A Journey to Change the Electric Vehicle Charging NetworkHelen TangBlockedUnblockFollowFollowingApr 16In January 2018, California Governor issued Executive Order B-48–18 calling for 1.
5 million electric vehicles (EV) and corresponding charging infrastructures by 2025.
EV charging network providers are still exploring new opportunities and making tradeoffs between market acquisition and huge costs.
Image source: https://images.
app.
goo.
gl/RbugGktch9zWFTkX7Where the Story BeganAs the UC Davis M.
S.
Business Analytics practicum team working with the Institute of Transportation Studies Davis (ITS Davis), we are expected to provide strategic ideas on the charging network optimization based on our analytics skills and business acumen.
The problems include:How many chargers are needed to serve the future number of EV usersWhere should the chargers be installed within California?Different from other university capstone projects which give a scenario and ask students to provide an ideal plan, our project is facing the real world.
We talk with our clients, research and observe the industry, adjust our work process according to the changes and uncertainties, and also try to make real impact on the business.
An Ambitious StartThe first time we met ITS, we were introduced lots of industry domain knowledge and meanwhile, we felt super excited and ambitious to work on this project.
We saw this a great opportunity to apply our analytics skills as data professionals and believed that we were going to change the industry！Image source: https://images.
app.
goo.
gl/xmyND3XhvP1hkRVMATo make our work more efficient, we assign different roles based on previous experience and skillsets, including team leader, data analyst, and business analyst.
We were soon given the first trial to work on the sample dataset to identify EV user charging patterns.
We brainstormed and decided to focus on the time and location patterns reflected from the dataset.
With the suggestion by our practicum professor, we also used Trello to track our weekly progress.
Applying R and Tableau learned from the classes, we found that the charging stations at a specific grocery store had the highest number of EV users and users tended to charge from noon to night rather than in the morning.
The findings were also agreed by ITS.
It was like a first step for us to success!But soon, we realized that it was really a small step.
A Journey of Failing and LearningThen we were given a dataset of 3 million rows to understand the EV customers.
We tried to upload it to Mysql and after 8 hours’ uploading, it failed many times.
The challenge came not only from the technical part but also from the business aspect.
‘What questions are we going to solve about the customers?’ We expected to tell a great story based on the data but we found it difficult to frame the problems.
There were too many aspects to discuss on and we felt lost when we tried to cover all the problems.
After several discussions within team and with our mentors, we decided to focus on two tasks first.
The first task was to segment customers based on products and conduct an exploratory data analysis on the charging behaviours of different segments.
Meanwhile, we were looking for a geocoding tool to calculate customers’ traveling distances to charge.
Without any experience with geographic information systems, we researched different possible methodologies.
Thanks to the Big Data and Cloud introductions by our professors during classes, we learned that Google Geocoding API would be a good choice.
Just before our first presentation to ITS, we showed the mentor our slides.
‘You are not telling a story.
These slides are just adding everything together.
You are not collaborating.
’ This was the comment we got.
We then realized that we each worked on different parts individually but never sat together to combine all the results carefully.
For the next 48 hours, we needed to rebuild our story.
Image source: https://images.
app.
goo.
gl/aRjLPcXRuoKVwFjEAWe reframed our storyline together and then helped each other to review the slides and provided advice for improvement.
We also simplified the slides to make them more visual and reader-friendly to our audience.
Fortunately, we achieved the client’s recognition after the presentation.
Always Reflect On What You LearnedImage source: https://images.
app.
goo.
gl/ADRGKAkiPg7Hk91r8Apart from the real-life experience gained from the practicum project, the learning process is also a highlight.
We can always practice technical and business skills, get feedback, and make improvements.
As so far, I want to record the three major learning points:Getting familiar with the data and doing some experimental analysis can help define what questions are important to ask.
Also, framing a business problem and structuring the questions before applying a bunch of analytics techniques on the data can improve the efficiency.
Tasks can be assigned to each team member to allow everyone to bring expertise and creativity.
However, team collaboration and mutual learning cannot be neglected.
Delivering a work product is not just to put every piece of work together, but to reach overall agreement and build a cohesive story.
Business/Data analysts are not data machines to apply programming techniques but thoughtful professionals to deliver business values based on the findings from data.
We have learned quantitative skills from the classes, and we also need to keep practicing critical thinking process and storytelling abilities in practicum project, in which we are making real impact to the business world.
.. More details