CR Tactical proposes to participate in the Fire-Faster-Cohort and provide a Cannon Ammunition Vision System (CAVS) including redundant on-board inventory management in order to provide the real-time disposition of self propelled howizter class five (CL V) ammunition supplies. CAVS would fulfill the vision and onboard-inventory component of a Fire-Faster System of Systems. The envelope of processable CL V includes shells, fuzes, propellant (both in and out of casings), stub cases, and fuzed shells. The CAVS includes a camera based sensor suite designed to scan a shell, fuze, propellant, stub case, or shell/fuze combination and create imagery and 3D data of the outer surface of the ammunition item. The CAVS analyzes the imagery and 3D data in real-time and compares the result to a database of known ammunition characteristics to ascertain the classification (database identification), Department Of Defense Ammunition Code (DODAC), weight markings, nomenclature, LOT#, and serviceability determination (presence of visual defects and proper installation) of the CL V. The CAVS will support integration with the Extended Range Cannon Artillery (ERCA), M992A3 Carrier Ammunition Tracked (CAT), and the M109A7 Self-Propelled Howitzer (SPH), and potentially other cannon artillery systems . Because it is provided in electronic form, the data produced by the CAVS will be immediately available to the Fire-Faster control computer and Paladin Digital Fire Control System (PDFCS) for inventory visibility. Electro-mechanically, the CAVS is an assembly that includes real-time processing, motorized CLV manipulation, environmentally sealed ammunition entry apertures, a redundant pair of hybrid low-power Field Programmable Gate Array, Central Processing Unit, and Graphics Processing Unit System on Module, stereo camera pairs, and illumination. This assembly can be used as a self contained unit for testing and can be decomposed to integrate with the larger Fire-Faster SoS assembly. The CAVS is based on an advanced but proven family of products that Carnegie Robotics has developed for the water and sewer utility inspection industry that produce 2D and 3D data-sets using machine vision and monocular and stereo structure from motion in order to produce as built diagrams, defect modeling and detection, and plot and plan maps from environmentally sealed and rugged camera assemblies that traverse the pipe or manhole. This technology used in reverse, that is scanning an object inside of a cylinder rather than scanning the cylinder, is the central proposed concept of CAVS.
