Support BikePortland

PBOT places bet on ‘smart city’ tech to count bikes and make streets safer

Posted by on April 19th, 2018 at 10:48 am

PBOT will install 200 sensors to gather more data about how streets are being used.
(Photo: J. Maus)

Portland’s bike planning is about to get smarter.

“The city has the right ideas but they don’t have the right tools.”
— William Henderson, Ride Report

The Portland Bureau of Transportation announced two major data-driven projects last week aimed at making biking safer. PBOT purchased data from Portland startup Ride Report, and plans to install smart safety sensors on three of the city’s most dangerous streets.

Together, these projects will flood planners with new data on bike travel patterns (and a whole lot more).

“Now you see more and more data becoming available from traffic sensors, smartphones, bike share,” says William Henderson, one of Ride Report’s co-founders (whom you might recall from our 2015 profile). “We want to help cities effectively use that data.”

Image from recent PBOT announcement.

This type of collaboration is part of the “smart city” approach to urban planning that’s gaining traction across the country. Cities are scrambling to partner with tech companies, and tech companies are grabbing for public dollars. The mobility services sector— think carsharing, mobile ticketing and any project coming out of company’s like Portland-headquartered moovel — is booming.

There’s been plenty of talk about smart city projects for transit and cars, but less for bikes. Portland will be the third city, behind San Diego and Atlanta, to pilot the traffic sensors. Ride Report has made it into 13 cities.

“The city has the right ideas but they don’t have the right tools,” says Henderson. “They don’t move like a tech company. On the other hand, you have tech companies that don’t understand how cities work. It’s important to have something in the middle.”

In the old days, when PBOT wanted to see how many people were biking at a specific location, planners collected surveys and sent out legions of employees and volunteers to manually count them. That process could take up to six months.

The smart sensors automate at least part of the process. Unlike humans, they can count several things at once — such as when and where people cross the street, and how fast they’re going. The sensors’ cameras can even pick out different shapes and sort them into cars, bikes and people, kind of like how Facebook recognizes your friends in photos.

PBOT plans to install 200 sensors on light poles in three high-crash corridors. The agency was restricted to light poles owned by Portland General Electric, one of the project partners. They settled on 122nd between Burnside and Duke, SE division between 11th and 22th, and Hawthorne between 11th and 46th.

Advertise with BikePortland.

The total price tag is just over $1 million. That includes the sensors, installation costs and the software to make sense of all the data.

The sensors will turn the tables on street design. Instead of painting a crosswalk and hoping people use it, planners can do the opposite. They can look at where people actually cross, according to the sensors, and fill in the missing crosswalks.

Capturing the quality of the bike trip

The sensors can tell how fast you’re driving, where you’re going and how you’re getting there. But they can’t tell another critical variable — how the trip makes you feel.

“What you’re not going to get from the sensors is reports on comfort level,” PBOT Communications Director John Brady shared with us in an interview this week. “Just sitting on the sidewalk doing a bike count you can’t get that qualitative data.”

That’s where Ride Report comes in. With the Ride Report app, people on bikes track themselves with their phones, and report the ride’s overall vibe: “stressful,” “chill,” or “mixed.” Users logged 1.2 million bike trips in the first year. That activity shows up on a user-generated map as a web of red, green and yellow lines.

Ride Report user-generated heat map of Portland central city.

Then planners just have to find the fattest red lines and steer public dollars in that direction.

You’re probably wondering where the chill rides are, and which routes to avoid. The answers aren’t surprising to anyone who’s spent more than a few days in the saddle in Portland. The danger zones include North Interstate, the southern end of Mississippi Ave., Milwaukie, and Woodstock.

By entering into an official partnership with Ride Report, PBOT planners get added data visualization superpowers not available to the general public. They can see the difference between peak and off-peak hours, for example. Henderson says, “we’ve seen facilities that are really good during off peak hours, and degrade during rush hour.”

PBOT is paying $25,000 for the data, and spending another $5,000 on staff to figure it out. Henderson says the price tag is probably less than what the agency spends collecting data for a single project.

“They definitely shouldn’t spend money on us before they spend money on bike lanes,” Henderson says. “But you have to spend on data.”

Who gets counted – and how — matters

“This is just one new tool for us.. We are mindful of its limits and the opportunities.”
— Dylan Rivera, PBOT

We don’t know how this project will affect low-income and minority people who choose to bike and walk. At a recent event hosted by Forth, an electric vehicle advocacy nonprofit, OPAL’s deputy director Vivian Satterfield warned against tech-driven approaches to transportation issues. Many low-income people use smartphones, she said, but mostly for text and email. They’re probably not using Ride Report. Planners might need to find other ways to make sure they’re concerns are reflected in the data-driven designs.

