Dynamic sensor selection for robotic systems

Abstract

A new technique for selecting, in real time, different sensing techniques for a robotic system has been developed. The proposed method is based on stochastic dynamic programming, which provides an effective solution to multi-stage decision problems. A t each stage in the decision process a sensor selection controller has the option of consulting a new process monitoring technique to improve the knowledge of the task or terminating the decision process without any further information gathering. The sensor selection controller has been successfully implemented for the real-time control of a planar robotic assembly task in a discrete event control framework. One of the monitoring methods used is based on Hidden Markov Models, where the average recognition rate was 87%. Larger recognition rates for the HMM method have been demonstrated by 141. The rate of 87% was chosen to show the efectiveness of the dynamic sensor selection method. The experiments show that the method performs better than any individual process monitor. A successful event recognition rate of 97% with an average CPU time of 0.38 seconds is achieved when two force monitors and one position monitor are available to the sensor selection controller.

Sensor Selection via Convex Optimization

“…The sensor selection problem arises in various applications, including robotics [HM97], sensor placement for structures [Kam91,KP02], target tracking [WYPE04,IB05], chemical plant control [KP99], and wireless networks [ZG04]. Sensor selection in the context of dynamical systems is studied in, e.g., [GCHM06,KP98,Osh94].…”

Sensor Selection for Event Detection in Wireless Sensor Networks

“…x + = max (0, x), S = S −1/2 0 Lemma 2 says that, in order to solve (9), it suffices to search over the set R ⊂ R 2 . For n ≥ 3, the set R is compact and convex, as the image set of a unit sphere under two quadratic mappings, see [24].…”

“…We explained how to solve (10) by parameterizing ∂R. Now we explain how a solution of (10), x , y , is sufficient to reconstruct v , a solution of (9). Vector v in Lemma 2 can be found as follows.…”

“…subject to e (j) e (j) = 1 (16) where m (16) is simply an instance of (9) in the reduced dimensional space R n−j+1 , for which we have developed a global solution. Algorithm 1 outlines the overall approach.…”

“…different contexts, including target tracking (e.g., [7]), target localization (e.g., [8]), robotics (e.g., [9]), and wireless sensor networks (e.g. [10], [11]).…”