Artificial Intelligence/Machine Learning on Safety Critical Systems
Risk Profile - What needs to be considered in a safety critical application
Verification and Validation - Statistical nature of AI/ML * Simplification of the code * Human in the loop
Data acquisition * Space One offs * What data is used for training?
NASA Software Engineering Requirements interpretation for AI/ML - Walk through some of the key interpretations: Artificial Intelligence And Software Engineering
Mr. Scott Tashakkor is currently serving as the Acting NASA Technical Fellow for Software. In this role, he works as the technical authority for software to the agency and sets software engineering requirements. In his career he has focused on flight and ground software development, security, verification, and validation. He started out working computational models for launch vehicles and the associated system components. From there he moved into writing the embedded flight software for Space Launch System (SLS) and other systems. Mr. Tashakkor has an undergraduate degree in Aerospace Engineering from the University of Illinois Urbana-Champaign, and has two master’s degrees, Aerospace and Computer Engineering.
Thu 6 AugDisplayed time zone: Pacific Time (US & Canada) change
10:20 - 11:50 | |||
10:20 25mTalk | Artificial Intelligence/Machine Learning on Safety Critical Systems Resilient AI Scott Tashakkor NASA | ||
10:45 25mTalk | From Algorithms to Flight Software: Resilient AI Applications for Onboard Science Instrument Autonomy Resilient AI Jack Lightholder Jet Propulsion Laboratory, California Institute of Technology | ||
11:10 25mTalk | AI That Knows How to Come Home: Preventing Hidden Alignment Drift in Space Systems Resilient AI | ||
11:35 15mTalk | Open Discussion Resilient AI | ||