MINIATURE MICROSENSOR ARRAY BIOMIMETIC "ELECTRONIC NOSE" FOR THE EARLY DETECTION OF AIRCRAFT FIRES.

PRESENT DAY FIRE DETECTORS NORMALLY EMPLOY A SINGLE SENSOR, SUCH AS AN OPTICAL DENSITY OR IONIZATION DETECTOR, THAT HAS WELL KNOWN LIMITATIONS. ALARM DECISIONS ARE THUS BASED ON A SINGLE MEASUREMENT THAT IS OFTEN INACCURATE, AND INSENSITIVE. WE PROPOSE TO USE AN ARRAY OF VERY SMALL AND INEXPENSIVE SOLID-STATE DEVICES THAT ARE HIGHLY SENSITIVE TO A VARIETY OF PHYSICAL EFFECTS AND CHEMICAL VAPOR PRODUCTS ASSOCIATED WITH THE EARLY STAGES OF COMBUSTION. THE BIOMIMETIC "ELECTRONIC NOSE" IN THIS STUDY WILL INCLUDE FOUR SURFACE ACOUSTIC WAVE (SAW) CHEMICAL MICROSENSORS THAT WERE DEMONSTRATED IN PREVIOUS WORK TO DETECT THE ORGANIC VAPORS RELEASD DURING THE INITIATION OF COMBUSTION. A MINIATURE, MULTI-SENSOR ARRAY WITH AN ON-BOARD NEURAL NET PATTERN RECOGNITION ALGORITHM WILL BE REFINED, BUILT, AND DEMONSTRATED. IT IS EXPECTED THAT THIS INSTRUMET WILL BE ABLE TO DETECT AND IDENTIFY THE SOURCE OF THE COMBUSTION (E.G., PAPER, ELECTRICAL INSULATION, FUEL, CIGARETTS, ETC.) WITH SIGNIFICANTLY HIGHER SENSITIVITY AND RELIABILITY THAN EXISTING SYSTEMS. IT IS ANTICIPATED THAT FIRE DETECTORS MANUFACTURED USING THIS SAW MICROSENSOR ARRAY TECHNOLOGY WOULD HAVE A RELATIVELY LOW COST.