Charging Station Allocation
Identyfing location factors before expanding to new places
Challenge
One of the leading Electric Vehicle rapid charging networks
Charger utilization is the most important business KPI
Implementing a predictive model to optimally locate chargers was a business priority
Traditional consultancies proved expensive, lacked innovation and featured exorbitant on-going maintenance costs
Designed a scalable data platform with continuous integration capabilities for new data sources
Combined company utilisation data with 500+ data features from open source and private purchased sources (everything from climate to road traffic and car ownership to amenities and socio-demographic factors)
Built a comprehensive predictive ML model that successfully explained a large percentage of the target variables
Visualised results within existing PowerBI instance to minimize costs and maximize usability
Solution
Value
Key business stakeholders have a high degree of confidence in model outputs
Model directly informs whether sites are approved for build or not
New data sources can be easily added and automatically fed into the model, thanks to the flexible and thoughtful design of the solution
4x cheaper than alternative solution and a better outcome achieved
Roles
Cloud Engineer, Spatial Data Scientist, Machine Learning Engineer, Project Manager
Tools
Sector
Automotive, Energy & EVs