Research Projects

Safe navigation for manned and unmanned platforms across marine, aerial, and ground domains requires advanced situational awareness. These platforms rely on a range of sensors, such as cameras, radar, sonar, LIDAR, GNSS, IMU, AIS, DVL, etc. alongside environmental sensors (e.g., pressure, temperature, and anemometers), depending on the platform domain. In manned systems, operators manually monitor sensor feeds, but fatigue and errors in judgment can compromise performance. Unmanned systems require real-time onboard sensor data processing to create situational awareness grids for effective navigation and mission planning. This project addresses these challenges by integrating AI/ML techniques with sensor fusion to process data from multiple sensors. Incorporating environmental sensors (e.g., wind, waves, pressure, and temperature) will further enhance understanding of the operating environment, enabling improved mission planning and optimization. Another critical challenge is navigation in GNSS-degraded or denied environments, such as indoor or military zones, where GNSS signals are disrupted by spoofing, jamming, or poor reception. The project will develop radar-, sonar-, and vision-based Simultaneous Localization and Mapping (SLAM) algorithms tailored to specific domains, providing robust navigation solutions under such conditions. To ensure independence from OEM software and SDKs, the project will develop proprietary systems to process raw sensor data, including radar images and sonar waterfall plots. These systems will be integrated with ML models for precise object detection, classification, and anomaly detection. Extensive datasets from both open and closed sources will be curated to train these algorithms for Automatic Target Recognition across multiple sensors for diverse operational scenarios and domains. The project aims to develop a domain-specific sensor suite capable of object/target detection, localization, collision avoidance, anomaly detection, and mission plan optimization. Building on the PI’s previous work in vision-aided speed modulation for high-speed marine crafts (Unmanned Surface Vehicles - USVs), the research will further enhance mission planning, structural integrity, and operational safety under varying sea-state conditions. Deliverables include: 1. A modular sensor suite adaptable for multi-domain platforms. 2. Case studies to validate its applications: o A USV equipped with a sensor suite for navigation in open water bodies. o An Unmanned Aerial Vehicle (UAV) or Unmanned Ground Vehicle (UGV) demonstrating vision-based SLAM in indoor environments. The project will culminate in proof-of-concept demonstrations, showcasing its contributions to improving situational awareness, autonomous navigation, and mission planning in civil and defense applications. By achieving these objectives, the research will advance the fundamental understanding of environmental perception and its practical applications across multiple domains.