Vadum brings critical experience and proven capabilities to the Missile Defense Agency (MDA) Scalable Homeland Innovative Enterprise Layered Defense (SHIELD) initiative. These programs include novel Machine Learning (ML) techniques to complete object and target classification on time-series data, enabling DoD customers to classify objects and targets from single sensor or multi-sensor time series data. Utilizing feature extraction and ML modeling and training, Vadum was able to output the object and target captured in the highly noisy data and measure level of confidence for the classification. This application has direct applicability to MDA for target and object classification on either single or multi-sensor fusion data.

For another DoD customer, Vadum developed a Machine Learning-based targeting algorithm to predict future states of incoming threats. This application utilizes threat kinematics and ML model based algorithms to predict where in space (position) and time a threat would be. This enables incoming object tracking and for planning interceptor engagement. This system predicts the future state of threats without requiring full or consistent track information. Vadum’s application works with just a few tracks from a sensor to produce a predicted future state. This application is based on near-peer adversarial missile threats and thus has direct application to the goals of MDA’s SHIELD.