(NEPTUNE) Threat Detection & Tracking for Vehicle Gunners
Abstract
ABSTRACT: Vehicle gunners need to improve the speed and reliability of threat detection and targetacquisition/engagement. Lidar specifically designed for A) long-range, B) rapid, accurate, andflexible beam-steering, C) detection using laser pulse shapes, D) complementing cameraimaging, E) tracking objects i.e., the ability to follow moving objects with adaptive scanning,and E) integration with Intelligent Agents -- can form the basis for enhanced protection ofoperators and gunners for military trucks and light tactical vehicles (e.g., JLTV). Imaginesensors / AI will mark regions of the field of view as both important (e.g., object predicted tointersect with vehicle pathway) and which lack sufficient resolution for high qualityclassification. These regions will be adaptively probed with dense point clouds concentratedwhere disambiguation is required in order to fully recognize objects in the field of view.Commercial and capital markets anticipate a tidal wave of change in the automobile andtruck markets and have unleashed torrents of effort and capital to position themselves for thearrival of autonomous vehicles. This new trillion-dollar sector of the economy will be based onseveral key technologies which are coming to fruition at the same time: A) algorithmicdevelopments in artificial intelligence and machine learning, B) AI chips which are afundamental change in the nature of computing hardware, shifting from von Neumannarchitecture to distributed memory and computation mimicking neural structures of the brain (seefor example research and commercial projects in neuromorphic computing, Brainchips Akida,Intels Loihi, or Graphcore), C) 3D Sensors lidars, radars, and ultrasonic sensors can produce3D point clouds (azimuthal, elevation, and range), providing a new basis for precisionpositioning, object detection, classification, and tracking, and D) the electrification andoptimized coordination of vehicle fleets.In previous work (Neptune 1.0), the ASU group designed a lidar to complement camerabasedimage processing, which we believe will be the work-horse of vehicle sensor suites. Thecamera will mark and hand-off to the lidar regions that are both of high interest and ambiguous.The lidar prototype has: A) 1550 nm wavelength for longer range, B) a MEMS mirror for beamsteeringwhich is flexible enough to scan boundary regions around objects, and C) an analogdigital converter approach to detection, allowing more advanced utilization of the returningpulse-shape and exploration later of AI-chip neural-net based detections. A rack-mountedprototype is currently being tested for short-ranges, up to 10 meters in range. Pulse energy ismeeting specifications for long range. Configurable MEMS mirror scanning is online andworking, and the analog digital converter is working for portions of the full dynamic range.Much of the lab work was done by veterans at ASU.These threads of development to-date will be brought together and adapted to enhancemilitary vehicle operator safety in hostile environments. Basic premises for the proposal are: 1)Sensors for autonomous vehicles should be designed from the whiteboard down with IntelligentAgent integration in mind, 2) Multiple sensor data streams will need to be fused, 3) Intelligenceshould be built into the sensing process at the lowest level, producing inherently sparse datasets,5) The majority of threats to the safety of the vehicle operators and occupants come from a smallpercent of difficult to classify objects and obstructions within the field of view, and 6) Higherlevel information such as extended-perception is of most value to the human operator.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 29, 2020
- Source ID
- N000142012247
Entities
People
- Ronald Calhoun
Organizations
- Arizona State University
- Office of Naval Research
- United States Navy