Article Preview
TopIntroduction
As technology advances, there are more options available for pervasive monitoring. Sensor miniaturization, wireless communication and increasing processing power in smaller packages allow more efficient, reliable and convenient systems, at least from the end-user perspective. In healthcare, one of the main driving forces behind ubiquitous computing is the increasing need to move patient care from the hospital to non-standard settings such as homes, nursing homes, improvised waiting areas, hazardous locations or the battlefield. For at-risk patients, such as those with chronic diseases or the increasingly aging population, being able to live in a familiar and comfortable environment improves quality of life and frees hospital resources. In disaster situations or during seasonal or regional disease outbreaks, response teams move from the hospitals to improvised settings to care for multiple casualties with varying levels of urgency. Firefighters, hazmat teams and soldiers need real-time, ubiquitous monitoring to detect life threatening events. Existing solutions from industry, academia and the military share the same goal of developing unobtrusive, reliable and pervasive monitoring systems.
Powerful, disposable computers, wireless technologies, sensors and energy storage have made possible the development of Body Sensor Networks (BSN) (Aziz et al., 2008). These networks are ubiquitous, allowing patient supervision wherever they may go. Personalized health care is a natural extension of these BSN. Future challenges are user acceptance (from both patients and practitioners), ease of use, and avoiding data flooding with little information.
This paper presents a particular experience of embedding a pervasive system in a healthcare setting and the steps required in the design and test of such a system in the hospital. We show benefits from fusing information from different sensors, the complications that arise when dealing with untethered subjects and finally the evaluation in a real environment. We also show how the implementation of this system required the skills of a multi-disciplinary team.