Research Proposed: Research is proposed to design and develop a novel concept for a family of low-cost, low-power smart sensors applicable to a variety of platforms and monitoring environments. Integrated power management and communications will enable the sensors to be used widely without adverse impact on the platform while still providing actionable information to other nodes that need to be informed. Integrating advanced machine learning networks directly into the sensor processing provides for low latency detection of faults without any offboard communications requirements. Problem Statement: Traditional health monitoring systems for vehicles and aviation are expensive and tailored for specific platforms, making them prohibitively expensive for wide distribution aboard ships or on lower-cost ground platforms (e.g. JLTV, ARV, MTVR). While the Internet of Things (IoT) has greatly advanced technologies for consumer sensing and detection, military-grade solutions are still expensive and dedicated. Plan/Process Outline: The effort will begin with the conceptual design covering key design blocks, requirements, and support components to monitor the health of typical rotating machinery. Following that, a modeling and simulation effort will provide an initial assessment of the concept and alternatives. The initial design specification and capabilities description will prepare for a prototype solution in Phase II.
Benefit: Successful completion of this project will provide a novel concept design for a family of compact devices, approximately 1 cubic inch in volume, that is affordable at scale costing $100 or less per node. Integrated sensing will cover typical line replaceable unit (LRU) health sensing including vibration, temperature, and electrical current. External communications, over CAN or IEEE 1451 wireless, will allow nodes to be linked together to share the output of their onboard processed health sensing. Fielding this capability will provide low-cost health monitoring, increasing the ability of the warfighter to perform reliability-centered maintenance and to trust their equipment.
Keywords: Signal processing, Signal processing, wireless sensors, condition-based monitoring, On-board Analytics, neural network chips, integrated processing, anomaly detection, CBM
