Real-Time Integrity/Durability Monitoring of Composite Structures Using a Fiber Optic Sensor System and Neural Network Processing

The overall objective of this project is to develop a reliable, low-cost neural network architecture for sensors, and real-time processing to continuously monitor the structural health of the thick composite structure used in armored vehicles. The Phase I demonstration will be implemented with advanced fiber optic sensors for acquiring health and status data from a sample composite structure, a monitoring architecture for interconnecting the sensors, and use of advanced neural network techniques for learning the composite structure characteristics and discerned changes in the structural health of the composite. A team with expertise in composite materials, embedded fiber optic sensors, neural networks, and computer systems proposes to produce a proof-of-concept prototype using real hardware. The Phase I success is ensured through the leveraging of existing R&D by using fiber optic sensor prototypes developed at FIMOD, thick composite technology developed at Simula, and the neural network implementations of Sedona Scientific.