Technology Areas

Dynamic Search Region Calculation for Search and Rescue Planning

Knight Scientific Systems uses advanced mathematical concepts to combine topological map data with human performance models, producing dynamic maps of where lost people could be found.  The image on the Left shows a probability density of where the lost person is likely to be found (red), and the image on the Right shows the growth in time of the maximum travel boundary (red boundary) and encompassed paths (green lines).

Detection Persistence Applied to Search and Rescue

Modern Machine Learning networks excel at detecting humans under benign conditions, but show dropouts in coverage when faced with realistic scenarios.  Recent research at Knight addresses this shortfall to improve detector performance for search and rescue applications, maintaining track of people even when they are temporarily obscured by the terrain.

3D Models from Full Motion Video

Combined with the proliferation of imaging sensors across the world, structure from motion approaches provide the ability to generate three dimensional models for vehicles, buildings, and landscapes of interest.  Knight Scientific Systems uses modern processing approaches to deliver 3D models from photographs and videos.

  • Visual Odometry and Photogrammetry
  • SVO
  • REMODE
  • Gaussian Splatting
  • NERF
  • SLAM

Multi-Spectral Object Detection

Traditional Object Detection (OD) algorithms are trained only on visible imagery, meaning the ability to detect vehicles, people, signs, etc., is only mature for visible imaging systems.  Because Mid-Wave InfraRed (MWIR) and Long-Wave InfraRed (LWIR) sensors play an important role in detection and surveillance missions, Knight is developing multi-spectral OD tools to bring the power of machine learning to all imaging wavebands.

Machine Learning Enhanced Detection, Tracking Identification and Characterization

Classical approaches to processing intelligence, surveillance and reconnaissance (ISR) sensor data are significantly improved when enhanced with modern Machine-Learning technologies like feature detection and classification. Our team is at the forefront of research and development of AI/ML applications to military missions. Our AI/ML activities include the following processes and tools.

  • Data Collection and Feature Extraction
  • ImageDet
  • YOLO
  • Tensor Flow
  • PyTorch

Sensor Data Processing and Exploitation

As sensor technology matures and sensor systems proliferate, data processing and exploitation must be automated and accelerated.  Our team excels at designing and deploying modern algorithms for data fusion and information extraction.  Our experience spans the following topic areas.

  • Technology Evaluation for Novel Sensor Technology
  • Data Visualization
  • Control systems, Detection, Tracking, Closed-Loop Tracking
  • Multiple Hypothesis Tracking
  • Target Recognition
  • Target Signatures, BRDF, Atmospherics
  • SWIR/MWIR/Visible Image Analysis and Data Fusion
  • Drone-based Imaging, Detection, Tracking, and Characterization
  • Multi-Modal Sensing and Data Fusion

Digital Engineering

Our team uses the full ecosystem of digital engineering tools to provide value to our customers; from modeling systems using SysML, to building databases with Object Role Modeling, to simulating multi-domain missions and engagements.  Digitial Engineering efforts include:

  • SysML Modeling for Complex Systems
  • Target Vulnerability and Aimpoint Analysis
  • Database Design and Implementation
  • Object Role Modeling for Data Schema Design
  • Scalable Database Solutions
  • Cameo Systems Modeler

Multi-Domain Modeling, Simulation and Analysis

Multi-domain simulations provide a valuable context for demonstrating the impact new technologies have on the battlefield.  We are expert users of the Department of Defense’s framework AFSIM: Analytic Framework for Simulation, Integration and Analysis.  Our team uses a variety of tools and approaches to predict system performance:

  • DoD Multi-Domain Simulations in AFSIM 
  • Targeting Timeline Analysis
  • Sensor Coverage Predictions
  • Satellite Constellation and Sensor / ISR Performance Predictions for SIGINT and EO/IR Pathways
  • Design, Development, Testing and Delivery of Software Tools
  • Algorithm Development
  • Data Visualization
  • AFSIM Plugin Development

Path Planning

Autonomous path planning is a critical technology for drones, manufacturing robotics, self driving vehicles and space vehicles. We have extensive experience in path planning for robotic linkages and quad-rotor drones. We utilize a wide variety of path planning techniques including:

  • Probabilistic Roadmap Methods

  • Rapidly Exploring Random Trees

  • Stochastic Reachability Analysis

  • Reinforcement Learning Based Planning

  • Machine Learning Based Planning

Multi-Spectral Rendering and Synthetic Target Signature Generation

Synthetic signature generation provides an analytical capability for testing targeting, tracking, and other image processing algorithms in a simulated environment.  We use a variety of approaches to generate multi-spectral target signatures as well as the accompanying sensor data output.  We are experienced users of the following technologies:

  • Bidirectional Reflectance Distribution Function
  • Unity Engine
  • Unreal Engine 5
  • Atmospheric Turbulence Models