Westpark combines what locals see with data we can observe — weather, time, events, school schedules — to produce a simple prediction for every downtown lot. Here's exactly how.
Prediction tiers
Open Westpark on any phone — no download, no account required to check conditions. You'll see a table of all 20 downtown retail lots, color-coded by predicted occupancy: Easy, Moderate, Busy, Very Busy, or Slammed.
If you want to report what you see at a lot, tap the lot name and select Open, Filling, or Full. That's it — two taps. If it's your first report, we'll ask for a first name and email so we can track your contributions and award points. No password, no download, no app store.
Your report is immediately factored into the prediction for that lot, weighted alongside reports from other locals and the signals the model tracks on its own.
Every prediction combines live community reports with five observable signals:
Produces a lot prediction: Easy / Moderate / Busy / Very Busy / Slammed
A Friday evening in July with a Levitt concert looks very different from a Tuesday morning in January. The model accounts for all of it.
Predictions show as Easy, Moderate, Busy, Very Busy, or Slammed — color-coded, at a glance. We use tiers instead of percentages on purpose. A percentage implies precision we don't have. A color tells you what you need to know: head there first, or try somewhere else.
Westpark launched with a hand-tuned prediction model — rules we wrote based on what we know about Westport's parking patterns. This gives us a reasonable baseline from day one, but it's not perfect.
Accuracy improves as local reports accumulate. We estimate we need roughly 2,000–3,000 reports to build a statistically reliable baseline across all 20 lots and time windows. In the early months, treat predictions as directionally useful — a good first guess, not a guarantee.
As the community grows, the model transitions from hand-tuned rules to machine learning trained on real local data. Patterns will emerge that no human would have thought to program — like which lots fill fastest after a rainstorm, or how a school play-off game ripples through downtown traffic at 4pm. The more you report, the smarter it gets for everyone.
Every report earns points. Points unlock status in the Westpark community — not titles, not badges, but actual standing among the locals who know downtown best.
Status tiers: Neighbor → Regular → Local → Insider → Legend. Points thresholds: 0, 50, 150, 350, and 700. Your tier is visible on your profile and on the community leaderboard.
For now, status is bragging rights. We have ideas for what it could mean down the road — but we'd rather earn the right to make promises than make them before we have.
Not all reports are weighted equally. New users' reports carry 0.5x weight while the system learns them. After 5 reports, that rises to 1.0x. Reporters whose observations consistently match consensus earn a 1.5x credibility multiplier.
This protects against noise and rewards accuracy — the people who report most reliably have the most influence on predictions.
No. Westpark is an independent, community-built tool. We're not affiliated with the town, the DPIC, or any official body. We use publicly available information (the Westport community calendar, public lot data) and reports from residents.
No sensors, no cameras, no hardware of any kind. Westpark is entirely software — a prediction model trained on human reports, weather data, and calendar events.
In our model, directional accuracy (is this lot easy or busy?) should reach 85–92% as the community grows. We're honest that we can't predict a spontaneous accident that blocks a lot entrance or a pop-up event that isn't on the calendar. The goal is to help you make a better first decision — not to guarantee a spot.
Yes, always. Westpark is free for residents to use and will stay that way.
No account needed to check parking conditions. Westpark is a web app — there's nothing to download from the App Store or Google Play. If you want to report a lot condition and earn points, we'll ask for your first name and email. We send you a magic link — no password ever.
All 20 public downtown retail lots: Parker Harding, Baldwin, Sigrid Schultz, Taylor, Imperial Avenue, Upper Library, Jesup North, Jesup South, Town Hall, Senior Center, Police Lot, Gorham Island, Bay Street, Myrtle Avenue, Avery Place, Elm Street, Jesup Road, Church Lane, Main St North, Main St South. Train station and beach parking are not yet covered — we plan to add those in 2026.
Not yet. We started with downtown retail lots because that's where the daily frustration is most acute and most solvable with community data. We plan to add Saugatuck Station and Compo Beach seasonal coverage in 2026. We'll share updates as we make progress.
We collect your first name and email to identify your reports. We never sell your data. Full details in our Privacy Policy.
Use our contact page. We read everything.