SpotAngels – Using crowds to solve the car parking problem.

SpotAngels is a San-Francisco based start-up founded in 2014 and a graduate of Y-Combinator summer 2014 batch. The company focuses on solving the car parking problem. While it is still at its early stage, SpotAngels is developing an innovative way of leveraging crowds to solve the car parking issues for car owners.

All of us who drive and live in urban metropolitan areas have had a frustrating parking experience, be it spending too much time looking for an available parking spot, or getting our car towed after overlooking the parking restriction. SpotAngels is a company aiming to solve those two parking problems, for current drivers but also ultimately in the self-driving world.

SpotAngels is a San-Francisco based start-up founded in 2014 and a graduate of Y-Combinator summer 2014 batch. The company focuses on solving the car parking problem. While it is still at its early stage, SpotAngels is developing an innovative way of leveraging crowds to solve the car parking issues for car owners.

Currently the company has an iPhone consumer App that is made available to users for free. The App automatically detects when a car is parked, and sends notifications to the users regarding parking restrictions (e.g. “Street cleaning in 20 minutes”). In the near future, the App will also allow users to see a real-time map of parking availability.

Crowdsourcing Parking Restrictions  

Parking restrictions are complicated, not uniform across cities, and there is no public central database making them available – not even cities can provide a database of their own parking regulations. In order to build a database of parking restrictions in a short period of time, SpotAngels have used a combination of proprietary computer vision technology to read parking regulation signs, which provided a majority of the data, combined with a crowd-sourcing model where users can take a picture of the parking sign and share it with the company. The snapshots below show how the user can send a picture on the App. Despite the recent re-launch of the App, the team has been receiving a large number of data from its users – which have made processing the data in a timely manner challenging.

Next Step – Crowdsourcing Parking Availability

Parking restrictions are just one part of the problem. The other key problem is the ability to identify available slots for parking in real-time. Multiple innovative crowd-sourcing approaches are currently being explored by the team.

One approach would be to collect data in real-time from the App users when they park their car, and use predictive models to predict how likely is that street / area to be busy at that time. The challenge with this approach resides in the fact of reaching a critical size in order to be able to make useful predictions.

Another way to crowdsource parking availability would be to partner with a pool of car owners who ideally spend a lot of time driving around (e.g. taxis, ride-sharing companies) and ask them to scan the streets in real time and collect data when they see a free parking spot. The challenge with this approach would be the need to develop a new ‘scanning’ technology as well as the ability to reward the pool of scanning cars – which would typically involve paying a small fee to them.

Business Model and Value Capture

As mentioned above, the B2C App is made available to users for free. The company has basically two options in terms of value capture:

The first option is to continue focusing on the B2C space and develop new innovative ways of monetizing the app usage (Just like Waze did through advertising or may be up-selling new features).

The second option would be to keep the B2C App free of charge without advertising, and selling the parking availability and restrictions data to businesses who want to integrate it in their own software (e.g. just like Google Maps is selling the mapping service to apps like Uber).



[Option 2] Glassdoor: Rant about your boss AND help the community

Student comments on SpotAngels – Using crowds to solve the car parking problem.

  1. I wonder if SpotAngels will really be able to transition to crowdsourcing parking availability. If its user base is broad enough, parking availability might turn into an efficient market – as soon as I see the free parking space on my app, someone else will already be there?

    1. Thanks Meili for your comment. The fact that the market becomes more efficient doesn’t mean that the service becomes less relevant – in fact, it’s the opposite. In your scenario, the ‘someone else’ can only already be there if they have themselves used the app.

  2. Totally agree with Meili’s point about parking availability. In high-traffic urban areas at peak times, cars often stall in traffic to wait for a car to leave a parking spot, so there is very little time when the space is not occupied. I don’t think the current availability feature is very useful unless you’re dealing with a parking lot that has multiple open spaces and wouldn’t fill up immediately. The data on parking restrictions does seem useful but strikes me more as a feature of another platform (e.g., Waze) vs. a standalone application.

    1. Great point Lacy. I agree – the fundamental assumption behind the parking availability service is that parking spots are not 100% utilized.

  3. Great post – I hope SpotAngels is able to help me find a parking spot regularly one day!

    About the parking restrictions component, I’m curious about two things: (1) How is the company thinking about accuracy? For instance, parking restrictions change from time to time. Is there a way for the team to use data from public records, noting when parking rules are changing? Also, for NYC at least, some times parking rules are in effect and other days they aren’t (e.g. certain rules are apply if it’s a certain holiday, if there’s a parade, if there’s an anticipated weather event, etc.). How is the company thinking about incorporating these nuances? (2) How is the company thinking about liability? I’m sure (unfortunately) there’s someone who will want to sue the company if his car gets towed because he followed inaccurate guidance on the app…

    1. Thanks Lauren for your comment.

      (1) I think the fact the data on restrictions is always changing is an opportunity for the company: If the data was static then the company would create less and less value over time. Unfortunately cities are not very good at updating and keeping track of new rules – so the solution will probably be either crowdsourcing when somebody finds an obsolete rule, or some primary data collection for example through image recognition (just like Google StreetView is regularly updated). With regards to your second point, the company now captures all sort of rules including complicated ones that only apply on specific days / hours.

      (2) Haha good point. I don’t think this is something that they have thought about yet!.

  4. Interesting — I also agree with the above comments about parking availability. Additionally, in a world of autonomous vehicles, won’t each vehicle know where the next parking spot they should go to if they can map where all the cars are in the system and optimize for the best outcome? I wonder whether this company will become successful by tapping into a current annoyance, and how long they will last before another platform incorporates some of their features or their value prop proves to be not as attractive.

    1. Thanks for your comment Ophelia – great point. Two answers: First, even though the first autonomous cars will arrive soon, it will take a lot of time (20 years?) to replace the entire car fleet and have a 100% driverless car fleet. Second, autonomous cars will still need to know if they are allowed to park in a given spot.

  5. I think this is extremely interesting, as i absolutely hate driving around the block looking for parking spaces. I wonder if there was any way they could work with the city to either “connect” the parking meters already installed or work to deploy an “Internet of Things” style solution to try an overcome the issue of having enough human sources to be accurate.

    1. Thanks for your comment. Great idea – if the park meters are smart, they could know when a spot is busy and when it is not. Ultimately, the vision for SpotAngels is to become a digital park-meter that you have in your phone, potentially including services like payment.

  6. Great post, thank you. I like Andrew idea’s above about potentially connecting with parking meters/physical sensors to know about the availability of spaces/rules rather than relying on human sources. When I think about experiences that I’ve had looking for and then finally parking in urban areas, I usually get out of my car, pay for parking, and then start heading to my destination ASAP. Maybe it’s just my personality, but it seems like it would be a hard sell to get people to take pictures/report data after finally finding a parking space.

  7. Great post – thank you. Parking tickets are the worst, so glad someone is trying to help prevent them! I wonder if they can team up with the city government facilitate payment (both of parking and tickets) via their app to increase adoption. That way, people can download the app to quickly pay for parking instead of having to walk to a meter, and SpotAngels can collect the data at the same time.

    1. Thanks for your comment. Payment is indeed a great opportunity that the team has indeed thought of for their long-term vision. The way we pay for parking today is inefficient and obsolete. This however looks like a winner takes / monopoly opportunity and SpotAngels would need to prove itself before being able to get traction from cities.

Leave a comment