Highly Automated Driving on Highways Based on Legal Safety

Vanholme, Benoit; Gruyer, Dominique; Lusetti, Benoit; Glaser, Sébastien; Mammar, Saïd · 2013 · Crossref

DOI: 10.1109/tits.2012.2225104

archive: archived pipeline: cataloged verified

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

Summary

This dissertation addresses the challenge of enabling highly automated driving on highways by introducing the concept of "legal safety." The research is motivated by the need to integrate automated vehicles into mixed traffic with human drivers without requiring modifications to existing infrastructure or other vehicles. The core problem is ensuring safety in an environment where the automated system cannot rely on perfect communication with other road users. The author proposes that driving systems should base their decision-making on traffic rules, allowing them to predict the legal trajectories of other objects and handle worst-case scenarios defensively. The methodology involves a systematic design process following the V-cycle, moving from system requirements to component design, implementation, and validation. The legal safety framework is built upon three sets of rules: traffic rules for interacting with the environment, human rules for interacting with the driver (based on the HAVEit project’s horse-rider metaphor), and system rules for internal component coordination. The central contribution is the detailed design of the decision component, which generates and evaluates subject trajectories. This component uses a zone model to predict object trajectories based on traffic rules and generates multiple candidate trajectories for the subject vehicle, considering friction limits, human comfort limits, system capabilities, speed limits, and interactions with other objects. The system selects the optimal trajectory for normal operation and a fallback trajectory for system failures. The legal safety system was implemented on both personal computers and automotive Electronic Control Units (ECUs). It was validated through integration and testing on three demonstrators: the LIVIC Legal Safety Demonstrator, the HAVEit Joint System and Architecture Migration Demonstrators, and the ABV Low Speed Demonstrator. Validation scenarios included various highway driving situations. The results demonstrated that the decision, control, and Human-Machine Interface components could be achieved with state-of-the-art technology. The system successfully maintained safety by adhering to traffic rules and defensive driving principles, ensuring that accidents are avoided if all objects respect traffic rules, and mitigating risks if they do not. The significance of this work lies in providing a viable path for highly automated driving on highways under current international law, which permits high automation but not full automation. By relying on legal safety, the system ensures compatibility with human-driven traffic without needing cooperative infrastructure. The dissertation concludes that while decision, control, and HMI components are currently feasible, legal safety perception technology is expected to become available in the medium term. This approach facilitates the gradual introduction of automated driving systems, enhancing transport safety, comfort, and environmental friendliness while maintaining legal compliance and operational safety in complex, mixed-traffic environments.

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-nvidia on 2026-08-10; verification: verified.

Topics

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

• situational awareness
• driverless ads
• automation surprise
• lane positioning

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Methodological Resource: tool software
• Theoretical Contribution: computational model, conceptual framework