US Cities’ Smart Parking Projects – Roundtable Highlights

What does the city want from a smart parking solution? It's a question we asked during our January roundtable featuring representatives from three U .S. cities. Westerville, Ohio, San Francisco, California and Philadelphia, Pennsylvania. In 2020, Westerville began deploying ground sensors by fiber to gain data on parking trends. in its popular uptown district. This is an area home to restaurants and businesses. We do not charge for our parking, which is probably unique compared to the bigger cities that we'll be speaking. So we view this as a community benefit, having this new technology where we can actually monitor the parking situation and hopefully encourage people. go directly to their destination and parking and not circling blocks looking for on -street parking. Meanwhile, San Francisco got its start years earlier with a pilot project in 2011 that established a demand -responsive pricing system at 7 ,000 meters. The pilot expanded in 2018 to all the city's 25 ,000 or so meters. Parking rates on every blockchain from $0.50 to $8 an hour. depending upon demand. The data from the sensors fed into an app for drivers and a parking database run by the city. While there were some concerns about people getting priced out of parking what we found is that generally demand balances itself so that even if there's a block where you have like eight dollars an hour say in the Union Street or Maina District you can go a couple a smart parking project that addresses congestion. It's looking to test technologies that would enable the city to digitally map its curb space so that it can create a reservation-based loading zone. We defined the project requirements using five types of applications. One was an application to support mapping and location type of location of elements. The second was an application to digitize the posted curbside. regulation, because if you ever come to Philadelphia, you'll have to read a dictionary posted on the sidewalk to figure out, can I park here? The third was an application for a user or somebody in the public or fleet to use one of these smart loading zones. So look up where they're allowed to park for a given day or time. And can they use it based on physical and virtual availability. The fourth application we had was something that the city would use from a management perspective, to analyze how the use is happening, what spaces are being allocated to what kind And lastly, we wanted to have an application that would allow enforcement officers to look up a license plate and say, somebody's in the loading zone, smart loading zone. Is that vehicle supposed to be there and for the time and the day they are there? Pilots are an easy way for cities to get... started with smart parking it's interesting to learn about the consideration philadelphia has taken as it moves to deploy its project in march specifically why it's looking for just one vendor cities are generally like at least speaking philadelphia side are bad at integrating products post at like you know procurement because suddenly that escalates cost to a level that we're not prepared to pay and the vendor is not prepared to do the work and you're suddenly in this gray zone where who's going to you know prepare this api that we thought would be part of the product and nobody did and then then you're left with a product two products that don't talk to each other so that's the concern i have in having more than one window how do you measure that you're actually making a change in the city through your project i would direct you to what was just released the curb data specification that was released by Open Mobility Foundation. It has a set of strategies in API form that allows one to think through how to sort of wrap your head around this whole issue of what's the asset, what's the event you want to manage, and what's the metric you want to track. So we looked at metrics coming from... speed circling for people looking for parking was causing significant amount of congestion so we actually had some floating car tests that were done um another major and almost even more important kpi was uh we looked at transit operating speeds and determined whether we saw a positive improvement or decline we actually did see a positive movement in the neighborhoods where we ran the pilot program data is the fuel for every city it's what enables it to operate more efficiently and ultimately address its top challenges. And in San Francisco, the data actually took on a life of its own. The sensor data that we have, robust real-time sensors, stopped working. They promised five years of battery life, we got three and a half. So we came up with a new algorithm that took the expected occupancy results based on sensor data and then... paired that with payment data and allowed us to use payment data as a proxy for real -time occupancy. So we've been using that algorithm where the rates based on payment data predict what occupancy should be on the block. For the last now seven years, we're actually now in the process of updating that algorithm using LPR data for occupancy. So we're hoping to use actually LPR readers now to give us a better snapshot of what's happening. on the street and so what do you do with the sensors uh we've left the sensors in place and so as the paving projects occur the sensors are removed lessons learned that wasn't covered by the vendor um and we done we didn't have that in our grant funding to remove this so what's the acceptable level of accuracy for a smart parking solution to be deemed a success interestingly the two use cases discussed relied on ground sensors to collect the data. This compares to cameras and software solutions used by other cities. We're very concerned because it's a historic area. We have a lot of trees. We have a lot of historic buildings. So for us, the camera technology, it wasn't going to be able to provide us with the level of accuracy that we wanted. A large truck parked in the lot could certainly block a number of spaces. If you're trying to provide like... generalized occupancy, you might not need as high of an accuracy as if you're trying to provide loading zone directions to a vendor that's going to guarantee that at 5 .30 p .m. there will be a yellow zone available on Pine Street for UPS2 lights. I think it really depends also on that use case. So that's a look at the highlights from Current Insights Roundtable on Smart Parking. For more coverage like this, be sure to stick with Current Insights, and if you have a question or comment, on our round table, be sure to hit the button below.