TOPGUN detects and classifies ships in color video using state-of-the-art deep learning techniques, which can be used by any manned/unmanned vessel to avoid collisions. TOPGUN detection is reliable even in cluttered scenes, and detected ships are classified into categories, including cargo vessel, sailboat, and military ship, to support safe navigation for unmanned vessels and increased situation awareness for manned vessels. We have demonstrated TOPGUN live on the water using commercial cameras and processing hardware. We are combining TOPGUN with whale and obstacle detection on a smart camera with an embedded processor to create a maritime smart camera that provides situation awareness and collision avoidance for military and commercial vessels. Ideal partners are building unmanned surface vessels, smart cameras, or ship surveillance and security systems.
Ecological Advanced Support Interface Toolkit for Heads-Up Attention to Improve Warfighter Knowledge (EASI-HAWK) enables better spatial orientation through natural visual and auditory cues that extend beyond the foveal visual system; seamlessly transitioning pilots from aided to unaided vision. EASI-HAWK, an auxiliary toolkit, supports head-up displays (HUDs) and head-mounted displays (HMDs) under development for the F/A-18 and F-35 variants, enhancing pilot effectiveness. Charles River Analytics, a leading provider of innovative R&D solutions for increasingly complex and important human-systems challenges seeks integration with HUDs and HMDs and EASI-HAWK’s visualization display components with a number of military aircraft and land vehicles. The underlying display design principles provide benefits for guiding effective display criteria to support remotely piloted and pilot-optional aircraft, as well as augmented display devices for private and commercial pilots.
Net-centric fleet operations generate an abundance of health and status data that can require skilled engineers to interpret to identify and isolate failing systems (Tier III support). DATEM, a TRL-8 machine learning system targeting Ship's Signals Exploitation Equipment (SSEE) Increment F, instantly pinpoints failed ship sub-component (e.g., DATEM automatically identifies 91% of failures) just from ships’ health and status signals, enabling Tier I technicians to resolve Tier III-level troubleshooting. This results in quicker, more accurate failure resolutions at lower tiers of support, which maximizes up-time, increases operational availability, and lowers costs. For 35+ years Charles River Analytics has been solving critical DoD research and operational problems using Artificial Intelligence. Beyond deploying to SSEE, we aim to solve other Navy critical health and status data understanding problems by adapting DATEM to new systems.
Navy system operators must protect their software applications from cyber-attacks without impacting the performance of mission-critical systems. Detecting Anomalies in Application Memory Space (DAAMS) is a machine learning enabled software framework that efficiently monitors application memory spaces to automatically detect and report known and unknown cyber-attacks as they occur. DAAMS is primarily designed to detect cyber-attacks on Navy ship-based systems such as AEGIS and SSDS, yet it can be applied to any system that may be vulnerable to attacks on application memory, including real-time and embedded systems. Charles River has over 30 years of steady growth providing innovative, cost-effective solutions through intelligent systems R&D. Our goal is to integrate and transition this technology into government and prime contractor systems to increase protection from cyber threats.
Lack of real-time prognostics leads to inefficiencies in preventative and corrective maintenance resulting in wasted resources, increased cost, and reduced mission readiness. SNAPPR is a hardware health-monitoring tool that provides real-time prediction of faults and system health, using these predictions to recommend maintenance actions enhance mission readiness and control costs. SNAPPR technology is modular and extensible, simplifying application to other systems requiring preventative and corrective maintenance. SNAPPR is being demonstrated on the Aegis Radar sub-system and undergoing functional verification. Charles River Analytics is an applied S&T company that works in many areas of Artificial Intelligence, Machine Learning, and Cognitive Systems Engineering with applications ranging from robotics, sensor processing, and autonomous system to decision assistance, interactive training, and advanced human interfaces and visualizations. The ultimate goal is to integrate and transition this technology to the prime contractor for the Aegis radar.
Charles River Analytics is a leading provider of innovative R&D solutions for increasingly complex and important human-systems challenges developed Blended Advanced Decision GUI Environment for Reasoning Support (BADGERS) enabling shipboard maintainers to rapidly analyze system status and predicted malfunctions, evaluate high-level mission impacts, and efficiently make maintenance decisions through intuitive and innovative data visualizations. BADGERS will help the execution of maintenance analysis, planning, and execution in the Navy community, including for complex, mission-critical systems such as Aegis Weapon System (AWS), the Operational Readiness Test System (ORTS) and the Integrated Condition Assessment System (ICAS) by combining an advanced ecological approach to supporting maintenance display visualization design. We seek to demonstrate its capabilities on the AEGIS deck and would entertain licensing BADGERS technology to lead system integrators (LSIs).
