Newman R.M.
Coventry University, UK
Keywords: arrays, communication, digital processing, microsensor
In 1994 the author patented an intelligent vibration sensor. The device was essentially an accelerometer with integrated analog and digital processing capability as shown in figure 1, which allowed a large proportion of the diagnostic functions of a complete vibration monitoring system to be localised at the accelerometer. This was a response to one of the major problems encountered with design of such systems at the time. The number of accelerometers in a system had escalated steadily, causing severe data processing bottlenecks. An array of intelligent sensors provided a parallel machine, the processing capacity of which expanded to meet the number of sensors used. By localising signal processing and reduction the data communication bandwidth was also reduced. At the time the proposal was at the limits of available integration technology, and the patent was not utilised. With current fabrication processes it is feasible to integrate a micromachined accelerometer along with a substantial processing capacity onto a single chip making the intelligent sensor a realistic and very economical possibility. In fact, such a device could be sufficiently inexpensive to make possible the building of systems using much greater numbers than considered before, given the open ended processing capacity of such systems. This paper considers the properties and some of the potential of such systems. An amendment to the original intelligent vibration sensor architecture is proposed, which consists of the addition of three further communications links and a dedicated communications processor (figure 2). The device then becomes an intelligent array vibration sensor, which can be connected together in a square, two dimensional array of sensors of any size required (figure 3), which could be integrated permanently into the fabric of the machine or vehicle under test, the main constraints being those of power consumption and the ability to calibrate sensors and handle their failure. Previous work by Gaura et. al. shows how the individual sensors might be calibrated and linearised using neural network methods. Such an array may be used in a number of ways, which are considered fully in the paper, including some preliminary analysis of the systems requirements, using ?-calculus. i) The large number of processors in the array may be used to enhance reliability by allowing for redundancy. Means would need to be developed for automatic detection of faults and allocation of tasks to adjacent sensors. ii) Different analysis tasks may be distributed to different processors within the array, allowing a single system to perform several analyses simultaneously. iii) Recognition and analysis algorithms might be distributed to the array as a whole, using the complete system as a pattern recognition device for different spatial vibration patterns. Such patterns are often diagnostic of particular mechanical faults, and could be detected by such means. Systems are likely to use all three of these possibilities, providing systems with specialised processing and an element of redundancy, providing powerful diagnostic procedures with graceful degradation in the case of individual sensor failure. References 1. Newman, R.M. and Robinson, B. (1994)Intelligent Vibration Sensor, U.K Patent No GB 2251071, Grant Date 3 Aug 1994 2. Gaura, E. Rider, R.J. Steele, N. (2000) Closed-loop neural network controlled accelerometer, Proceedings of the I. Mech. E, Part I, Journal of Systems and Control Engineering, vol. 214, no.I2, pp.129-138. 3. Gaura, E. Rider, R.J. Steele, N. (2000). Developing smart micromachined transducers using feed-forward neural networks: a system identification and control perspective. The IEEE International Joint Conference on Neural Networks, IJCNN2000, Proceedings, ISBN 0-7695-0619-4, Vol. IV, pp. 353-358, Como, Italy. 4. Newman, R. M. (1998) The ClassiC Programming Language and Design of Synchronous Concurrent Object Oriented Languages, Journal of Systems Architecture, Elsevier Scientific, September 1998. 5. M. Chu, H. Haussecker, F. Zhao, (2002) Scalable information-driven sensor querying and routing for ad hoc heterogeneous sensor networks. Xerox Palo Alto Research Center Technical Report P2001-10113, May 2001. 6. Estrin, D., Govindan, R., Heidemann, J., Kumar, S., (1999) Next Century Challenges: Scalable Coordination in Sensor Networks. Proc. ACM International Conference on Mobile Computing and Networking, (Mobicon 1999) pp. 263-270. 7. Elson, J. ,Estrin, D. (2001)Time Synchronization for Wireless Sensor Networks, Proceedings of the 2001 International Parallel and Distributed Processing Symposium (IPDPS), Workshop on Parallel and Distributed Computing 8. Wang, A, Chandraskan, W (2002) Energy Efficient DSPs for Wireless sensor networks, IEEE Signal Processing magazine, July 2002, pp. 68-78. 9. Jaikaeo, C., Srisathapornphat, C., Shen, C.-C., (2001) Diagnosis of Sensor Networks, Proc IEEE International Conference on Communications, Helsinki, Finland, June 11–14, 2001. 10. Milner, R, (1999), Communicating and Mobile Systems,: the Pi-Calculus, Cambridge University Press, Cambridge, England.
Journal: TechConnect Briefs
Volume: 1, Technical Proceedings of the 2003 Nanotechnology Conference and Trade Show, Volume 1
Published: February 23, 2003
Pages: 266 - 269
Industry sector: Sensors, MEMS, Electronics
Topics: MEMS & NEMS Devices, Modeling & Applications, Sensors - Chemical, Physical & Bio
ISBN: 0-9728422-0-9