Certification scheme for artificial intelligence based systems

Weng, Joanna; Denzel, Philipp; Reif, Monika Ulrike; Schilling, Frank-Peter; Billeter, Yann; Frischknecht-Gruber, Carmen; Brunner, Stefan; Chavarriaga, Ricardo; Repetto, Marco; Iranfar, Arman

doi:10.21256/zhaw-30549

Please use this identifier to cite or link to this item: https://doi.org/10.21256/zhaw-30549

Publication type:	Conference paper
Type of review:	Peer review (publication)
Title:	Certification scheme for artificial intelligence based systems
Authors:	Weng, Joanna Denzel, Philipp Reif, Monika Ulrike Schilling, Frank-Peter Billeter, Yann Frischknecht-Gruber, Carmen Brunner, Stefan Chavarriaga, Ricardo Repetto, Marco Iranfar, Arman
et. al:	No
DOI:	10.21256/zhaw-30549
Conference details:	34th European Safety and Reliability Conference (ESREL), Cracow, Poland, 23-27 June 2024
Issue Date:	Jun-2024
Publisher / Ed. Institution:	ZHAW Zürcher Hochschule für Angewandte Wissenschaften
Language:	English
Subjects:	AI; Explainability; Trustworthiness; Safety
Subject (DDC):	006: Special computer methods
Abstract:	Artificial Intelligence (AI) and machine learning (ML) algorithms are making an impact in an increasing number of industries. AI models differ from conventional software due to their probabilistic decision-making process, with a heavy reliance on training data quantity and quality. Ensuring the trustworthiness of AI-based systems (AIS), including dimensions such as reliability and transparency, is becoming increasingly vital due to their widespread adoption. As regulatory standards are put in place, it becomes essential to have practical guidelines for certification. In this paper, we present an ongoing effort to develop a validated Certification Scheme for AIS. This scheme encompasses distinct objectives, criteria, and corresponding measures, as well as specific metrics and technical methods which support the implementation of trustworthy AI. A critical aspect of this scheme is the establishment of a clear connection between the set of requirements and the validated ML algorithms and methods used to evaluate the compliance of AIS. We provide a tangible example of the workflow for the reliability dimension on a hypothetical real-life use case: employing the Yolo5 model for the detection of construction vehicles in a diverse image dataset of construction sites. This example demonstrates the step-by-step process of the Certification Scheme from establishing initial requirements to selecting and applying technical methods for two example objectives.
URI:	https://digitalcollection.zhaw.ch/handle/11475/30549
Fulltext version:	Accepted version
License (according to publishing contract):	Licence according to publishing contract
Departement:	School of Engineering
Organisational Unit:	Centre for Artificial Intelligence (CAI) Institute of Applied Mathematics and Physics (IAMP)
Published as part of the ZHAW project:	certAInty – A Certification Scheme for AI systems
Appears in collections:	Publikationen School of Engineering

Files in This Item:

File	Description	Size	Format
2024_Billeter-etal_Certification-scheme-for-AI-based-systems.pdf		1.48 MB	Adobe PDF	View/Open

Show full item record

Weng, J., Denzel, P., Reif, M. U., Schilling, F.-P., Billeter, Y., Frischknecht-Gruber, C., Brunner, S., Chavarriaga, R., Repetto, M., & Iranfar, A. (2024, June). Certification scheme for artificial intelligence based systems. 34th European Safety and Reliability Conference (ESREL), Cracow, Poland, 23-27 June 2024. https://doi.org/10.21256/zhaw-30549

Weng, J. et al. (2024) ‘Certification scheme for artificial intelligence based systems’, in 34th European Safety and Reliability Conference (ESREL), Cracow, Poland, 23-27 June 2024. ZHAW Zürcher Hochschule für Angewandte Wissenschaften. Available at: https://doi.org/10.21256/zhaw-30549.

J. Weng et al., “Certification scheme for artificial intelligence based systems,” in 34th European Safety and Reliability Conference (ESREL), Cracow, Poland, 23-27 June 2024, Jun. 2024. doi: 10.21256/zhaw-30549.

WENG, Joanna, Philipp DENZEL, Monika Ulrike REIF, Frank-Peter SCHILLING, Yann BILLETER, Carmen FRISCHKNECHT-GRUBER, Stefan BRUNNER, Ricardo CHAVARRIAGA, Marco REPETTO und Arman IRANFAR, 2024. Certification scheme for artificial intelligence based systems. In: 34th European Safety and Reliability Conference (ESREL), Cracow, Poland, 23-27 June 2024. Conference paper. ZHAW Zürcher Hochschule für Angewandte Wissenschaften. Juni 2024

Weng, Joanna, Philipp Denzel, Monika Ulrike Reif, Frank-Peter Schilling, Yann Billeter, Carmen Frischknecht-Gruber, Stefan Brunner, Ricardo Chavarriaga, Marco Repetto, and Arman Iranfar. 2024. “Certification Scheme for Artificial Intelligence Based Systems.” Conference paper. In 34th European Safety and Reliability Conference (ESREL), Cracow, Poland, 23-27 June 2024. ZHAW Zürcher Hochschule für Angewandte Wissenschaften. https://doi.org/10.21256/zhaw-30549.

Weng, Joanna, et al. “Certification Scheme for Artificial Intelligence Based Systems.” 34th European Safety and Reliability Conference (ESREL), Cracow, Poland, 23-27 June 2024, ZHAW Zürcher Hochschule für Angewandte Wissenschaften, 2024, https://doi.org/10.21256/zhaw-30549.