Simulation-Based Elicitation of Accuracy Requirements for the Environmental Perception of Autonomous Vehicles

Philipp, Robin; Qian, Hedan; Hartjen, Lukas; Schuldt, Fabian; Howar, Falk

doi:10.1007/978-3-030-89159-6_9

Cited by 6 publications

(2 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similar concepts have also been applied to study the sensitivity of a specific planner to perturbed perception results [33], [113]. Alternatively, this process can be used to derive acceptable perception perturbations for a single scenario [70]. However, this approach is limited by the availability of the planner [103].…”

Section: Safety Aware Metricsmentioning

confidence: 99%

See 1 more Smart Citation

SHARD: Safety and Human Performance Analysis for Requirements in Detection

Mori,

Peters

2024

IEEE Trans. Intell. Veh.

View full text Add to dashboard Cite

Automated driving requires reliable perception of the environment to ensure the safety of the driving task. One common perception task is 3D object detection, which aims at perceiving location and attributes of dynamic objects. This task is typically evaluated on different benchmark datasets, which each propose different metrics. However, these different metrics generally lack consistency and bear no relation to safety. Most notably, there is a lack of consistent definitions of pass/fail criteria for any given detection metric. In this work, the issue is addressed by systematically considering safety and human performance across different aspects of the object detection task. This approach yields interpretable detection metrics as well as thresholds for pass/fail criteria. Furthermore, a validation approach leveraging a prediction network is introduced and successfully applied to the requirements. A comparison of existing detectors shows that current perception algorithms exhibit failures for a majority of objects on the nuScenes dataset. Therefore, the results indicate the necessity of explicit safety consideration in the development of perception algorithms for the automated driving task.

show abstract

Section: Safety Aware Metricsmentioning

confidence: 99%

“…Furthermore, it suffers from ambiguities and challenges present in the planning task [31]. In addition, the validity is limited to the specific implementation of the planner [70]. While safety is considered in these metrics, no generally applicable pass/fail criteria for perception are identified.…”

Section: Safety Aware Metricsmentioning

confidence: 99%

SHARD: Safety and Human Performance Analysis for Requirements in Detection

Mori,

Peters

2024

IEEE Trans. Intell. Veh.

View full text Add to dashboard Cite

show abstract

Formal Methods for a Digital Industry

Howar

Hungar

Rausch

2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Integrating scenario- and contract-based verification for automated vessels

Hake,

Reiher,

Mentjes

et al. 2024

J Mar Sci Technol

View full text Add to dashboard Cite

Scenario-based verification defines the current state of the art for examining a vessel’s control systems for reliability and safety. However, software updates after release can only be covered to a limited extent. To take changes to a deployed system into account, the design and test phase must be harmonized with the operational phase. For all phases, regulatory, technical and safety requirements provide the scope to which the development process and the scenario-based tests need to be aligned and whose specifications the System under Test (SuT) must adhere to during operation. For this reason, a procedure is needed that converts the requirements into a format that can be utilized across all phases and measured in a structured manner comparing the original system to the updated version. This work does so by combining scenario-based verification methods with formal composition and monitoring techniques based on contract-based design into an integrated development approach. It is shown how safety requirements can be transferred into a Verification Descriptor that in turn provides the foundation for the division into model-based system development, contract-based virtual integration testing and a scenario-based test environment. For the entire lifecycle of the System under Test (SuT) to be included, the extended scenario and contract descriptors are carried forward up to the operational phase, such that the previously defined properties of the SuT can be monitored and validated during runtime. The approach is designed alongside a minimal-viable system and evaluated on an actual implementation of a safety-critical maritime LiDAR-based positioning system.

show abstract

Simulation-Based Elicitation of Accuracy Requirements for the Environmental Perception of Autonomous Vehicles

Cited by 6 publications

References 27 publications

SHARD: Safety and Human Performance Analysis for Requirements in Detection

SHARD: Safety and Human Performance Analysis for Requirements in Detection

Formal Methods for a Digital Industry

Integrating scenario- and contract-based verification for automated vessels

Contact Info

Product

Resources

About