behavior predictions, are insufficient for rigorous safety certification because they cannot guarantee that all possible future positions are accounted for. Specifically, exact computation of occupancy sets is infeasible for complex dynamic models that enforce realistic constraints, such as road boundaries and speed limits. To solve this, the paper proposes a method to compute tight overapproximative occupancy sets—regions that enclose all reachable positions of other vehicles—allowing a trajectory planner to verify if emergency maneuvers are collision-free in all possible future scenarios. The approach utilizes a dynamic vehicle model defined by state-space equations incorporating position, velocity, and acceleration, subject to five specific constraints: maximum speed limits, prohibition of backward driving, engine power limitations, maximum absolute acceleration (tire friction), and road boundary adherence. Because the system is hybrid and complex, the authors employ reachability analysis on simplified abstractions of the model rather than simulating individual trajectories. They prove that intersecting the occupancy sets of these abstractions yields a valid overapproximation of the true occupancy. The computation is divided into two components: determining the upper bound of occupancy along the road path using monotone dynamics, and determining the lateral boundaries using three distinct methods. Method A considers only absolute acceleration limits, resulting in circular occupancy regions. Method B decouples longitudinal and lateral dynamics to strictly enforce road boundaries, using an axis-aligned box approximation for acceleration. Method C combines both approaches, using Method A for the initial phase of movement and Method B for subsequent phases to ensure tight bounds near road edges. Numerical examples demonstrate that the proposed occupancy computation is both efficient and sufficiently tight. The results show that the method accurately captures the boundaries of reachable positions, including complex maneuvers like lane changes, without including unrealistic behaviors such as leaving the road. The combined Method C provides a tighter approximation than either Method A or B alone, particularly in scenarios involving curved roads and lane changes. The authors illustrate that the computed sets allow the collision avoidance system to identify safe maneuvers almost until the last possible moment, maximizing the utility of the vehicle's dynamic capabilities while maintaining rigorous safety guarantees. The significance of this work lies in its contribution to the formal verification of autonomous driving systems. By providing a computationally efficient way to generate guaranteed-safe occupancy sets for complex vehicle dynamics, the method enables collision avoidance systems to be certified for safety. Unlike probabilistic approaches, this deterministic overapproximation ensures that no unsafe behavior is missed, addressing a critical gap in current autonomous vehicle research. The framework is also adaptable for collaborative emergency maneuvers using vehicle-to-vehicle communication, where known occupancies of cooperating vehicles can further refine the safety checks.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• anticipation