A new paper on “Specifying for Trustworthiness” in Autonomous Systems has been published in Communications of the ACM (CACM).
Key findings include:
Autonomous Systems are increasingly becoming part of our daily lives. To demonstrate the trustworthiness of an autonomous system, we must first specify what is considered trustworthy.
This article looks across the range of autonomous- systems domains and identifies some of their key specification challenges.
Key intellectual challenges involved with specifying for trustworthiness in autonomous systems cut across the domains and are aggravated by the uncertainty in which the autonomous systems must operate.