School of Engineering and Information Technology
University of New South Wales - Canberra - Australia
Trusted Autonomy research focuses on the design of algorithms and methodologies to simulate, optimize, data mine and implement systems to make their interaction with other systems and humans happens naturally including computational models to analyse and understand this interaction.
Trusted Autonomy refers to two or more interacting and self-governed autonomous intelligent systems (including humans) where one side of the interaction is willing to delegate a task that will make it vulnerable to other parties in the interaction who are willing to accept and can autonomously perform the task.
Trusted Autonomy is when an autonomous agent becomes willingly vulnerable by delegating a task to itself or to another autonomous agent.
Autonomy is the freedom to make decisions subject to, and sometimes despite of, the constraints in the environment according to internal laws and values governing the autonomous agent.
Trust is a social contract between two agents. The truster delegates a task to the trustee while carrying the risk that the trustee might beuntrustworthy. The trustee accepts the task while, implicitly or explicitly, promising of being trustworthy.