Background

This is a DIY traffic monitoring project, using open-source software from developer Claude Pageau, who graciously published his package for all to use on github at https://github.com/pageauc/speed-camera. Thanks Claude! This data was collected with a calibrated webcam, but this is not as precise as a radar or laser installation that the town would use to collect similar data, so it may be taken with a grain of salt. In aggregate however, these readings are highly likely to convey the correct speeds and volumes on the street.

My motivation to collect this data is driven by frustration with the Brookline Transportation Dept’s approach to traffic calming, and is intended to determine whether and to what extent speeding occurs on Beals St, where I live in Brookline with my wife and two small kids.

Beals submitted a petition more than 6 years ago, per the Transportaiton Department’s Traffic Calming program, requesting that we be considered for traffic calming measures. It seems a slam dunk per the program’s ranking criteria (the JFK house draws pedestrians all summer, kids walk to school on the street, and it is used as a cut-through to Babcock, is required for many motorists to access FRR school during drop-off and pick-up hours due to the many one-way streets in North Coolidge Corner). Since then however, our street’s request has sat in an ever-growing queue of similar requests that have been prioritized, seemingly arbitrarily, over ours.

One approach to pushing for the town to address our street’s safety concern is to demonstrate that it is valid. The data here contains a week of traffic statistics observed over the course of March 8 - 15, 2022.

Key Findings

Lots of people don’t speed down Beals. But lots of people do speed down Beals.

  • The 50th percentile speed is 23.44 MPH
  • The 85th percentile speed is 30.70 MPH
  • The most traffic occurs during school pick-up and drop-off times at FRR
  • During weekdays, about 700 cars per day drive down Beals St
  • At its busiest times, well over 50 cars per hour drive down Beals St
  • Every day about 90 cars drive faster than 30 MPH down Beals St
  • Every day about 17 cars drive faster than 40 MPH down Beals St (I even clocked a few people a day who approach 50 MPH)
  • Some people are great! Every day about 200 cars drive below 20 MPH. Thanks folks!

Volume

Let’s start with some basics: how many cars drive down Beals every day and during what hours?

This verifies Beals’ status as a cut-through: significantly higher volume on weekdays than on weekends, with Tuesday, Thursday, and Friday seeing well over 700 cars down the street.

The by-hour data also verifies Beals is used to access the FRR school dropoff and pickup zones, with additional, more muted spikes in volume during traditional rush hours.

Speeds

Now let’s dig in and look at the overall distributions of the speeds of those cars. The key numbers the town uses to assess whether traffic calming is needed include volume and 50th vs 85th percentile speed. For Beals our 50th percentile speed is 23.24 MPH, while the 85th percentile sits at 30.62 MPH, well above what one would want for a residential street. According to AAA’s pedestrian risk analysis from 2011[1], the risk of a fatal impact is 10% at a speed of 16 mph, 25% at 23 mph, 50% at 31 mph, 75% at 39 mph, and 90% at 46 mph.

According to my data, more than half of all drivers Beals exceed 23 MPH down Beals, and 2.6% exceeded 39 MPH.

Ideally the 50th percentile is at or below 20, with an 85th at 25 or lower.

Is it not worth the minor inconvenience and <$10k installation costs to elminate the most lethal driving speeds from this densly packed residential street?

Clearly a lot of drivers are taking it slow down Beals, but there is an alarming number that travel over the legal limit of 25, and the lethal buffer of 30, where the likelihood of a fatal incident is considerably higher. On every day of the week except Saturday (huh, guess people are nicer on Saturdays…), the 85th percentile speed down Beals is well over 30 MPH. Not great.

The uniformity of the distribution reflects my experience of the street though: There is a breadth of speeds! Some people care, some people don’t :)

One more question I have is whether people tend to speed more at night or not. My anecdotal experience is that they do, but let’s see if the data bear that out: In fact, nighttime driving appears to have, if anything, slightly lower typical speeds, with 50th and 85th percentiles coming in slightly below daytime driving values. My personal experience is not validated by the data, and my just be the bias I experience spending more time at home at night, and noticing speeding cars more easily during quieter times or when it disturbs my sleep.

This concludes my review of this very interesting data. Thanks again to Claude, and please share this with members of the T-board if you know anyone!

  1. https://aaafoundation.org/impact-speed-pedestrians-risk-severe-injury-death/