Difference between revisions of "AI Safety Aviation"

From MIT Technology Roadmapping
Jump to navigation Jump to search
Line 11: Line 11:
[[File:AVIAIDecomp.png | 600px | Decomposition of 3ADC]]
[[File:AVIAIDecomp.png | 600px | Decomposition of 3ADC]]


[[File:AVIAIDSM.png | 600px | DSM of 3ADC]]
[[File:AVIAIDSM.png | 400px | DSM of 3ADC]]


= Roadmap OPD =
= Roadmap OPD =

Revision as of 06:07, 5 November 2024

Roadmap Overview

Focuses on the integration of artificial intelligence, particularly machine learning, into safety-critical aviation systems while maintaining the highest standards of safety and trustworthiness. It encompasses a phased approach, starting with human assistance/augmentation (Level 1 AI), progressing to human-AI teaming (Level 2 AI), and ultimately advancing to more automated solutions (Level 3 AI). Key aspects include developing AI assurance frameworks, addressing human factors in AI integration, implementing safety risk mitigation strategies, and ensuring compliance with evolving regulatory requirements. This technology aims to enhance aviation safety, efficiency, and performance across various domains including aircraft design and operation, air traffic management, and unmanned aircraft systems.

Scope Down Idea (Draft)

  • Hybrid cockpit with AI supporting pilot by improving situational awareness and translate pilot intent to automated systems.

DSM Allocation

Decomposition of 3ADC

DSM of 3ADC

Roadmap OPD

OPD


Figures of Merit

Important FOMs for 3AISC
Figure of Merit Description Trends Units
Incident Rate Rate of incidents and accidents normalized by flight time aircraft decreasing

incidents passenger mile

System Maintainability How much time is spend maintaining aircraft systems, measured by comparing maintenance time per flight time decreasing

maintenance hours passenger mile

Response Time Time required to identify and mitigate safety issues flat

minutes

System Uptime Amount of time aircraft is available to be dispatched on a mission as a percentage of wall time increasing

%

Cognitive Load on Crew As measured by industry standard NASA TLX score flat

unitless

Detail of Accidents per passenger hour

Normalized Accidents per Year, United States General Aviation Fleet