Assuring the Safety of End-to-End Learning-Based Autonomous Driving through Runtime Monitoring

Grieser, Jörg; Zhang, Meng; Warnecke, Tim; Rausch, Andreas

doi:10.1109/dsd51259.2020.00081

Cited by 12 publications

(4 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Runtime enforcement goes beyond monitoring and attempts to enforce certain safety properties. Existing works [8,21,25,43] already propose a few methods for runtime enforcement of ADSs. AVGuardian [25] performs static analysis of the communication messages between the ADS modules to generate control policies and enforce them during runtime.…”

Section: Related Workmentioning

confidence: 99%

“…Shankaro et al [43] define a policy using an automaton and enforce the car to stop when the policy is violated. Grieser et al [21] build an end-to-end neural network (from LIDAR to torques/steering) that implicitly picks up safety rules. Simultaneously, the distance to obstacles on the current trajectory is monitored and emergency brakes are applied if a collision is likely.…”

Section: Related Workmentioning

confidence: 99%

“…This is the idea of runtime enforcement, a technique that observes the execution of a system and then modifies it in a minimal way to ensure certain properties are satisfied. In AVs, runtime enforcement has been applied, for example, to monitor basic safety properties such as the avoidance of collisions, applying the emergency brake before they are violated [21]. Avoiding collisions, however, is not enough in general.…”

mentioning

confidence: 99%

See 2 more Smart Citations

REDriver: Runtime Enforcement for Autonomous Vehicles

Sun,

Poskitt,

Zhang

et al. 2024

Proceedings of the IEEE/ACM 46th International Conference on Software Engineering

View full text Add to dashboard Cite

Autonomous driving systems (ADSs) integrate sensing, perception, drive control, and several other critical tasks in autonomous vehicles, motivating research into techniques for assessing their safety. While there are several approaches for testing and analysing them in high-fidelity simulators, ADSs may still encounter additional critical scenarios beyond those covered once they are deployed on real roads. An additional level of confidence can be established by monitoring and enforcing critical properties when the ADS is running. Existing work, however, is only able to monitor simple safety properties (e.g., avoidance of collisions) and is limited to blunt enforcement mechanisms such as hitting the emergency brakes. In this work, we propose REDriver, a general and modular approach to runtime enforcement, in which users can specify a broad range of properties (e.g., national traffic laws) in a specification language based on signal temporal logic (STL). REDriver monitors the planned trajectory of the ADS based on a quantitative semantics of STL, and uses a gradient-driven algorithm to repair the trajectory when a violation of the specification is likely. We implemented REDriver for two versions of Apollo (i.e., a popular ADS), and subjected it to a benchmark of violations of Chinese traffic laws. The results show that REDriver significantly improves Apollo's conformance to the specification with minimal overhead.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

REDriver: Runtime Enforcement for Autonomous Vehicles

Sun,

Poskitt,

Zhang

et al. 2024

Proceedings of the IEEE/ACM 46th International Conference on Software Engineering

View full text Add to dashboard Cite

show abstract

“…Regarding vehicle traversability (or vehicle going) safety, a study by [5] implemented a safety solution to avoid collisions (the authors referred to their approach as Simplex-Drive).…”

Section: The Contribution Of the Study Was The Fusion Of Informationmentioning

confidence: 99%

Predicting Autonomous Vehicle Navigation Parameters via Image and Image-and-Point Cloud Fusion-based End-to-End Methods

Beyçimen

Ignatyev

Zolotas

2022

2022 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI)

View full text Add to dashboard Cite

This paper presents a study of end-to-end methods for predicting autonomous vehicle navigation parameters. Image-based and Image & Lidar points-based end-to-end models have been trained under Nvidia learning architectures as well as Densenet-169, Resnet-152 and Inception-v4. Various learning parameters for autonomous vehicle navigation, input models and pre-processing data algorithms i.e. image cropping, noise removing, semantic segmentation for image data have been investigated and tested. The best ones, from the rigorous investigation, are selected for the main framework of the study. Results reveal that the Nvidia architecture trained Image & Lidar points-based method offers the better results accuracy rate-wise for steering angle and speed.

show abstract