My Commute Data Collection

By | August 29, 2014

I enjoy data. I like seeing details on the mundane, figures on the minutiae and histogram on the meaningless. I track which songs I listed to on Last.fm, how many steps I do with FitBit and for the last two and a half years, have been keeping track of when my Sherwood Park transit bus would arrive in the morning and afternoon.

What started as a way for me to become familiar with the frequency of the buses turned in to a habit and each time a 425 or a 401 bus would arrive in my life I would take a screenshot on my phone and mark the date and time of that specific bus in a spreadsheet.

On September 2 the Sherwood Park transit routes are changing and my routine will have to change. All the data I have collected since April 2012 is now useless, but instead of putting that to waste let’s see what information I can pull out of this spreadsheet.

The timing of my morning bus was captured 454 times and my afternoon bus was captured 375 times. This difference is attributed to not starting to capture afternoon bus time until nearly three months later. There are also times where I would take an ETS bus elsewhere to Edmonton, commute on the LRT to ride home with Jenna, or find another way home. There were times I would drive to the transit center and not need the local commuter so on those days I would only record my afternoon numbers.

The morning information was useful to determine when the bus would arrive at home. The afternoon data was used to see when I typically left work. My average work hours could be determined from below, so a typical day to and from work is from 6:49 AM until 4:44 PM, which is closer to eleven hours than ten when you add the time for the commute time home.

Morning commute data:

  • Earliest bus pick up: 6:23 AM
  • Latest bus pick up: 7:11 AM
  • Average pick up: 6:49 AM
  • Mode: 6:56 AM

I spent the first 2/3 of my commute to TELUS going in at 6:56 AM and started to go in 15 minutes earlier catching the 6:40 AM bus in 2014. This explains why the Average is closer to the Mode since the majority of my mornings started at the later, 6:56 AM, time.

Afternoon commute data:

  • Earliest bus pick up: 2:37 PM
  • Latest bus pick up: 6:41 PM
  • Average pick up: 4:44 PM
  • Mode: 4:22 PM

Number of times caught the afternoon bus:

  • …before 4PM: 8
  • …after 5PM: 100
  • …after 6PM: 14

For the first six to eight months at TELUS I typically caught the 4:21 PM bus. As my responsibilities grew so did my time at the office and I ended up catching a 5:00 PM or later bus. I aim to be logging off between 4:30 PM and 5:00 PM, but I found the afternoon bus schedule to be inconsistent so needing to be in the elevator at a certain time to catch a bus didn’t matter because I would need to wait until closer to 5:00 PM to catch a bus anyway. As someone who strived for consistency in data the variance in the afternoon schedule bothered me, but eventually a bus did arrive and I would always get home.

Let’s view this data in graph form!
Morning commute data for the 425:


Afternoon commute data for the 401:

What does all this mean? Very little. My mornings were pretty set and the afternoon was a gamble on when I would leave and when a bus would be there to take me home. I like knowing when my morning bus arrives and plan my routine around that and the afternoon information is purely for entertainment purposes to see what time I leave the office.

I will continue monitoring the bus schedules in September but will start a new worksheet and maybe in a few years time post a similar summary of my data to see if my working hours have changed at all.

4 thoughts on “My Commute Data Collection

  1. Brendan Young

    Hey Sean, what program/service did you use to make those graphs?

  2. Sean Post author

    @Brendan: I used simple Excel to make the graphs. I had the raw data stored in a separate Excel Worksheet and made a line graph charting time of pick up (on Y-axis) and date of pick up (on X-axis). It took a bit more effort to get the data to appear but that may have been an issue with how the raw data was stored than the graph itself.

Leave a Reply

Your email address will not be published. Required fields are marked *