Parking Signs are Confusing - AI Can Fix That.
The City of Los Angeles is notorious for having convoluted and confusing parking signs, with locals and visitors alike lamenting unforeseen parking tickets. By utilizing artificial intelligence, What the Park? tells the user if they can or cannot park at a given spot, accompanied by a visualization.
Defining Problems
People want more intuitive parking signs, but custom signs for each street is a lot of work.
In 2015, the City of Los Angeles briefly introduced parking signs that were designed by Nikki Sylianteng, who came up with visual representations to address the confusion caused by convoluted parking restrictions.
However, due to budget and logistical issues, the initial 100 signs would be the only signs to be made and put up on streets.
Despite the failed adoption of the redesigned parking signs, many agree that they were more intuitive. I gathered online opinion on what Angelinos thought about Sylianteng's signs:
User Insight #1: Standardized Layout of Information
“One of the benefits of the new sign is the standardized layout of information. If you're trying to park on a Wednesday afternoon in an area where there are a half-dozen parking times and a sign for each, you have to read them all to figure out which applies to you that day. With the new sign you know exactly what part of the sign to look at.”
User Insight #2: Easy to Understand
“I for one think these are MUCH easier to read and understand. Having seen them in person I can see them much easier from my vehicle.”
User Insight #3: Still Too Complicated
“Can we have signs in plain English?!? Or some form of app that you can enter the code and it dumbs it down for you.”
Negative Feedback
“I don’t find it to be any better than what we have today”
“These are a no from me. Maybe because I grew up here, but I’ve never struggled to read any parking sign”
It's worthwhile noting that some people were against the new signs. However, a majority expressed positive opinions, denoting that there is indeed a problem to be solved.
With this information, we can define the problem:
Drivers prefer more intuitive parking signs, however, the city remains limited by budget and logistics.
Exploring Solutions
Leveraging AI = No Physical Limitations
Using Artificial Intelligence to scan parking signs and generating visual diagrams
Having an app generate Sylianteng's visualizations lessens the burden on the city by not requiring custom signs, while offering drivers the convenience of having more intuitive ways of understanding parking restrictions.
Opt-in Locating Saving for faster load times
Storing location data along with each scan will provide more pre-processed data for the AI to work off of. By pulling up past scans from similar locations, the AI can recognize patterns more easily and won't have to start computing from scratch each time.
Competitive Analysis
Upon some research, there were a few programs that offered similar services:
Parky.AI
Parky.AI is a mobile app where the user takes a picture of the parking sign, then the app tells the user whether or not they can park there.
Pros
Straightforward user flow
Easy to understand results
Cons
Home screen with button to open camera is unnecessary (the user will only use the app to take a picture)
Perhaps too little information in results
5 free scans, subscription required after
Parking Sign Interpreter GPT
Powered by ChatGPT-4o, this web based chatbot analyzes the parking sign if the user provides a picture and time.
Pros
Detailed analysis
Free
Cons
Browser based
Too much manual input
Potentially too much information
User-Centered Design
Uncertain waits are longer than known, finite waits.
Parky.AI would sometimes show the current applicable parking restriction along with the time remaining. However, other times it would simply say: "No parking restrictions apply at the current date and time." What does that mean? When can I park until? What if I'm parking overnight?
In the results page, showing the results along with a visualization provides the user with a more comprehensive understanding, eliminating any lingering doubts. There is also the option to remind the user to move their car, with a timer that can be customized in the settings.
User Feedback: "It'd be nice if I could tell where I am"
In addition to showing the visualized parking signs, one user commented that it would be nice to see where on the calendar they are.
I experimented with adding an indicator that is commonly used in calendar apps, however, that would interfere with the iconography used in the diagrams. I decided to make the color brighter and add a hover animation instead for an unobtrusive indication of which slot the user's currently in.
Accessibility: Color Blindedness
When designing, Nikki Sylianteng had color blinded drivers in mind as well; her iterations can be seen below:
In addition to the iconography, she added patterns to distinguish the green and red areas. Similarly, I also added icons and animations as indicators other than color. However, some users still expressed difficulty in understanding.
User Insight: "As a colorblind person, … it's just one big blob."
In response to Nikki Sylianteng's original design, someone commented:
"As a colorblind person, I can say that sign would be impossible to decipher from behind the windshield of a moving car. At that distance, it's just one big blob."
When looking through the eyes of someone with protanopia, even the patterns become hard to discern.
To make sure all users can easily understand the visuals, I added a toggle for a color blind mode.
Nikki Sylianteng's designs had a physical constraint where the same sign would have to be used by non-color blind and color blind users alike. However, that limitation disappears on an app.
Revisiting Past Scans
By keeping a history of past scans, users can go back and look at the parking rules for locations they may frequent, and add to a favorites list.
By keeping a history of past scans, users can go back and look at the parking rules for locations they may frequent, and add to a favorites list.
Tracking Impact
Since this is a case study, there aren't any tangible metrics to work off of. However, here are KPIs that would indicate the app's success in helping people.
Primary KPIs
User Engagement: Daily/Monthly Active Users, Retention Rate
User Interaction: Scans per user, Reminder set rate, Feature Usage.
Supporting KPIs
User Feedback: Surveys to understand if users felt helped
Impact on Parking Violations: Track revenue from parking tickets alongside app usage metrics (many external factors to consider, however, would still be an interesting statistic to keep an eye on)
Takeaways
Overcoming Fiscal + Logistical Obstacles
Though the redesigned parking signs initially received a lot of praise, there was a limit to how widely it could be implemented throughout Los Angeles. Each street would require a customized sign with its own set of rules, and it would have to be compliant with and enforceable by Los Angeles Municipal Code.
However, all of those problems could be overcome by placing the burden entirely on a mobile app. With the advent of artificial intelligence, it's exciting to think of other areas of life that could potentially be improved.