Unmanned Aerial Vehicles (UAVs) are rapidly proliferating across commercial, government and civilian domains for inspection, surveillance, emergency response and delivery services. However, high failure rates continue to undermine mission success, safety and regulatory approval. To address this, a retrofit health and usage monitoring system (HUMS) tailored for small multirotor UAVs was developed through iterative design, simulation and experimental validation to enable condition-based maintenance.

A comprehensive literature review identified prevalent UAV failure modes and assessed suitable technologies for continuous monitoring. Fiber Bragg Grating (FBG) sensors emerged as most promising due to attributes like high sensitivity, low size/weight, multiplexing ability, and durability. Subsequently, the applicability of these technologies to address failure modes related to structural, electrical, temperature, vibration, and environmental factors were evaluated. System architecture options like mesh networking provided redundancy against individual node failures.

Based on this analysis, custom design requirements were established for a retrofittable HUMS prototype able to interface with different UAV types. The functional prototype implements distributed temperature and vibration sensors connected via Zigbee to a central microcontroller with WiFi telemetry to ThingSpeak cloud analytics. Controlled lab integration on a quadcopter successfully demonstrated real-time streaming of sensor measurements and threshold-based anomaly detection. The simplified proof-of-concept establishes core capabilities, despite limitations in robustness, range and advanced analytics.

Ongoing work focuses on stacking reliability through improved packaging, resilient communications and ease of installation. Transitioning to LoraWAN aims to expand coverage for remote deployments. Edge computing and machine learning techniques will elevate diagnostic intelligence. Testing will shift from lab settings to diverse field conditions spanning various UAV types, flight profiles and live missions. A preliminary techno-economic analysis projects attractive returns on investment through reduced downtime and maintenance costs, substantiating commercial viability.

This research pioneers an innovative HUMS architecture tailored for size, weight and power constraints of small aerial platforms. By facilitating the transition to predictive maintenance, the system promises to significantly bolster efficiency, safety and longevity in UAV operations. The successful demonstration of a pragmatic HUMS holds disruptive potential to overcome reliability barriers hindering the next evolution of ubiquitous and autonomous unmanned flight.