Assurance of Machine Learning-Based Aerospace Systems: Towards an Overarching Properties-Driven Approach

Paul, Saswata; Prince, Dan; Iyer, Naresh; Durling, Michael; Visnevski, Nikita; Meng, Baoluo

doi:10.21949/1528237

i

Assurance of Machine Learning-Based Aerospace Systems: Towards an Overarching Properties-Driven Approach

2023-09-01
By Paul, Saswata ; Prince, Dan ; Iyer, Naresh

File Language:

English

Details

Creators:

Paul, Saswata ; Prince, Dan ; Iyer, Naresh ; Durling, Michael ; Visnevski, Nikita ; Meng, Baoluo
Corporate Creators:

GE Research
Corporate Contributors:

United States. Department of Transportation. Federal Aviation Administration
Subject/TRT Terms:
Publication/ Report Number:

DOT/FAA/TC-23/54
DOI:

https://doi.org/10.21949/1528237
Resource Type:

Tech Report
Geographical Coverage:

United States
Edition:

Phase I Report
Corporate Publisher:

United States. Department of Transportation. Federal Aviation Administration. William J. Hughes Technical Center
Abstract:

Traditional process-based approaches of certifying aerospace digital systems are not sufficient to address the challenges associated with using Artificial Intelligence (AI) or Machine Learning (ML) techniques. To address this, agencies are evaluating an alternative Means of Compliance (MoC) called the Overarching Properties (OP). The goals for this research are to develop recommendations and assurance criteria and to explore safety risk mitigation approaches for such AI/ML-based software systems. This document outlines a novel foundation for the application of OPs to support the assurance and certification of complex aerospace digital systems consisting of AI/ML-based components. To this end, we first select the use case of a Recorder Independent Power Supply (RIPS) system. We then perform a Functional Hazard Assessment (FHA) to identify a set of hazards associated with the RIPS and design a set of appropriate requirements to mitigate those hazards.
Format:

PDF
Funding:

692M15-22-T-00012
Download URL:

https://rosap.ntl.bts.gov/view/dot/68923/dot_68923_DS1.pdf
File Type:

[PDF - 2.86 MB]
Collection(s):

Artificial Intelligence Deployment Center of Excellence (AI COE) ; US Transportation Collection
Main Document Checksum:

urn:sha-512:bc4d39e4dbf7d7ffa99a5533425f0724356f59a9cf3583327633ae33888dda529ef1f99af81954c867c02e8fa321a406648dd044d34d495173c11ac90458912b

File Language:

English

ON THIS PAGE

Details

Found an issue?

Send us an email at:

publishinghd@cdc.gov

ROSA P serves as an archival repository of USDOT-published products including scientific findings, journal articles, guidelines, recommendations, or other information authored or co-authored by USDOT or funded partners. As a repository, ROSA P retains documents in their original published format to ensure public access to scientific information.