SOCIAL

MENU

GET IN TOUCH

ankitsneh03@gmail.com
Back

OCTANE: Digital Twin Platform

Year

2026

Tech & Technique

Python, FastAPI, SimPy, AWS Fargate, TimescaleDB, Docker, React

Description

A comprehensive Full-Stack Digital Twin platform designed to simulate and optimize city-scale battery swapping networks for the Indian EV market.

⚠️ Note on Live Demo: The frontend is live, but I have spun down the AWS Backend (Fargate/RDS) to save costs after the hackathon/demo period. The architecture and code are fully viewable on GitHub.

Key Features:
  • 🏙️ City-Scale Simulation: Event-driven SimPy engine modeling entire battery ecosystems.
  • 🧠 AI SOC Estimation: Ensemble ML (LightGBM, GRU) for precise State of Charge prediction.
  • 🌡️ Physics Engine: Models thermal dynamics and cooling constraints of batteries.
  • 💰 Greenfield Planner: AI-driven economic optimization for new station deployments.

Technical Highlights:
  • Architecture: Chose AWS Fargate over EKS for 60% lower ops overhead and 40% cost reduction.
  • Database: Implemented TimescaleDB for high-frequency time-series telemetry.
  • Optimization: Built a custom "Out-of-the-Box" discrete event simulation using SimPy that models Indian traffic and festival patterns.

My Role

Backend Lead & Cloud Architect
I architected the entire simulation engine and cloud infrastructure:
  • Simulation Core: Built the SimPy-based discrete event engine handling thousands of swap events.
  • ☁️ AWS Architecture: Designed a serverless-first architecture using Fargate, reducing cold start times and costs.
  • 🔒 Security: Implemented multi-layer security with JWT, VPC isolation, and WAF.
  • 📊 Data Engineering: Managed the flow from raw telemetry to actionable ML inference.

ANKIT SNEH

ankitsneh03@gmail.com