Micromobility Data Analysis: Interview with movmi's Jie Chen
Tell us a little bit about yourself and background.
Jie Chen has a MSc in Computer and Mathematical Sciences by Tohoku University (Japan), a MSc in Mathematics by McMaster University and is looking forward to getting her Math PhD in Summer 2023 (McMaster University). In 2021, 2022, she served as sessional faculty at McMaster University. As the developer of the topology database flatknotinfo.com, she is keen to apply her experience in math research and transportation data analysis to innovative solutions.
Watch the full interview where Jie breaks down the findings from her data analysis work for movmi below. 👇
What work have you done for movmi?
Jie analyzed trip data available from Mobi by Shaw Go’s website and supporting cycling infrastructure data from Open Street Maps. The data was from August 2022 to January 2023, separating e-bike and non-ebike trips to see if there is a difference in trip patterns and usage. Jie then used open-elevation to query the elevation of each station, used networkx, osmnx packages to pull out shortest paths and Tableau for data visualization.
Some of the questions we were trying to answer with this data analysis was,
- How do e-bikes change trip patterns?
- Where is the most demand for electric bikes?
- How do e-bikes positively impact the residents? With an aim to gain infrastructural support.
We believe that Jie’s findings will serve as a baseline comparison with those related to providers in other markets in the future.
What were some of your findings from the Mobi by Shaw Go data?
Trip Volume By Routes
Last August, Vancouver’s public bike-share system Mobi by Shaw Go boosted its service by adding 500 e-bikes for the first time. The new pedelecs can be charged at 30 e-bike stations and bring the Mobi fleet to 2,500 bikes and 250 stations.
After analyzing the data available, Jie discovered that Non-ebike riders tend to stay within their neighbourhoods whereas e-bike riders venture outside and into other regions. This indicates that e-bikes provide more transportation options to individual communities within the city.
The data also showed that neighborhoods located on the outskirts of the city have a larger proportion of e-bike trips, especially during the winter months. The further distance from the downtown core leads to longer and more frequent e-bike trips.
Mobi E-Bike Volume at 6am: City Outskirts
Interestingly, the data shows that at 6am, 1/2 of trips from Grandview-Woodland, 3/4 of trips from Kensington-Cedar and 3/4 of trips from South Cambie are e-bike trips. This is shown in the the Fall/Winter data (November to January) that we had access to.
6am is also the peak of e-bike volume, implying e-bikes are a preferred alternative to public transportation. This preference is stronger during the winter season, suggesting e-bike may provide higher comfort level than the usual mechanical bikes.
E-bikes have the potential to increase social equity within the transportation network. The data shows that the people who live on the outskirts of the city, those who may not be able to afford accommodation closer to downtown, have a preference for the e-bike system. We can also see from the data that Grandview-Woodland, Strathcona and Kensington-Cedar have the largest community pass proportion. Having a discounted e-bike price would probably increase demand ever further and offer more people an alternative to public transportation.
Trip Volume by Elevation Change
Not surprisingly, the elevation pattern within the city mirrors the topographical features of Vancouver. Riders prefer e-bikes for uphill trips. For example, downtown is at lower elevation compared to the neighbouring areas. If we take a look at e-bike departures, we can see that usage is higher in the areas with lower elevation – meaning, more people are choosing to use the e-bikes when venturing from downtown into higher elevated areas, compared to the mechanical bikes.
What were some of your findings from the Hello Scoot data?
Loop Trip Volumes (Clustered Data)
Hello Scoot is a dockless electric moped scooter service located in Tahiti, French Polynesia. It is a solution that unclogs the roads in the most ecological way, while offering the island a simple and practical mobility solution.
In order to better understand the service, Jie clustered together data points. From the original data, it showed that just under 10% of trips were loop trips. However, if you take the data from individual users and track their trip behaviour, it shows that 40% of trips follow the A-B, B-A (looped) pattern. This means that 50% of trips are looped which is quite unique and interesting for a dockless service.
Loop Trip Volumes (Clustered Data)
When exploring the most popular and occurring short trip paths (more than 100) from the data, Jie also discovered some intriguing findings. Of the top four trip paths, two of the highest occurring paths were two way (East to West, West to East) one path was a one way route (to the east) and the last highest occurring path was once again looped. Using this data, Jie noticed that many hotels sit along these high occurring paths, including the looped path.
An explanation for the high number of looped paths from this dockless scooter service might be due to the lack of public transportation in the region and a high level of active users, either commuters or tourists visiting the island. Perhaps a next step for Hello Scoot would be to consider partnering with hotels on the island to establish a station-based network.
