Evision - Charging station recommendation

Landing page
landing page content
content
sign in
basic question
location
input detail
genrating
Info
newz
location
final report
2d site
3d site

Inspiration

The rapid growth of electric vehicles is creating a need for reliable and strategically placed charging infrastructure. We were inspired to build Evision to solve the problem of placing charging stations based only on current demand. Our goal is to predict future EV demand and identify where infrastructure will be needed next.

What it does

Evision analyzes EV registration data from VAHAN, existing charging-station data, and geographic information to:

Forecast future EV demand using XGBoost Identify EV and infrastructure hotspots using DBSCAN Analyze geographic factors using GeoPandas Calculate location suitability using MCDM Rank and recommend the best locations for new charging stations How we built it

We cleaned and combined VAHAN datasets by fuel type, vehicle category, and manufacturer to create an EV master dataset. We performed feature engineering and trained an XGBoost Regressor for future EV demand forecasting. Open Charge Map data was used to analyze existing infrastructure, while geospatial analysis and DBSCAN identified potential hotspots. Finally, MCDM combined multiple factors to generate a location suitability score.

Challenges we ran into

Our biggest challenge was handling and integrating large VAHAN datasets with different structures. We also faced challenges in cleaning location data, matching RTO information, processing geographic coordinates, and combining demand forecasts with existing charging infrastructure.

Accomplishments that we're proud of

We built an end-to-end system that goes beyond simply displaying charging stations. Evision connects EV demand forecasting, infrastructure-gap analysis, hotspot detection, and location ranking into one decision-support platform.

What we learned

We learned how to work with large real-world datasets, perform feature engineering for time-series forecasting, apply XGBoost for prediction, use DBSCAN and GeoPandas for spatial analysis, and combine multiple factors using MCDM to support real-world decisions.

What's next for Evision – Charging Station Recommendation

Our next goal is to make Evision more useful for investors, charging-network operators, and governments by adding real-time data, improved traffic and grid information, charging-station utilization, ROI/payback estimation, and interactive investment reports. This will help users understand not only where to build, but also why a location is a good investment.