AI and Algorithm Development
DRA has extensive experience building novel model-based algorithm designs using Artificial Intelligence (AI), Machine Learning, and other optimization techniques. DRA engineers can develop quantitative problem descriptions and model the problems in a meaningful way. Engineers can then evaluate multiple approaches to solving the problem while analyzing tradeoffs between computational constraints and performance requirements. Using a spiral design approach, iterative lessons learned are used to shape concepts into algorithms that provide the best solutions. This approach has assisted DRA in developing algorithms to solve challenging problems in the following areas:
- Image processing
- Object classification
- Sub-pixel object detection
- Signal detection and characterization
- Sequence analysis
- Object tracking
- Control applications
DRA develops a computer vision algorithm to perform real-time sub-pixel object detection and track aircraft at ranges greater than 15 nautical miles.