Navy – 3D models of maritime platforms from UAV data
PROBLEM
Joint Survivability Tactics Validation Unit (JSTVU) uses methods such as trials and modelling & simulation to assess the survivability tactics of platforms across Navy, Army, and Airforce.
To assess the survivability and operational effectiveness of a platform the development of signatures is critical.
Currently the only methods for developing a signature models of platforms is through costly and time consuming measurements, or computer simulations of outdated 3D models.
However, Defence requires a fast and inexpensive method of signature analysis to support changes through platform life and during operations. Effective signature analysis through computer simulation is dependent on highly representative 3D models of the platform of interest. Defence requires a capability to develop these 3D models.
SOLUTION
Shoal supported JSTVU, in collaboration with Defence Science and Technology Group (DSTG), to develop the concept and assess the feasibility of the capability across two phases: Phase 0 and Phase 1.
The capability uses commercial technologies and UAV-based data capture to produce highly representative 3D models to support decision making. The capability offers the benefits of:
- Representative 3D models for versatile use in signature analyses
- Fast and simple deployment with low operating and support costs by using commercially available technologies
- Allow for high-fidelity calculation of signatures for different EM bands and different incident angles.
Commercial technologies employed include: LiDAR, photogrammetry, artificial intelligence & machine learning, uncrewed aerial vehicles (UAV), and high-performance computing.
In Phase 0, a broad set of data capture technologies were explored through a technology survey. This was complemented with a trial conducted on a merchant vessel.
Phase 1, built upon the work from Phase 0, to build a deeper understanding of the benefits and limitations of photogrammetry. This included exploring and applying techniques in artificial intelligence and machine learning to the data. Phase 1 included a trial on an Armidale Class Patrol Boat to baseline the time and effort required to produce a fit-for-purpose 3D model and explore challenges with photogrammetry on a grey vessel.
RESULTS
Shoal effectively and rapidly supported JSTVU in the assessment of the feasibility of the application of commercial technologies to mature the concept of the capability. Deliverables under Phase 1 included Feasibility Study Report, Concept of Operations, and a Plan for Phase 2.
Shoal now has a deeper understanding of the technologies needed to produce a fast and highly-representative 3D model, including a plan towards operationalising the capability.