Sensor fusion and threat analysis

A counter-uncrewed system (C-UxS) rarely gets one clean view of a target. Different sensors each see part of the picture, at different rates and with different errors, and third-party sensors add more. The software has to merge all of it in real time into tracks an operator can trust, and say how much each track matters.

That is the job of the Sensor Fusion Engine and the threat analysis inside DroneSentry C2, DroneShield's command-and-control product suite. Details here are limited to what the company has made public.

• Led the research and development of ThreatAI, an industry-first AI threat analysis capability for DroneSentry C2, from honours project to product.
• Designed, researched, developed and maintain the Sensor Fusion Engine and the threat analysis algorithms deployed in DroneSentry C2.
• Built and optimised machine learning models and classifiers: traditional ML, deep neural networks, attention-based architectures and computer vision.
• Implemented non-AI fusion: probabilistic tracking, physics-based motion models and complex decision logic.
• Wrote the backend services, REST APIs and databases behind it in Python, Go, C++ and SQL.
• Built the internal tooling that makes it testable: simulators, data replay systems, benchmarking frameworks, annotation tools and data-injection pipelines.
• Automated MLOps and CI/CD on GitLab for training, evaluation, deployment and large-scale benchmarking, and introduced ClearML as the company's MLOps and DataOps platform.
• Worked with the AI, Vision AI, data, hardware, field, application software, third-party sensor and demo teams through release testing, field trials, validation and customer demos.

The fusion engine and threat analysis ship inside DroneSentry C2. The work began as my honours project, an industry-linked research project on AI threat analysis, which earned the University Medal. It led to the AI Rising Star of the Year award (Australian AI Awards, 2024), Rising Star of the Year at the Australian Defence Industry Awards (2025), and the National Industry PhD that starts in 2026.