At Iconal I worked on two projects in the security sector.
The first was developing a machine learning algorithm to classify custom X-Ray images of bags into threat vs non-threat. The X-Ray sensor setup was custom-build, and used multi-sensor fusion to arrive at the classification. I hold a patent for the algorithm, and was responsible for the entire data algorithm design process, from data collection to model training and deployment. I gave a talk on the algorithm development process at Cambridge University.
The second project was consulting on a multi-sensor project using time series classification methods.