BEIS explores using geospatial data for EV charge point mapping

The Department of Business, Energy and Industrial Strategy recently took part in the OS Map & Hack, to understand the datasets available for EV charging and infrastructure. Members of the BEIS Advanced Analytics team, a central team of data scientists, explored how the team could effectively use and access geospatial data to answer specific issues. The team shared their experience below:

Our participation in the hackathon and materials developed during the hackathon were a learning and development exercise only and not reflective of government policy.

The concept

Before the hackathon and on the first morning of the event, we discussed what we wanted to get out of the experience. Our focus was on developing our own capability for using OS Data and APIs; as we did not have much experience with geospatial analysis, we decided a good outcome would be familiarity with the wealth of data available from OS.

To focus our learning and development, we examined the ‘Taking Charge’ hackathon challenge and aimed to map where charge points currently exist in proximity to business sites (with the secondary idea of evaluating how much space was available at these sites). The idea was to create a basis for evaluating where future charging points could be best installed.

The data and output

We set out by exploring the documentation of the APIs and the OS Data Hub to see what was most appropriate for mapping businesses and charge points. We spent some time, as a team, working through the python data science tutorials provided by OS to acquaint ourselves with the OS Features API.

In investigating the ‘Taking Charge’ challenge, we examined the different facets of the OS Features API (such as ZoomStack and MasterMap layers), as well as the OS Maps API and OS Linked Identifiers API. We also brought in AddressBase data from OS and polygons from the ONS’s geography portal to map businesses, and data from the National ChargePoint Registry on existing charge point locations with python.

We used all this data to calculate distances between nearest existing charge points and businesses, whilst also examining the size of the area of sites that belonged those businesses. Our final output was a jupyter notebook heavily based on the excellent python tutorials published by OS and directed at presenting back what we had learned to our team.

The experience

Overall, the experience of the hackathon was great to introduce us to all the data OS had on offer, as well as how this could be accessed and manipulated in python through the APIs. Members of the team found the experience to be an “enjoyable and challenging experience”, and “a brilliant opportunity to seek out the potential of geospatial data” for our team.

The objective of building an interactive map with multiple data sources was outside of our usual work for many of us - but the hackathon helped us to remove the barriers of doing this type of work and using this kind of data. In a short amount of time, we were able to go from understanding the challenge and the syntactic building blocks to producing a useful proof of concept map with python, having learned about multiple packages, techniques and APIs along the way.

The team used a combination of OS APIs and our Python API wrapper. For more information on how to access and use OS data via our API suite, visit the OS Data Hub and view our tutorials and documentation. Public sector organisations can access our geospatial data and services when registering for the Public Sector Geospatial Agreement (PSGA).