Trusted Human-Autonomy Teaming in Teleoperations

Funding: DST Group and Australian Army

Scheme: Human Performance Research Network (HPRNET)

The primary aim of this project is to develop methodologies for improving, and metrics for measuring, mission effectiveness of human-machine teams.

Decoupling the operator from the operating platform reduces risks for the operator and reduces cost, where a remote operator can operate multiple assets simultaneously. Teleoperation, especially in a war-fighting context, comes with its own challenges. Teleoperation removes soldiers from danger, but mission effectiveness can be limited if we do not understand operators' cognitive performance in these settings, and how it relates to combat effectiveness. Cognitive performance may be impacted by fatigue, multi-tasking, switching from the civilian to the war-fighting setting, uncertainty about the uncontrolled remote environment or surprise that increases workload (e.g. operator needing to integrate information rapidly from limited sensors).

With these challenges in mind, the primary aim of the project is distilled into four sub-questions:

  1. Using wearable technologies, how can we automatically recognise human activities during teleoperation of autonomous systems? This recognition task facilitates intent management and the development of mutual predictability. Examples of sensors we will consider for this sub-question include:  Kinect, EEG, ECG, physiological, speech, and IMUs. 
  2. What is an appropriate set of cognitive and behavioural indicators for humans, and task-complexity indicators for the task, to enable efficient monitoring of these indicators in order to load-balance the distribution of sub-tasks among different human and non-human actors? 
  3. What is an appropriate methodology for real-time load balancing between humans and autonomous entities in remotely supervised tasks? A closed-loop system will be developed to allow cognitive indicators from humans and task indicators from the overall system (including autonomous entities) to be used to ensure appropriate load balance between the human and the machine to maximise the effectiveness of the mission. 

How can we assure trustworthiness in a team made of humans and autonomous systems during teleoperation? Trust assurance is necessary to ensure the effectiveness of teleoperation missions. 

External Links: http://www.dst.defence.gov.au/hprnet

Return to Trusted Autonomy >>