Systems for
complex operations.

ECHELON analyzes images from active incidents with computer vision, running object detection and scene-level analysis across structural damage, hazards, crowds, and traffic. It then calculates risk to nearby infrastructure assets, traces the impact across dependent and cascading assets, and fuses the evidence into a final conclusion with recommended actions.

• FastAPI / API ArchitectureCore
• Computer Vision / OpenCVCore
• YOLOv8 / Object DetectionCore
• WebSockets / Real-Time SystemsCore
• PostgreSQL / SQLAlchemy
• Docker / Pytest

ATLAS converts aerial imagery into semantic terrain maps and rover routes. Mask2Former runs locally with PyTorch to classify roads, open ground, and obstacles. The application builds a weighted traversability grid, plans routes with A*, and reruns perception on selected route corridors before replanning.

• Python / PyTorchCore
• Mask2Former / TransformersCore
• Computer Vision / Semantic SegmentationCore
• A* Path Planning / Traversability CostmapsCore
• FastAPI / WebSockets
• JavaScript / Canvas

A real-time warehouse fleet simulator coordinating multiple robots through shared space. A* routing, cell and edge reservations, yielding, and replanning prevent collisions; a traffic- and energy-aware scheduler improves throughput while battery management, failure recovery, persistence, and live event streaming keep operations resilient and observable.

• A* / Collision-Aware RoutingCore
• Traffic + Energy SchedulingCore
• FastAPI / WebSocketsCore
• React + Three.js / 3D UICore
• SQLite / State Recovery
• Docker + AWS EC2 / CI

A warehouse logistics prototype that turns requests, inventory, fleet capacity, and live routing into clear delivery commitments for food-bank operators.

• FastAPI / Python BackendCore
• OSRM / Route OptimizationCore
• MapLibre GL / Geospatial UICore
• Decision Engine / Fleet AllocationCore
• REST APIs / Async HTTP
• Pytest / Systems Validation