Trusted Autonomy (TA) is an emerging field of research that focuses on understanding and designing the interaction space between entities, each of which exhibits a level of autonomy. These entities can be humans, computer controlled machines, or a mix of the two. The aim of the TA group is to create a trusted, cooperative, seamless, symbiotic, and organic team of humans and machines to solve complex problems in an uncontrolled real-world uncertainty-rich environment.
Our group has expertise in traditional machine learning, navigation and control of autonomous vehicles, developmental robotics, computational motivation and computational red teaming. In this respect, our group is unique in Australia because of its mix of expertise, and its ability to innovate a concept and take it from an idea all the way through science and engineering to a technological solution.
The group currently consists of 6 academics, 7 postdoctoral researchers and over 20 research students. We also have a number of other collaborating academics within UNSW who make a strong contribution to the group. In the past five years, the group has been granted over $2.5 million in external research funding and produced over 200 publications in collaboration with associated academics, research staff and students. Group members currently hold (as of late 2016) 4 ARC Discovery projects and a number of Defence and industry funded projects.
Our research program is structured in four broad areas listed below:
Program 1: Computational Intelligence for Trusted Autonomy: Modular neural and neurofuzzy ensembles; Neural controllers in environments with high level of uncertainty and conflicting objectives; Deep neural networks for sensor to effectors modelling.
Program 2: Human-Machine Interaction: Understanding, modelling and shaping trust in humans and machines; User profiling in complex uncertain environments using internal (trust, motivation, power) models and data driven (EEG, physiological data, and others) models; Principles of experimental design for in-situ environments.
Program 3: Data and Decision Analytics: Big data analytics; Autonomous optimization and reasoning; Computational red teaming.
Program 4: Unmanned Vehicles: Ground control of unmanned vehicles; On-board autonomous control embedded systems; On-board sensing and fusion models.
More information can be found at the group’s external webpage which be found at http://trustedautonomy.net
CURRENTLY FUNDED RESEARCH PROJECTS
User-Task Adaption for effective interactive simulation environments (ARC Discovery Project 2016-2020)
Challenging Systems to discover vulnerabilities using computational red teaming (ARC Discovery Project 2014-2016)”
Reactive Planning under Disruptions and Dynamic Changes (ARC Discovery Project 2017-2019))”
Robust Configuration of Evolutionary Algorithms (ARC Discovery Project 2015-2017)
Trusted Human-Autonomy Teaming in Teleoperations (DST Group /Australian Army)