RFID and sensor device to help older people stay at home
The development of a new device by Intel Research in the US that combines passive radio frequency identification (RFID) technology with movement and location sensors has led an Australian research group to begin a study into its application in health and aged care.
The study, led by University of Adelaide Associate Professor Michael Sheng, will investigate the use of the device, called the Wireless Identification and Sensing Platform (WISP), with context-aware computer algorithms to create a low-cost, passive monitoring system that could help keep older people in their homes for longer.
Dr Sheng and his colleague Damith Ranasinghe have extensive knowledge of information systems and RFID. A large part of the project will be the development of a fully automated, context-aware, common sense-based activity reasoning engine that can recognise common activities automatically and without input from users.
Dr Sheng describes context aware technologies as applications that can understand a person's environment and surroundings, but that can also make some decisions.
“With this project, the first part is we will try to detect human activity,” Dr Sheng said. “For example, a person is walking or sleeping or lying down. But this is not enough to make the right decision.
“This is about context-aware, common-sense interpretation of activities. For example, walking: if a person is walking around the house in daytime, that means he's okay and he's quite mobile. But if we consider the time – time is context information – if the walking is happening very early in the morning then there might be something wrong.
“Another example is lying down. If we can see the context of the location, if we detect that he is lying down in the bedroom, he might be okay. But if the old person is lying down in the kitchen, perhaps there is something wrong. That is what we mean by context aware. Basically we need to consider all of the information, and then based on that we use common sense knowledge and we get the right information.”
There are many low-cost technologies currently being used to monitor older people in their homes, but most have serious limitations, Dr Sheng said. Video monitoring is very intrusive and also requires a person at the other end to be on the watch constantly.
Pendant alarms are the most common safety device deployed, allowing the wearer to call for help if they have had a fall or are in difficulty. The limitation with these devices are that they are battery-powered and rely on the older person, who may have cognitive impairment, to remember to change the battery and to wear the pendant in the first place.
Low-cost sensors are also used, mainly attached to clothing or a bracelet, but these also require the person to remember to wear the sensors and to change the battery.
Instead, this project will look at using passive RFID tags, which are very cheap and don't need energy. RFID tags run on radio waves emitted by a reader and can be deployed on any object.
“Our project is very ambitious,” Dr Sheng said. “Basically we don't ask the people to wear anything and they don't have to do anything. We are trying to use passive RFID tags that we attach to objects that the people interact with a lot, like a teapot. The software then tries to interpret that he is going to make a cup of tea.”
The idea of using passive RFID tags is nothing new, but Dr Sheng said they were previously limited as they were only capable of doing object identification. The recent development of the WISP device by Intel Research now means that passive RFID tags have been combined with motion and location sensors such as accelerometers – commonly used in smartphones – into one cheap device.
It is understood that when the WISP device is mass manufactured, it will cost about $1 each – sensors are cheap and RFID tags cost about 20c. The device will be able to do the same activity recognition as wearable and battery-powered sensors, but without the drawbacks of user involvement, expense and maintenance.
The main work is in developing a fully automated approach that takes the data collected from the sensors and makes sense of it in a person's normal activities. It will also apply an activity reasoning engine so interventions can be made quickly.
“In this project we are not going to invest in new hardware devices,” Dr Sheng said. “We will use this new Intel device and focus on the software part. We can collect the data and look at how we can analyse this data to interpret what the person is doing.
“On top of that we want to identify all of his normal activities like walking, sleeping and cooking, but also context-aware, common-sense interpretation to give a more accurate information about him.”
The research is being funded under the Australian Research Council's Discovery Project scheme over three years, in a collaboration with the University of Queensland and the University of Washington.
Posted in Aged Care