These days, curbs are for more than just parallel parking. With the rise of e-commerce and ride hailing, and with all of the new shared vehicles cropping up, they’re becoming the key to understanding city streets. Curbs are used for picking up and dropping off passengers, delivering packages, docking bikes, and getting on scooters. Particularly as all of these activities become more digital, it is essential that people are able to know how, when, and where they can use the curb.
The crazy part about all this activity is that, in most American cities today, nobody has a digital record of what the curb rules are at any given time or place. Not the city, not the delivery company, not the ride-hail company, no one. The result: more congested, complicated, inefficient and dangerous streets for anyone trying to get from point A to point B.
When we set out to build a parking solution for Google Maps, we assumed that this data would be readily available. We didn’t necessarily think there would be a well-designed API but we at least thought the data would exist. We were wrong. When we tried to figure out how to solve the challenge, we found consultants who specialize in this type of collection. They take to the streets with a wheel, pencil and clipboard. Given the level of detail and the number of features that we wanted them to collect, the cost was astronomical.
Like us, you may be thinking there must be an easier way. Street View imagery! What could be better for solving this problem than using the images Google and others collect along nearly every street in the country? Put some computer vision algorithms on those images and problem solved, right? Wrong! It turns out that when you want precise positioning of signs and the ability to read them, car-mounted imagery is next to useless. Many signs are blocked by trees or other obstructions. Plus, the minor inaccuracies in the position and heading of the vehicle add up, so there’s not a great way to know exactly where they are on the curb. Besides these hurdles, curb cuts and curb paint are not consistently visible.
Because of this, we knew that we had to find a way collect curb data from the sidewalk. The data had to be accurate and the cost had to be reasonable. When we worked with consultants, we knew that despite their time-consuming process, they were good at getting the necessary information. What if there was a way we could use technology to super-charge their work?
So, we decided to build an app and accompanying management system that gives surveying superpowers to anyone. Now, we’re announcing that this system, Coord’s Surveyor, is available for anyone to use. By using Surveyor, not only can we easily find the exact positions of parking signs and other assets in a fraction of the cost and time that traditional methods require, but we can also use this output to automatically produce a detailed and accurate map of the underlying curb rules. Over the past several months, we’ve used Surveyor for collecting data on more than 12,000 curbs across both San Francisco and Los Angeles, and piloting its external use with organizations, such as Streetline, who are interested in developing a digital record of curbs in other cities. According to Mark Noworolski, CTO of Streetline: “Streetline relies on accurate curb information to provide its real-time parking availability solution. We were thrilled to have had the opportunity to field-test Surveyor which can supplement our policy gathering solution with additional metadata.”
Now, let’s answer some questions you might have about the app itself.
How does the app work? And why is it so much faster?
Traditionally, the surveyor has to measure their position manually while writing down the text and details of every sign. As you can imagine, these are both incredibly time-consuming. With Coord’s Surveyor, the surveyor marks the beginning of the curb by tapping a button in the app and then simply walks down the sidewalk, only stopping to take pictures of signs or to note the location of other curb features, such as curb cuts or curb paint. Once they’ve reached the end of the block, they mark the end of the block in the app. As they do this, the app measures the curb for them. Not only does this speed up the process, but it also makes it much less susceptible to manual mistakes of measurement or sign transcription.
How does the app know the location of the signs?
You might have guessed that the app uses GPS technology, but unfortunately, in many places, GPS is not accurate enough for this purpose. Particularly in cities, GPS is routinely off by 10 feet or more — it could even place a parking sign on the wrong side of the street.
Instead, Surveyor’s patent-pending algorithm uses augmented reality technology initially developed for gaming. It uses your phone’s camera and accelerometer to keep track of your movement, so it can can very accurately position two points relative to one another. So when a surveyor marks the beginning of a curb, that starts the measurement. Then, when the surveyor takes pictures of each curb feature, we can use our technology to position them in relation to one another and snap them to the location of the actual curb on a map. Learn how we built the curb map here.
How to Get Surveyor
As part of the launch, we’re making Surveyor available for free one-month trials through the end of 2018. We’re excited to work with all kinds of organizations, including government agencies, transportation consultants, mobility companies, and more.
Once you start your trial, you’ll be able to install the app on any iPhone. Better yet, the app doesn’t require a cellular plan or uses any data while surveying is ongoing: a connection to WiFi twice a day is all you need. In addition to the app, we’ll give you access to the management tools as well as training materials to ensure your surveying process goes smoothly.
With the growing importance of curbs in our cities, having a digital picture of the curb rules is essential for the future of urban transportation. That is why we built the Surveyor app and why we’re making it available externally.