Predict future demand for Electric Vehicles (EVs) and public access EVs charging stations at zip code level in NY state.
Even though the EVs market is increasing due to the tendency of green energy, the infrastructure to fully tranform to EVs is insufficient. This causes range anxiety meaning that people are worried about how far they can travel in an electrical car before its charge is out. This project attempts to address the insufficiency of charging stations by predicting how many charging stations are required to be installed in each zip code in NY. Up to 2020 in the data, the total registered EVs was about 100K, and charging stations were about 3K in NY state meaning 3 charging stations per 100 EVs. The project provides the following information.
Current EVs on route
Current EVs charging stations
Actual EVs vs Prediction of EVs
Actual EVs charging stations vs Prediction of the stations
EVs Tranform is a project that predicts the EVs market and potential charging stations for users (such as state, car companies, and electric companies for installing EV infrastructe) to accelerate the use of EVs.
The main steps involve in the project were;
- Gathering several data
- NY State EV dataset (2011-2020)
- NY State EV charging stations dataset (2011-2020)
- Census dataset (2011-2020)
- Data Wrangling (cleaning, merging datasets, etc.)
- Ploting the actual EVs vs predicted EVs and actual charging stations vs predicted stations
- Creating Choropleth mapping at the zipcode level to show these actuals and predicted values for the year of 2020.
- Python
- Pandas
- Matplotlib
- Geocoder to get lats/longs of addresses
- SKLearn pipelines
- Altair -- Choropleth Mapping, plotting