Data Sensor Fusion for Autonomous Robotics

Multi-sensory information is a generic concept since such information is of concern in all robotic systems where information processing is central. In such systems for the enhancement of the accurate action information redundant sensors are necessary where not only the number of the sensors but also the resolutional information of the sensors can vary due to information with different sampling time from the sensors. The sampling can be regular with a constant sampling rate as well as irregular. Different sensors can have different merits depending on their individual operating conditions and such diverse information can be a valuable gain for accurate as well as reliable autonomous robot manipulation via its dynamics and kinematics. The challenge in this case is the unification of the common information from various sensors in such a way that the resultant information presents enhanced information for desired action. One might note that, such information unification is a challenge in the sense that the common information is in general in different format and different size with different merits. The different qualities may involve different accuracy of sensors due to various random measurement errors. Autonomous robotics constitutes an important branch of robotics and the autonomous robotics research is widely reported in literature, e.g. (Oriolio, Ulivi et al. 1998; Beetz, Arbuckle et al. 2001; Wang and Liu 2004). In this branch of robotics continuous information from the environment is obtained by sensors and real-time processed. The accurate and reliable information driving the robot is essential for a safe navigation the trajectory of which is in general not prescribed in advance. The reliability of this information is to achieve by means of both physical and analytical redundancy of the sensors. The accuracy is obtained by coordinating the sensory information from the redundant sensors in a multisensor system. This coordination is carried out by combining information from different sensors for an ultimate measurement outcome and this is generally termed as sensor fusion. Since data is the basic elements of the information, sometimes to emphasize this point the fusion process is articulated with data as data fusion where the sensor fusion is thought to be as a synonym. Some examples are as follows. “Data fusion is the process by which data from a multitude of sensors is used to yield an optimal estimate of a specified state vector pertaining to the observed system.” (Richardson and Marsh 1988) “Data fusion deals with the synergistic combination of information made available by various knowledge sources such as sensors, in order to provide a better understanding of a given scene.” (Abidi and Gonzales 1992)