INTELLIGENT SYSTEMS FOR THE IN-SITU EVALUATION OF MATERIALS AND STRUCTURES

ADVANCES IN MEASUREMENT TECHNOLOGY HAVE MADE AVAILABLE LARGE QUANTITIES OF POTENTIALLY VALUABLE DATA REGARDING THE PROPERTIES AND BEHAVIOR OF STRUCTURES AND MATERIALS IN-SITU. THE COMPLEXITY OF THE DATA, THE SIZE OF THE DATABASE, AND THE AVAILABILITY OF ADVANCED MODELLING TECHNIQUES HAVE MADE DATA ANALYSIS A COSTLY PROCESS REQUIRING A HIGH DEGREE OF EXPERTISE. THIS PROPOSAL DEVELOPS THE CONCEPT OF AN "INTELLIGENT" SYSTEM, BASED ON PRINCIPLES OF ARTIFICIAL INTELLIGENCE (AI), FOR THE ANALYSIS AND INTERPRETATION OF IN-SITU MEASUREMENTS. SUCH A SYSTEM WOULD COMBINE STATISTICAL SIGNAL PROCESSING AND PATTERN RECOGNITION WITH RULE-BASED "EXPERT" REASONING DEVELOPED FROM KNOWLEDGE OF THE MEASUREMENTS AND THE SYSTEM BEING MEASURED. SIMILAR SYSTEMS HAVE BEEN DEVELOPED FOR MEDICAL, GEOLOGICAL AND MILITARY APPLICATIONS, AND CAN BE ADAPTED TO SUCH CIVIL ENGINEERING APPLICATIONS AS GEOTECHNICAL SURVEYS, HAZARDOUS WASTE SITE CHARACTERIZATION, AND INSPECTION OF PAVEMENTS AND BRIDGE DECKS. THE OBJECTIVES OF THE PROPOSED WORK ARE TO IDENTIFY IN-SITU MEASUREMENT APPLICATIONS TO WHICH AI SYSTEMS CAN AND SHOULD BE ADAPTED TO CONFIGURE A GENERALIZED AI SYSTEM TO ADDRESS THESE APPLICATIONS, AND TO EVALUATE THE FEASIBILITY OF DEVELOPING A WORKABLE SYSTEM FOR EACH APPLICATION.