UTA Researchers Develop Radar-Based System to Predict Drone Failures at Low Cost
As drone fleets expand across commercial and industrial sectors, preventive maintenance is emerging as a critical challenge for operators. A team at the University of Texas at Arlington (UTA) is addressing this issue by developing a low-cost radar-based system designed to detect mechanical anomalies in unmanned aerial vehicles (UAVs) before they lead to failures. Led by Dr. Dianqi Han, assistant professor in the Department of Computer Science and Engineering, the project uses millimeter-wave radar to monitor key operational parameters such as propeller rotation speed, vibration intensity, and flight trajectory to flag early signs of wear or damage in real time.
The proposed system is lightweight and scalable, relying on millimeter-wave radar technology that allows a single unit to monitor the mechanical health of multiple drones simultaneously. Unlike traditional manual inspections, which are still the norm in the industry, the UTA system enables continuous, remote diagnostics from distances exceeding 100 meters. Han emphasizes that as drone fleets age and scale up, the current reactive approach—where failed drones are simply replaced—will no longer be cost-effective or operationally sustainable. The new system aims to shift the paradigm toward predictive maintenance, extending operational lifespans and reducing downtime.
The radar technology chosen for this system, developed by Texas Instruments, is both affordable and powerful. It can detect slight variations in mechanical performance that typically precede serious malfunctions. This includes detecting unbalanced rotors, decaying motor speed, or early structural weakness in drone joints. According to Han, the full monitoring system, including a radar unit and a basic laptop for processing, could be deployed for under $600. This makes the system particularly attractive for drone delivery companies and operators managing large autonomous fleets, where any unexpected failure can disrupt operations at scale.
The system is currently undergoing testing in controlled environments. Initial results demonstrate its ability to accurately measure critical flight parameters and identify anomalies linked to hardware degradation. The research team is now working on correlating specific signal patterns with known failure modes to improve the predictive capabilities of the platform. This will allow the system not only to flag problems but also to provide diagnostics and alerts tailored to specific components or systems, helping maintenance teams take targeted actions before a drone becomes non-operational.
Once the radar-based prototype is validated, the next phase will focus on securing funding from federal agencies or private sector partners to scale the technology. Challenges remain, including ensuring consistent radar coverage for different drone models and adapting the system to environmental variations such as wind, temperature, and electromagnetic interference. However, Han is confident that commercialization is possible within two years. The project demonstrates how relatively simple and cost-effective tools can dramatically improve the reliability and safety of drone fleets, offering a compelling solution to a growing problem in the UAV sector.