Hivemind Nova is the first artificially intelligent quadcopter for battlefield applications. It is designed to reduce risk to forces and missions. Hivemind Nova learns from experience to access and navigate denied or high threat environments. Without a remote pilot or human input, it autonomously explores buildings, urban canyons, caves, tunnels, and other high threat environments ahead of friendly forces. It live-streams HD video and a map of the building layout to operators. Hivemind Nova can also conduct searches without an RF link, be flown manually, and be used as organic outdoor ISR. Hivemind Nova learns and continuously improves over time. The more it is used, the more it learns, and the more capable it becomes.
Hivemind Argos provides real-time analysis for immediate detection and classification of information using trained neural networks. Trained on people and vehicles, and trainable to find arbitrary signatures of interest; Hivemind Argos notifies analysts when it identifies objects of interest, so an analyst can spend time analyzing instead of watching video feeds. It automatically generates and distributes intel products and reports to analysts, and also catalogs the data which is analyzed by Hivemind to identify patterns of life and activity based on areas of interest.
Hivemind is a leading AI that ingests data from individual robots, teams of robots, and external sources like weather, video feeds, or social media, to learn and formulate new skills. With Hivemind, robots improve daily to accomplish new missions. It is a framework for encoding information that is readily expanded and recalled to facilitate AI operations. This framework consists of learning techniques; online, deep, and lifelong; and a knowledge base where data is stored, maintained, and queried via the learning techniques. Learning occurs on the edge (individual robots), in the collective (teams of robots), or through exo-intelligence sources (weather, social media, video feeds).
These new skills and features can be ported to different types and sizes of robots. For example, controls policies or a distributed perceptual model formulated by Hivemind for a small quadcopter could be transferred to a fixed-winged drone. All Shield AI products are architected around Hivemind.