Health Care and Sports Performance
Wireless sensor networks are being used to provide patient information in health care applications, and to provide real-time performance feedback to athletes and coaches.
Patient Tracking
Indoor localisation and movement tracking of residents in aged care facilities offers opportunities for improved care of elderly patients - in particular those who suffer from dementia and other wandering behavioural symptoms. Aged care residents with dementia who tend to wander away from the facility or intrude on other patients present a danger to themselves and other residents. The ability to determine a patient's location and current activity (i.e. standing, sitting, walking, or falling) will help to improve their safety as well as easing the health care worker's task in managing them. Monitoring motion activity allows for a patient's behavioural patterns to be understood, thus enabling improved and prioritised/individualised care towards residents with severe or unusual wandering problems.
- Our Fleck Nano platform is ideal for a mobile node due to its small form factor. It includes an onboard integrated microcontroller and wireless transceiver, and an accelerometer for measuring the inertial movement of a patient.
In conjunction with the Australian e-Health Research Centre, we have developed an patient localisation and activity monitoring system which uses a combination of radio frequency (RF) tracking and inertial sensors. Information from the inertial sensors is incorporated into a human activity model to determine the characteristics of a patient’s activity such as running, walking, postural transition, etc.
Our localisation network consists of static nodes placed at predetermined locations in a building. These nodes are based on our Fleck™ wireless sensor network platform, and are used to determine the presence of the user within a particular region of a building.
Patients or health care workers carry a mobile node which incorporates onboard motion sensors that detect their current heading and direction. The mobile node is based on our Fleck Nano platform, which has a form factor small enough to enable it to be worn unobtrusively in a device such as a wrist watch. The mobile node detects a person's footstep by measuring the acceleration generated by the heel strike. This in turn allows their stride length, walking speed and displacement distance to be measured. The mobile node continuously streams inertial and heading information to a base node via the static nodes deployed throughout the building.
The user's position is determined with a dynamic tracking model which uses a particle filter to combine information from the mobile node's movement information (heading), proximity information from the nearest static node, and the building’s floor-plan. We have demonstrated the position resolution of our localisation network to have a maximum error between 1m and 3.5m. The patient's currently location and activity can be viewed via a computer or mobile phone interface.
We are in the process of evaluating our localisation network and activity monitoring system for the purpose of tracking and studying behavioural patterns of residents with dementia in an aged care facility.
Improving Sports Performance
- Multiple wearable sensors gather performance data to provide real-time feedback to the athlete and coach.
Athletes and coaches are often working in the field where real time interaction and information is essential to athlete development. The athlete and coach are often separated which makes both measurement and communication difficult. Furthermore it is not always possible for specialist staff to routinely join field based training sessions and simultaneously access and comment on immediate and historical data.
ICT Centre scientists are working with sports scientists and elite-level athletes at the Australian Institute of Sport as well as colleagues from CSIRO's Materials Science and Engineering Division to develop new sensors, data-acquisition applications on mobile platforms, and database systems to store and analyse training, athlete testing and competition data. As part of a larger project investigating a range of personal performance technologies, we have developed a prototype Knowledge & Experience Network (KEN) that aims to support acquisition, storage and use of data by athletes, their coaches and sports scientists.
The first KEN system uses multiple wearable sensors connected via Bluetooth to a smart phone to provide real-time performance information to athlete, and via online storage to coaches and sports scientists. A mobile phone application has been developed to demonstrate acquiring and storing data in real time in the field with an interface for simultaneous voice communication and transmission of performance metrics. A variety of peripheral devices have been incorporated in order to explore the capacity and performance of the system.
We have conducted preliminary research into methods of data sonification (auditory visualisation) relevant to the application, and have created some small exploratory prototypes in order to demonstrate the feasibility of sonification on the phone itself in real-time. We aim to test different techniques for sonification to determine the effectiveness of this type of auditory feedback on athlete training regimes.
- By enabling athletes, coaches and sports scientists to access data and communicate more efficiently, we predict that athlete performance can be improved.
A trial of the KEN system was conducted in Narrabeen, north of Sydney, in October 2009. An elite Australian Kayaker participated with two CSIRO officers - one onsite in Narrabeen, the other in an office in Geelong, Victoria. This trial was designed to simulate a remote participation situation where a sports-scientist joined in a regular training session with the athlete and coach.
The athlete was wearing CSIRO-developed personal sensors that allowed measurement of body angles, connected to a 3G smartphone running custom-built software. The software on the smartphone transmitted the sensor data to a remote database, and this data could be accessed by the coach and sports-scientist with only a few seconds delay. The athlete was also wearing a headset microphone allowing real-time communication via VoIP on a 3G data network.
The trial demonstrated the feasiblilty of the KEN approach, allowing all three participants to communicate and for the coach and sports-scientist to provide instant feedback to the athlete on their performance. More trials will be conducted to improve usability and robustness of the system, and to test simultaneous participation by multiple athletes.
Contacts
Publications
- Matthew D'Souza, Tim Wark and Montserrat Ros, "Wireless localisation network for patient tracking". 4th International Conference on Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP 2008); Sydney, NSW. IEEE; 2008: 79-84.
- Lasse Klingbeil and Tim Wark, "Demonstration of a wireless sensor network for real-time indoor localisation and motion monitoring". 7th International Conference on Information Processing in Sensor Networks (IPSN 2008); St. Louis, Mo. IEEE; 2008: 543-544.
- Lasse Klingbeil and Tim Wark, "A wireless sensor network for real-time indoor localisation and motion monitoring". 7th International Conference on Information Processing in Sensor Networks (IPSN 2008); St. Louis, Mo. IEEE; 2008: 39-50.
-
Lasse Klingbeil, Tim Wark and Niranjan Bidargaddi, "Efficient transfer of human motion data over a wireless delay tolerant network". 3rd International Conference on Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP 2007); Melbourne,Vic. IEEE; 2007: 583-588.
