Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

SMIRK : A machine learning-based pedestrian automatic emergency braking system with a complete safety case

Socha, Kasper ; Borg, Markus LU and Henriksson, Jens (2022) In Software Impacts 13.
Abstract

SMIRK is a pedestrian automatic emergency braking system that facilitates research on safety-critical systems embedding machine learning components. As a fully transparent driver-assistance system, SMIRK can support future research on trustworthy AI systems, e.g., verification & validation, requirements engineering, and testing. SMIRK is implemented for the simulator ESI Pro-SiVIC with core components including a radar sensor, a mono camera, a YOLOv5 model, and an anomaly detector. ISO/PAS 21448 SOTIF guided the development, and we present a complete safety case for a restricted ODD using the AMLAS methodology. Finally, all training data used to train the perception system is publicly available.

Please use this url to cite or link to this publication:
author
; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Advanced driver-assistance system, Automotive demonstrator, Computer vision, Machine learning, Pedestrian automatic emergency braking, Safety case
in
Software Impacts
volume
13
article number
100352
publisher
Elsevier
external identifiers
  • scopus:85134628533
ISSN
2665-9638
DOI
10.1016/j.simpa.2022.100352
language
English
LU publication?
yes
additional info
Publisher Copyright: © 2022 The Authors
id
c7b33f7b-b69f-483d-880c-0f4a8ea7cb4d
date added to LUP
2022-09-02 09:10:09
date last changed
2023-11-20 09:30:31
@article{c7b33f7b-b69f-483d-880c-0f4a8ea7cb4d,
  abstract     = {{<p>SMIRK is a pedestrian automatic emergency braking system that facilitates research on safety-critical systems embedding machine learning components. As a fully transparent driver-assistance system, SMIRK can support future research on trustworthy AI systems, e.g., verification &amp; validation, requirements engineering, and testing. SMIRK is implemented for the simulator ESI Pro-SiVIC with core components including a radar sensor, a mono camera, a YOLOv5 model, and an anomaly detector. ISO/PAS 21448 SOTIF guided the development, and we present a complete safety case for a restricted ODD using the AMLAS methodology. Finally, all training data used to train the perception system is publicly available.</p>}},
  author       = {{Socha, Kasper and Borg, Markus and Henriksson, Jens}},
  issn         = {{2665-9638}},
  keywords     = {{Advanced driver-assistance system; Automotive demonstrator; Computer vision; Machine learning; Pedestrian automatic emergency braking; Safety case}},
  language     = {{eng}},
  month        = {{08}},
  publisher    = {{Elsevier}},
  series       = {{Software Impacts}},
  title        = {{SMIRK : A machine learning-based pedestrian automatic emergency braking system with a complete safety case}},
  url          = {{http://dx.doi.org/10.1016/j.simpa.2022.100352}},
  doi          = {{10.1016/j.simpa.2022.100352}},
  volume       = {{13}},
  year         = {{2022}},
}