Mapping Current EV Charger Infrastructure with Web Scraping and Big Data
By Anna Mowat
Realizing the clean energy transition will require a spectrum of social movements that will include, but are not limited to, replacing internal combustion vehicles with their electric counterparts, creating net-zero carbon buildings, and replacing fossil fuel plants with solar or wind farms. It is often difficulty to know exactly how and where the adoption of clean energy is happening, because these societal shifts are occurring in real-time and tend to be de-centralized. Therefore, it is rare to find an up-to-date database on a clean energy campaign or initiative, with easy-to-read information, that can keep us informed.
Indeed, one of the great challenges associated with the clean energy transition is the staggering amount of new information generated as the transition occurs. Unfortunately, much of the information is scattered across different websites and locations, and is stored in different formats (pdf files, excel sheets, web pages, and more) which makes analysis very time-intensive. While some datasets and analyses of the clean transition are published regularly—e.g., on a monthly basis—they are more often published annually.
At ACE Partners, we are continuously working on tools to collect and analyze current data on the clean energy transition, in real-time. These tools help us provide evidence-based advisory of policy, communications, network building, and more.
One of the tools we are working on identifies the location of the current EV charging infrastructure of a country. EVs will be a core part of the clean energy transition. Globally, the transport sector is responsible for around 40% of energy consumption and 28% of total energy-related CO2 emissions. In 2019, the transport sector accounted for roughly 18% of CO2 emissions in ASEAN. So, moving to electric vehicles in conjunction with switching to renewable energy sources will significantly reduce carbon emissions.
One method to assess the pervasiveness of electric vehicles is to map out how many charging ports are available, where they are, and which companies supply the chargers. Even a quick glance at such a map can give us insight on how easy it is to own an EV in the city or country in question.
Using Python programming code, we can “scrape” data from all of the EV chargers published in google maps and save them as an excel sheet dataset. A more in-depth description of the coding technique can be found in my article in Medium.
“We all know how to find the closest EV charger. All you have to do is just open your phone, type “ev charger” into google maps, and hit search. Boom, now you know where the closest twenty charging stations are.
So, in theory, you could:
(1) run this search,
(2) write down all the relevant search results,
(3) walk 10 km in one direction,
(4) and repeat.
While that is a super effective way to hit your daily step count, it is not realistic. Fortunately, there is a way to write code that convinces Google Maps that you are very into long walks and car chargers.”
-Anna Mowat, Scraping Google Maps to Measure Electric Vehicle Infrastructure: Real-time EV Charger Data V1.0
The initial case study presented focuses on Hong Kong. We collected a list of public EV chargers in Hong Kong and, using Python, automatically saved our dataset into a google sheet. This Python code can be re-run at any time to keep the dataset up to date.
This tool is only an initial step in using data to develop a full picture of EV infrastructure. Following the Hong Kong case study, we could continue our analysis by writing code that scrapes the Hong Kong Transport Department for the list of EV models currently approved for the road. We can then use the list to track the number of EV car companies in Hong Kong, visit each company’s page, and pull relevant EV information from their site (car sales, their EV charger maps, each EV model sold, etc.).
Data scraping tools can also apply to other facets of the clean energy transition, such as renewable energy development, energy efficiency potential, and policy status. While more exploration needs to be done, it may be possible to perform daily scrapes of renewable investment pipeline websites, while also using code to analyse and map grid transformation. For energy efficiency, map scraping may be able give us more accurate information on building area and building efficiency in different urban areas or zones. In short, the ability to use Python for data collection and analysis can open the door for quicker and more powerful analysis of the energy transition.
References:
[1] (7 September 2019). Sustainable Land Transport Indicators on Energy Efficiency and GHG Emissions in ASEAN - Guidelines. Association of Southeast Asian Nations. https://asean.org/storage/2019/03/Infographics-on-the-Guidelines-for-Sustainable-Land-Transport-Indicators-on-Energy-Efficiency.pdf (accessed 17 Feb 2021)
[2] Crippa, M., Guizzardi, D., Muntean, M., Schaaf, E., Solazzo, E., Monforti-Ferrario, F., Olivier, J.G.J., Vignati, E. (2020). Fossil CO2 emissions of all world countries - 2020 Report. EUR 30358 EN, Publications Office of the European Union, Luxembourg, 2020, ISBN 978-92-76-21515-8, doi:10.2760/143674, JRC121460. https://edgar.jrc.ec.europa.eu/overview.php?v=booklet2020
[3] Mowat, Anna (21 Jan 2021). Scraping Google Maps to Measure Electric Vehicle Infrastructure: Real-time EV Charger Data V1.0. https://medium.com/swlh/scraping-google-maps-to-measure-electric-vehicle-infrastructure-real-time-ev-charger-data-v1-0-77a2b4142a78
コメント