Imagery from submarine sensor masts often suffers from a variety of artifacts, which negatively impacts image quality and performance of downstream processing algorithms. Charles River Analytics, a leading provider of innovative R&D solutions for increasingly complex and important human-systems challenges developed Submarine Imaging Real-time Enhancement (SIREN) to detect and correct these artifacts in real-time, which is currently done manually. Beyond a set of “gold standard” video enhancement algorithms and novel artifact removal techniques, SIREN features an image analysis module that detects which artifacts are present and automatically applies the correct enhancement algorithms. Besides submarines, other Navy platforms using EO/IR sensors would benefit from an automated video enhancement system. Legacy security and surveillance systems could use SIREN to immediately improve video quality without expensive hardware upgrades.
When escorting ballistic missile submarines (SSBNs), Type Auxiliary Combat General Escorts (T-AGSEs) maintain their relative position using their dynamic positioning (DP) systems, which rely on scanning lasers to measure positions of targets mounted on SSBNs. Laser sensors performance degrades with rain, snow, fog, smoke, bright lights/sun, and/or excessive vessel motion. Charles River Analytics, a leading provider of innovative R&D solutions for increasingly complex/important human-systems challenges developed Sensor system for Precise Automatic Relative-position Keeping (SPARK) using short-wavelength infrared (SWIR) and covert radiofrequency (RF) sensors to track surfaced SSBNs under all plausible environmental conditions, doesn’t require Sailors to go on deck when underway, and meets regulatory standards for system redundancy by incorporating independent systems based on two different physical principles. Opportunities also exist for SPARK in the commercial maritime industry.
Charles River Analytics, a leading customer-focused provider of innovative R&D solutions for increasingly complex and important human-systems challenges, is developing Advanced Mission Display and Planning Tools (AMPT) providing operators of a multi-vehicle (manned and unmanned), multi-domain (air, ground and sea) common control station with decision support for real-time re-tasking and re-planning of multiple assets. Formal analysis efforts have identified task and information requirements for manned-unmanned teaming, which has driven the design of a set of display concepts and a prototyping and demonstration environment used to validate AMPT technology. We seek to fully integrate software into the PMA-281 Common Control System (CCS) for testing and performance validation and verification.
For 30+ years, Charles River Analytics Inc. has provided intelligent systems technology, software tools, and design/analysis services for government and private industry. We are developing an Embedded Architecture for Cyber-resilience (EAC) to protect cyber-physical systems from cyber threats, ensuring fault isolation for both system- and application-level software components and using novel machine learning algorithms to detect, locate, and automatically recover from compromises due to cyber-attacks. EAC is directly applicable to a number of military and commercial systems including satellites (e.g. CubeSat), avionics, unmanned systems, and industrial control systems. Early versions of EAC have successfully detected component faults and predictably and efficiently recovered from these faults to achieve mission success. We seek to license to large system integrators and integrate into Navy embedded control systems.
Modeling Cyber Behaviors to Wargame and Assess Risk (MOC-WAR) helps cyber defenders proactively adapt defenses by modeling adversary socio-cultural behaviors, motivations, and biases. Our hybrid approach combines the strengths of various computational models and shores up their weaknesses to enable advanced intelligence analysis. Our objective is to seamlessly integrate MOC-WAR into commercial tools already favored by analysts and defenders to provide a cost-effective solution. MOC-WAR is currently at TRL 3, based on our Phase I prototype implementation. We are currently expanding its modeling capabilities to include active goal management. Charles River Analytics is a provider of innovative R&D solutions for increasingly complex and important human-system challenges. Ideal partners for MOC-WAR transition are developers and/or acquisition communities that are creating tools for cyber defenders and analysts.
Despite advances in automated instruction, virtual environment-based (VE) shiphandling training currently depends on expert trainers working one-on-one with students, which creates significant bottlenecks in the VE-based shiphandling training process. The Shiphandling Educator Assistant for Managing Assessments in Training Environments (SEAMATE) provides instructors with the capability to supervise larger student cohorts and dynamically track student performance through event-based, automated performance feedback and efficient, contextualized alerts. SEAMATE’s instructor-oriented user interface (UI) supports at-a-glance awareness of multiple student performance, provides dynamic feedback to direct instructor attention, and improves instructor awareness of individual student progress and intervention needs. Charles River Analytics’ goal is to transition this technology into government and prime contractor virtual training systems for surface, subsurface, air-vehicle control training, and virtual-environment based part-task training systems.