Certifiable Artificial Intelligence Through Data Fusion
This work tackles the problem of ensuring reliability and trust in AI systems for stakeholders like developers and regulators, but it appears incremental as it reviews existing concerns and proposes a notional approach without new empirical results.
The paper addresses the challenge of certifying AI systems by proposing the use of design and operational test evaluation procedures to determine performance bounds, illustrated through an image data fusion use case for object recognition with precision versus distance considerations.
This paper reviews and proposes concerns in adopting, fielding, and maintaining artificial intelligence (AI) systems. While the AI community has made rapid progress, there are challenges in certifying AI systems. Using procedures from design and operational test and evaluation, there are opportunities towards determining performance bounds to manage expectations of intended use. A notional use case is presented with image data fusion to support AI object recognition certifiability considering precision versus distance.