And then there are the privacy concerns. At a time when people are re-assessing their relationship with Facebook over its rampant misuse of personal data, it’s likely the public will be skeptical of sensors of any kind.

And Sarah Iannarone, a PBOT Bicycle Advisory Committee member and local activist on many fronts whose daily critiques of Mayor Ted Wheeler seem to be a precursor to another run at his office, offered sharp rebukes to the smart city tact on Twitter yesterday. The “corporatist push toward ‘Smart City Portland’,” she warned, “never serve people over profits.” “Long-term prosperity hinges on climate adaptation and social cohesion, not tech,” Iannarone continued. “We need a ‘Resilient City’ era with a focus on stupid networks like sidewalks.”

PBOT isn’t blind to these concerns. “This is just one new tool for us, that will supplement our existing public involvement and data research,” spokesperson Dylan Rivera says. “We are mindful of its limits and the opportunities.”

Futuristic tools and future projects

Putting a number on subjective feelings like stress, Henderson says, will prove invaluable in the Central City 2035 plan. Staff working on Central City in Motion and projects in east Portland might also make decisions based on Ride Report, a spokesperson said. The agency could spend up to $254,000 over five years.

Right now, Henderson is enjoying sitting at the front of the peloton of bike-focused data startups. Henderson says he doesn’t have much competition. Moovel’s data, for one, is more general and transit-focused.

But Henderson is gearing up for a tight race. He hired two software engineers recently. He’s looking for two or three more. “Biking is really going to heat up,” he says. “We’ve been growing to meet that demand.”

With data pouring in from all sides, the challenge moving forward is figuring out what it means. Most of the money budgeted for the sensor project is going toward analytics. As smart cities grow, so will the need for data analysts and sophisticated statistics software.

“Everyone has so much data,” PBOT’s Brady says. “No one knows how to actually integrate the data to draw any conclusions. The next challenge is developing the analytics we can use.”

Smart City PDX Team Open House and Tech Expo
Where: Community Hall, Portland Community College – Southeast Campus, 2305 SE 82nd Avenue, Portland OR, 97216
When: Wednesday, April 25th, 2018 from 6:00 to 7:30 p.m.
Who: Representatives from the Office of Mayor Ted Wheeler, Smart City PDX Team, AT&T, Current by GE, Intel and Portland General Electric
Representatives from the Traffic Safety Sensor Project Team will be present to explain the project and answer questions. Representatives from Current by GE, AT&T, Portland General Electric and Intel will also be at the event with displays of the sensor technology.
Please RSVP with Anne Hill, Smart City Coordinator, at anne.hill@portlandoregon.gov

— Caleb Diehl, csdiehl16 [at] gmail.com

Never miss a story. Sign-up for the daily BP Headlines email.

BikePortland needs your support.

Greenfield Health
Please support BikePortland.

NOTE: We love your comments and work hard to ensure they are productive, considerate, and welcoming of all perspectives. Disagreements are encouraged, but only if done with tact and respect. If you see a mean or inappropriate comment, please contact us and we'll take a look at it right away. Also, if you comment frequently, please consider holding your thoughts so that others can step forward. Thank you — Jonathan

27 Comments
  • Blake April 19, 2018 at 11:06 am

    I don’t disagree with critiques of the problems that can come from technology implemented in a way that ignores possible problems with privacy and usage but I think it is a bit hyperbolic the way that Sarah Iannarone puts it about a “corporatist push toward ‘Smart City Portland’ [that] never serve[s] people over profits.” (I say that as someone who generally agrees with her, supported her candidacy in the last election and probably would support her again if she runs again).

    The idea that it is an either/or between “climate adaptation and social cohesion” and “tech” is just wrong. To get to the point where we can have a “Resilient City” needs to involve a prioritization of where to put scarce resources in “stupid networks like sidewalks”. The way we allocate resources is unequal today, as it prioritizes those with the most resources (including time) to dedicate to advocacy.

    Technology built on the idea that every death and serious injury is a failure with equal consequence (regardless of race, income, time spent in neighborhood association meetings, or emails and reports sent to PBOT) can help get resources to the right place, and hopefully stop the cycle of “death / advocacy / death / actions”.

    We already know where the biggest problems are — they are in underserved areas with speed-enabled roads — so as long as the technology collects data prioritizing these areas, offsetting implicit biases, maintains privacy, etc it should be encouraged, not denigrated.

