At present, athletes and soldiers are monitored under special laboratory conditions, have data returned to base after several days of logging, and may fill out surveys to determine their own performance. Coaches and commanders want real-time feedback of performance: the challenge is to interpret, transmit and display the information.

Data processing

Smart processing algorithms combine the sensor data to detect patterns such as fatigue. Creating such algorithms is the task of the Machine Learning sub-group. The data from sensors is combined to determine measures of fatigue and group-wide (macro) performance using distributed and heterogenous inference algorithms.

At the coach's end, the information must be understandable and useful in assisting decisions. This requires prioritizing based upon what is happening and the user's needs.

Individual analysis

Smart algorithms from multiple sensors operating on the user -- such as accelerometers and heart-rate sensors. Producing useful measures of performance including logging of activity and fatigue estimates. We are currently working with the Australian Institute of Sport swimming team to develop activity monitoring systems.   More information here

Team & Group analysis

How do groups of individuals interact - eg. in teams on on a sporting field? This component of the project will identify patterns and performance data based on the movement of teams and groups of individuals.

 

Cognitive performance

In order for a coach to provide useful feedback to an athlete, the coach must know how the future performance of the athlete can be improved without impacting on the current performance: standing on the sideline and issuing commands throughout the game will only be a distraction. This component will find optimal strategies to achieve performance improvement.

Transmit

The meaningful information is transmitted back to end-users wirelessly, in a time frame that ensures the data is up-to-the-minute. Our expertise in wireless signal processing, radio engineering and network design will be combined to produce programmable, software-defined local networks operating on ultra-low power. More details are available on the wireless sub-project