Kepler Deliverable - CV2X Geospatial Visualization Platform
Technologies: React, Redux, Node.js, Express, PostgreSQL/PostGIS, Kepler.gl, Mapbox GL
Collaboration: UTC Center for Urban Informatics & Progress - Chattanooga CV2X Infrastructure
- Architected full-stack dashboard with React 18, Redux state management, and Kepler.gl 3.2.5 for real-time CV2X vehicle and VRU tracking
- Designed PostgreSQL database with PostGIS GEOMETRY types, GIST spatial indexing, and stored procedures for ST_Contains/ST_Intersects queries
- Engineered RESTful API with Express and JWT authentication handling GeoJSON serialization and CRUD operations for geofences and SDSM events
- Built analytics dashboard with Recharts for traffic pattern visualization using time-series aggregation and JSON metadata queries
Coast Project - Smart RV Modernization System
Technologies: React Native, Node.js, Express, Raspberry Pi 4, CAN Bus, ModbusTCP, WebSocket
Collaboration: Aero Build Company (Nashville, TN) & UTC Center for Urban Informatics & Progress
- Developed React Native application with SAE J1939-compliant CAN bus communication controlling 14+ RV subsystems via RV-C protocol
- Deployed Node.js/Express API on Raspberry Pi 4 with PICAN3 interface parsing RV-C DGN messages for lighting and HVAC control
- Integrated Victron Energy Cerbo GX via ModbusTCP reading holding registers for battery SOC, solar yield, and power consumption metrics
- Implemented WebSocket protocol with Socket.io for sub-200ms command transmission and state synchronization between app and vehicle subsystems
VIGOR: Virtual Intelligent Generative Orchestrator for Mobility Rehabilitation
Technologies: Python, TensorFlow, Keras, Unity, C#, XR Interaction Toolkit
- Trained TensorFlow convolutional neural networks for real-time limb movement prediction using 4D sensor arrays (3D spatial + somatosensory)
- Implemented sensor fusion pipeline with Kalman filtering processing multi-modal input streams for precise motion tracking
- Engineered Unity VR environment with XR Interaction Toolkit for physics-based avatar simulation and inverse kinematics
- Integrated TensorFlow Lite model inference with Unity ML-Agents achieving <50ms prediction latency during VR rehabilitation sessions
CrowdCheckers - Real-Time Occupancy Monitoring Application
Technologies: Swift, iOS SDK, RESTful APIs
- Built native iOS application with Swift 5 using URLSession for RESTful API integration and real-time occupancy data synchronization
- Implemented MapKit framework with custom annotations and dynamic color-coded overlays for venue visualization based on capacity thresholds
- Designed Core Data persistence layer with NSFetchedResultsController for offline caching and optimistic UI updates during network latency
- Integrated UserNotifications with CLCircularRegion geofencing for proximity-based push alerts on crowd density changes