    It is not likely to be initially deployed the right way, nor is the idea of data-driven “Smart Cities” initiatives likely to reverse, so it seems more productive to focus on un-biasing the tech, and making sure data is collected and used in a way that explictly prioritizes unwinding existing inequities in our transportation netwrok.

    Recommended Thumb up 0

  • bikeninja April 19, 2018 at 11:28 am

    Once they have collected the Data then we will get targeted enforcement, Right?

    Recommended Thumb up 0

  • Glenn F April 19, 2018 at 11:42 am

    Portland Leaders need to learn ..its not all about Downtown and the inner areas..

    Recommended Thumb up 0

    • da rev April 19, 2018 at 3:40 pm

      when does PDX begin to include other aspects of the “Smart City” concept, besides cycling & PubTranspo?

      Recommended Thumb up 0

    • Kyle Banerjee April 19, 2018 at 4:24 pm

      This — I suspect the overall effect will be to direct more resources to sections that already get too much attention at the expense of areas that need it.

      Aside the fact that people willing to use an app to rate road sections can hardly be assumed to be representative, different sections of road represent totally different kinds of riders.

      For example, you may notice that the riding on Division is better than Clinton. Or that the Tilikum bridge with total separation from traffic and a whopping 40 feet or so of elevation gain gets a yellow stress rating while you can climb to the top of Marquam Hill on much steeper narrow streets with no bike lanes where cars move right along coded mellow green. Interstate from Moda to Mississippi which is flat with a great bike lane and virtually devoid of hook threats is high stress red while if you continue along Interstate further north where there are hook threats galore, parked cars, and no bike lane it’s only yellow.

      Most of sections I find troublesome have no color at all, and I don’t see other cyclists when I’m on them. If you want more riders, attention needs to be focused on places other than the areas that already support plenty of riders.

      Recommended Thumb up 0

  • Todd Boulanger April 19, 2018 at 12:10 pm

    This is a second big bike data step for Portland (first being annual network manual counts in the 90s/00s)…as I always used to tell our traffic engineers…you collect the data on the problems that you want to solve…and no data means no interest.

    Recommended Thumb up 0

  • Scott Mizée April 19, 2018 at 12:17 pm

    Speaking of smart city, what is going on with the Cycle Oregon donated bicycle counter on the west end of the Hawthorne Bridge? Last time I rode by it, it did not seem to register my passing.

    Recommended Thumb up 0

    • soren April 19, 2018 at 2:02 pm

      It’s been broken for years…

      Recommended Thumb up 0

      • Scott Mizée April 19, 2018 at 2:03 pm

        wow…. that’s a shame. Obviously, I have not ridden over the Hawthorne much in recent years….

        Recommended Thumb up 0

    • q April 19, 2018 at 9:09 pm

      They need to fix it to make that counter productive.

      Recommended Thumb up 0

  • Ned April 19, 2018 at 2:52 pm

    Protected cycle tracks on Hawthorne please!!!!!

    Recommended Thumb up 0

    • Doug Hecker April 20, 2018 at 1:03 pm

      On Hawthorne? We already have matching greenways that are up for or have been “improved.”

      Recommended Thumb up 0

      • soren April 20, 2018 at 1:25 pm

        the horror of bike infrastructure improvements…

        Recommended Thumb up 0

  • Tom April 19, 2018 at 3:55 pm

    Who makes the smart sensor units?

    Recommended Thumb up 0

  • Buzz April 19, 2018 at 5:32 pm

    Are these going to be like those sensors on the self-driving car in Arizona that failed to see that pedestrian?

    Recommended Thumb up 0

  • Resopmok April 19, 2018 at 6:11 pm

    No one knows how to use the data we have.. great.. when is it time to stop spending money and brainpower on collecting and start spending on actually building bike infrastructure that’s ALREADY BEEN PLANNED, like the BMP that’s sitting on a shelf gathering dust? Too many people in this kitchen are writing recipes and there’s not enough cooks to feed the people.

    Recommended Thumb up 0

    • Hello, Kitty
      Hello, Kitty April 19, 2018 at 8:08 pm

      Creating plans is fun. Caring them out… not as much.

      Recommended Thumb up 0

  • B. Carfree April 19, 2018 at 10:04 pm

    “The sensors will turn the tables on street design. Instead of painting a crosswalk and hoping people use it, planners can do the opposite. They can look at where people actually cross, according to the sensors, and fill in the missing crosswalks.”-JM

    This sounds somewhat like kin to the old line about building a bridge if and only if you see people jumping in to swim across. Painted crosswalks may be much more useful at places where people aren’t choosing to cross, but the lack of a painted crosswalk may discourage them at that crossing.

