Sensor Technology Box

A Sensor Technology Box for Smart Health (SenseBOX)

The U.S. Census Bureau reports that the U.S. population of people aged 65 and up will grow more than double in between 2010 and 2050. The market for remote patient monitoring is expected to grow from $10.6 billion in 2012 to $21.2 billion in 2017. Inspired by this growing market need, we propose to build an integrated system consisting of commercially available off the-shelf sensors and servers that can be deployed for independent living applications across multiple homes for monitoring activities of daily living (ADLs) of older adults. Existing smart home technologies are mostly targeted at technology enthusiasts, with some limited offerings in smart health. They provide simple monitoring and control, e.g., times of activity and the ability to turn on and off lights. They range from fairly inexpensive to very expensive while being produced in limited quantity. We propose to develop a system which can be very price-competitive, especially by partnering with people already producing devices that can be easily modified to run our software (removing most of the need for full-custom hardware development associated with a substantial cost). Existing technologies also fail to analyze data in real time; while they can provide information about e.g., weather a person has been active or taking medications, we can mine data for more subtle trends that may indicate depression or illness that could warrant intervention. When compared to existing smart heath technologies, we can provide more useful information (detection and assessment of both ADLs and cognitive impairments) while still meeting or beating the price of the competition and promote the large scale deployment in real-life setting.

System Development

A Linux kernel and base system has been built for the CloudEngines PogoPlug and tools to communicate with the Cao Gadgets Wireless Sensor Tags have been developed. The PogoPlug is currently acting as a bridge between the Sensor Tags and a standard 802.11N network in an in-lab test configuration. Tests have indicated the device is stable in this capacity, and recovers from power loss and outages on the external networks without issue. The PogoPlug is also successfully doing NAT translation in order to bridge additional Ethernet devices on to the network. Devices like an IP camera used for collection of ground truth benefit from such bridging.

IRB-approved test runs in a senior living community are planned within the next two weeks, with a wider test deployment planned in the coming months.

Installation Instructions: TBA

Hardware Components: Raspberry PI, PogoPlug, Z-wave, wireless sensor tags.

Software Components: Linux, openHAB, Klone web server.