A brief overview of ubiquitous computing research at the Laboratory for Communication Engineering, Cambridge
George Coulouris · 2004
We shouldn't expect technologies that impact peoplesdaily lives to be adopted before they are demonstrated as reliable, useful and simple to deploy. Measured against this criterion ubiquitous computing is still in its infancy. We present an overview of the spectrum of research in ubiquitous systems currently in progress at the LCE which aims to address some of these issues. Ubiquitous computing is undergoing a long gestation period. The factors contributing to this are similar to those that delayed the acceptance of other technologies such as personal computing and computer networking. Like those technologies, ubiquitous computing demands substantial investments in infrastructure and changes in human behaviour. Our role as technologists is to ensure infrastructure is reliable and cost-effective and to understand the space of context-aware applications. Research in ubiquitous computing at the LCE ranges from studies aimed at enhancing the accuracy, reliability and deployability of location-aware systems to investigations of augmented reality techniques as a means to add interactivity to a world in which computers and communication devices are pervasive. Between those extremes, we are also studying some possible approaches to achieving privacy in sentient environments and we have deployed several location-aware applications that seem to meet needs in our user community. Accuracy, reliability, deployability The Active Bat location sensing system (1) was originally developed and deployed in the AT&T Cambridge Laboratories and is now deployed in the LCE. As a research prototype, the Active Bat system was designed to provide the highest performance available with current technologies while paying due regard to user needs associated with its deployment, maintenance and use. The system tracks small wearable tags with an accuracy of a few centimetres and a time resolution of a few milliseconds. The tags are small and lightweight and have a battery life of 1-2 years in normal use. In addition to their tracking functionality, they are equipped with two buttons for user input and two LEDs and a small speaker for user feedback. This frugal provision for interaction is dictated largely by a desire to achieve a long battery life. Location tracking in the Bat system depends on the deployment of an array of ceiling-mounted receivers and the use of the signals that they receive to perform ranging calculations. The accuracy of these measurements is normally high, but environmental conditions have an impact on them through occlusion (producing weak signals) and reflection (producing ambiguous signals) and errors sometimes occur due to component failure. Investigations are in progress in the LCE into the detection and correction of all of these conditions. Another study has demonstrated that the mapping of a physical space in which the tracking system is deployed can be, at least to some extent, automated (2). These studies are aimed at producing a more robust and more manageable location tracking technology. Application deployment