Where and when to deploy your volunteer team?

Background

The objective of this project to help WomenTechWomenYes(WTWY) organization strategically deploy their volunteers by time and stations to maximize number of signatures from those who will attend the annual gala WTWF holds at the beginning of each summer.

Methodology

The data sources used for this study are MTA's turnstile data, MTA's station location data and google map's data of a company list suggested by the Grace Hopper Convention.

First, stations within 1/2 mile of the companies are selected because this is a reasonable distance for passengers to walk to their offices, otherwise they may choose another station.

Then, the list of stations is further shrunk by incorporating a weighted score based on the hourly throughput of the station and the number of companies within the range of 1/2 mile to the station. For each company, only the station with the highest score is taken into the final target list of stations where WTWY should send their volunteers.

Next, to better scale the turnstile data, our study group segment hours of a day into three shifts(Morning: 7am - 11am, Midday: 11am - 3pm, Evening: 3pm - 7pm) and aggregate entries and exits data into a matrix of 7 weekdays by 3 shifts(7x3).

Last, for each station in the target list, a schedule has been generated as 3 shifts which rank top three by values in the matrix. These 3 shifts are timeslots when the organization should send their volunteers to the targeted station. Furthermore, heatmaps are generated for visualization purpose.

Result

Now, let's see the target list of stations:

Targeted Station List

What does the final schedule look like?

Final Schedule

Targeted stations and companies visualized by map (The size of red stars reflects Toucan score of the very station):

Target stations(red stars) vs. companies(blue dots)

To better illustrate, we generate the heatmap of "Herald SQ" station as below. As shown in the heatmap, the darker the color of a rectangle box is, the busier the station is at the corresponding shift. We can observe that the Herald Square Station is busy with entry flow on Tuesday, Wednesday and Thursday evening and exit flow on those mornings. It makes sense because technical professions leave the station for work in the morning and go back to the station for home in the evening.

Heatmap of Herald SQ's hourly entries and exits

Conclusion

Instead of sending volunteers to all stations three shifts a day five days a week, it is more efficient to send them to targeted stations at the three shifts proposed in the schedule every week. Totally, there will be an 80% reduction in shifts needed compared to the shift strategy of last year.