Inspiration

EV chargers are computers on a stick. They run complex protocols like ISO 15118, handle payment and identity, and communicate directly with vehicles—making them an increasingly attractive cybersecurity target. Yet everyday drivers have no simple way to tell a trustworthy charger from a risky one.

We wanted to make charger security as intuitive as a map pin: a clear go/no-go decision before you plug in, without overwhelming users with CVEs, firmware versions, or security jargon.


What it does

EcoShield SecureRoute helps drivers find EV chargers that have been screened for common cybersecurity risks.

Driver Features

  • Browse nearby charging stations on a live map powered by the NREL Alternative Fuel Stations API (Orlando demo region)
  • View simple safety statuses:
    • Verified
    • ⚠️ Use with Care
    • Not Recommended
  • Open station details including:
    • Plug type
    • Estimated charging time
    • Last security check
  • Save favorites and recent stations with Firebase Authentication
  • Navigate directly to a charger using built-in directions

Behind the scenes, Google Gemini analyzes real EVerest / ISO 15118 charging session logs and assigns SAFE / CAUTION / COMPROMISED risk profiles that drivers can understand in seconds.


How we built it

Frontend

  • Expo
  • React Native (Expo Router)
  • Google Maps
  • Location Services
  • Light/Dark Theme

Backend

  • FastAPI
  • Firebase Authentication
  • Cloud Firestore
  • Offline-friendly synchronization
  • Secure Firestore Rules

Data Pipeline

  • NREL Alternative Fuel Stations API
  • ~50-mile Orlando dataset
  • Explicit station mapping between NREL station IDs and EVerest log files

AI Pipeline (Google Gemini 2.5 Flash on Vertex AI)

Model-in-the-loop Auditor

  • Reads raw charging session logs
  • Produces structured security verdicts
  • Returns:
    • Risk Score
    • Status
    • Findings

Driver Explainer

  • Converts technical findings into short, plain-language explanations anyone can understand.

Architecture

NREL Stations → EVerest Logs → Gemini Analysis → Firestore → Expo App

Approximately 200 analyzed charging stations are seeded into Firestore for the live demo.


Challenges we ran into

Bridging Two Worlds

Real NREL station metadata and EVerest logs don't share common IDs, so we built an explicit mapping system to ensure AI verdicts appear on the correct charger.

Designing for Drivers

Raw cybersecurity findings are valuable to engineers—but confusing for everyday drivers.

We redesigned the experience around:

  • Traffic-light risk indicators
  • Plain-language explanations
  • Actionable recommendations

Hackathon Integration

Connecting a FastAPI AI backend with a polished Expo frontend under tight time constraints while keeping both stable.

Scaling the Demo

We generated deterministic security profiles so the Orlando map contains hundreds of usable charging stations instead of only a handful of analyzed logs.

Cross-Platform Maps

Making Google Maps feel consistent across both web and mobile while respecting user permissions and location preferences.


Accomplishments that we're proud of

  • 🚗 A working Gemini-powered security auditor that produces structured SAFE / CAUTION / COMPROMISED verdicts from real charging session logs.
  • 🗺️ A driver-first experience that feels like a navigation app—not a cybersecurity dashboard.
  • 🔄 A complete end-to-end pipeline:

NREL → Gemini Analysis → Firestore → Expo App

  • 📍 Hundreds of Orlando charging stations with meaningful security labels for a live demonstration.
  • 💡 A product that's easy to understand:

Find a charger you can trust before you plug in.


What we learned

  • EV charging security is driven by protocol behavior (authentication timing, ISO 15118 handshakes, RFID replay attacks, tamper events), not just known CVEs.
  • Google Gemini can serve as a structured security analysis engine—not just a chatbot.
  • The hardest challenge wasn't detecting cyber threats—it was translating them into decisions everyday drivers can confidently act on.
  • Practical Firebase patterns including:
    • Public charger catalog
    • Owner-only user data
    • Offline persistence
    • Secure database seeding
  • Building a polished cross-platform Expo application with a seamless mapping experience.

What's next for EcoShield SecureRoute

  • 🌎 Expand beyond Orlando to additional charging networks and regions.
  • ⚡ Move from batch processing to near-real-time charger health monitoring.
  • 🔌 Integrate directly with charger operators and EVerest-compatible infrastructure.
  • 🤖 Add personalized recommendations based on:
    • Preferred connector types
    • Trusted charging networks
    • Route-aware safe charging suggestions
  • 🤝 Launch pilot programs with fleets, campuses, and charging providers looking to improve consumer trust through transparent cybersecurity insights.

Built With

Share this project:

Updates