    The data will be real. Determining what it means is going to be subject to incredible errors.

    Recommended Thumb up 0

    • q April 19, 2018 at 10:25 pm

      Exactly. Are people crossing somewhere because they want to? Or because they can’t cross at some other location that they’d prefer?

      Plus, after figuring out what it means, the biggest issue of all–what to do with it. Will it be, “Look at all these people crossing here–we need to put in a crosswalk”? Or, “Look at all these people crossing here–we need to put in a barrier.”

      Recommended Thumb up 0

  • Sam Churchill April 20, 2018 at 8:07 am

    I hope Portland doesn’t use an IoT system like AT&T’s Licensed Assisted Access (LAA) in the 3.5 GHz band. That would require end users to also become subscribers to AT&T’s current LTE network. Many carriers want to control access to “free” spectrum…not encourage it.

    A better solution, that is available now, is shared Multefire radios, using the unlicensed 3.5GHz (CBRS) band. Shared spectrum. Shared radios.

    Multefire uses the “free” 3.5 GHz band but can be used by any small business or individual. No license required. Phones and devices that use the MulteFire standard can seamlessly roam to any carrier when a user leaves the local service area.

    The Citizens Broadband Radio Service, in the 3.5 GHz band, is similar to WiFi. Some 80 MHz is available (free) to everyone, with 70 MHz is available to anyone who wants to buy a license. When LTE is used over WiFi (and 3.5 GHz), it can be more reliable than ordinary WiFi connections.

    Narrow Band IoT in 3.5 GHz enable cheaper bike tracking that uses much less juice then current systems which use an entire cellular channel.

    Let’s not turn our bus shelters and light poles into cell towers, each “owned” by different carriers. One radio can be shared by competing carriers. It benefits everyone.

    http://www.hayden-island.com/smart-neighborhoods/

    Recommended Thumb up 0

    • Sam Churchill April 20, 2018 at 1:06 pm

      Here’s more on P-DOT’s IoT plan.

      https://www.portlandoregon.gov/transportation/76735
      https://www.portlandoregon.gov/transportation/76736
      https://www.portlandoregon.gov/transportation/76738

      The total project budget is $1,012,000. This includes the cost of the sensors and the mast arms to hold them; the installation of the sensors; and the creation of the Portland Urban Data Lake (PUDL). The data lake includes the software and programming needed to collect and analyze the sensor data.

      The budget has been funded through the following sources:
      – $426,000 in general transportation revenue (GTR).
      – $461,000 in Transportation System Development Charges (TSDC). TSDCs are fees that developers pay when they build new developments in Portland. (More about TSDCs can be found here).
      – $125,000 from the project’s private sector partners. These partners also made in-kind contributions related to the handling of the data and the development of the software to analyze the data.

      I think it’s great that AT&T, Intel, PGE and Moovel are all working together on this project. I just hope that light poles and bus shelters will use technology like MuLTEfire in the future so that many different carriers can share one radio and one light pole on an equitable basis.

      I hope someone is writing a grant for shared 3.5 GHz IoT. Salt Lake City received a $4.3 million NSF grant to create a city-scale 3.5 GHz network. NYC is doing it too.

      Recommended Thumb up 0

  • soren April 20, 2018 at 8:57 am

    The agency could spend up to $254,000 over five years.

    How about building the fully-funded 100s, 130s, 150s, THOP, and 4M neighborhood greenways before buying planners/engineers expensive toys. These fully-funded projects have been repeatedly delayed while inner PDX projects move forward.

    Recommended Thumb up 0

  • paikiala April 20, 2018 at 12:19 pm

    “PBOT is paying $25,000 for the data, and spending another $5,000 on staff to figure it out. Henderson says the price tag is probably less than what the agency spends collecting data for a single project.”
    BS.
    No way hose counts on a greenway project cost $25k.

    Recommended Thumb up 0

  • Ron Kopald April 20, 2018 at 12:24 pm

    Anybody know why the northbound parking on Water near Bunk was made verboten? Cleaning up sight lines? That can be iffy sometimes wuth the big trucks rolling through. Most play nice though.

    Recommended Thumb up 0

  • Doug Hecker April 20, 2018 at 1:05 pm

    The placements of these sensors seem to be misplaced and probably misguided.

    Recommended Thumb up 0

  • paul h April 20, 2018 at 1:50 pm

    I’m skeptical of Ride Report’s data. There’s nowhere in the UI to tell it where stressful situation occurred, only that it happened somewhere along my, e.g., 19 mile ride.

    So I just mark all of my rides as chill, b/c well, they nearly always are.

    Recommended Thumb up 0