Multisensory Data Fusion for Ubiquitous Robotics Services

Abstract

Real time human tracking in indoor environment is highly demand and important to many applications for ubiquitous robotics services. There are several technologies for human tracking in the indoor environment, such as vision, pressure, infrared, and ultrasonic have been proposed. However, some technologies are costly, or have many restrictions when they are applied. For example, implant several cameras around indoor environment through real time image processing for tracking people is frequently discussed. However, the computer vision method does not function in dark situation and the privacy issue also generated. Recent advances in radio frequency, microprocessor and sensor technologies enable the wireless sensor network (WSN) system. Wireless sensor network system used for target tracking can be found (Liu, J. et al., 2007 ); (Hua Li et al., 2008); (Djuric, P. M. et al., 2008). In general, the smart home aims to provide appropriate intelligent services to assist the resident’s living. Autonomous and multi-functional robots plays important role in the smart home. However, if robots depend on their own sensors, the applications will be limited. Sensor network system uses multi sensor combined with microprocessor and radio transmission into a device and deploy in the monitoring environment. With the ambient information provided by sensor network based ubiquitous robotics services system, robots can serve people more quickly and accurately. There are a lot of researches discuss sensor network theories or applications. However, these researches (Yoon Gu Kim et al., 2006); (Taehong Kim et al., 2008) deal with implement sensor network in indoor environment just implant additional sensor nodes on the desks or ceiling. We propose a solution on implementing the sensor network system for ubiquitous robotics services in the indoor environment. Since most building have implant traditional smoke detector on ceiling for fire detection. We try to implement sensor network system in indoor environment and modify these smoke sensors. We developed a prototype of wireless multi-functional detector; the sensor consists of pyroelectric infrared (PIR) sensor for detecting human radiation, temperature sensor for measuring the ambient temperature and smoke sensor to replace the original smoke detector function on ceiling. With this multi-functional detector, robot can locate the remote people in the environment and provide the services. O pe n A cc es s D at ab as e w w w .in te ch w eb .o rg