Paying for parking is one of those small transactions that's always been quietly broken.
You land in an unfamiliar city. You download yet another app. Create yet another account. Guess how long you'll be parked.
Then you either overpay for time you don't use — or risk a fine for running over.
Move to the next city, and you start the whole cycle again with a different app. Strip away the convenience marketing, and what's left is a payments-and-trust problem.
A growing mix of players — carmakers, crypto projects, and everything in between — is now racing to fix it.
Key Points
- Most parking is still billed in blocks bought in advance, so drivers systematically overpay for unused time or get fined for overstaying.
- The apps meant to fix this hold sensitive data – ParkMobile's 2021 breach exposed around 21 million users' details, licence plates included – and the market stays split across dozens of operators and apps.
- The genuinely hard parts are physical and institutional: proving a car was really there, and getting thousands of operators and cities onto shared rails. Neither is a cryptography problem.
- Nor is cost a solved problem on either side: card fees are cheap per transaction but add up in volume, while continuous off-chain settlement swaps that for its own costs — opening and closing a channel, plus collateral tied up for the session. The comparison only works if both sides are actually counted.
The Mess Under The Meter
It's easy to underestimate how much this actually costs us.
A widely cited 2017 INRIX study found that U.S. drivers spent around 17 hours a year just hunting for a parking spot. That's roughly $345 per driver, burned on wasted time and fuel. On top of that, drivers overpaid for parking by more than $20 billion a year. In Los Angeles, parking is thought to cover something close to a seventh of all land.
The "smart parking" industry grew up to fix exactly this problem — sensors, cameras, guidance apps. It's now worth around $10 billion, and growing fast.
And yet the actual payment experience, for the person behind the wheel, has barely changed.
Two things keep it broken.
The first is prepayment. You buy a block of time up front. Pad it, and you overpay. Cut it close, and you're gambling on a citation.
The second is fragmentation. Every operator and every city runs its own platform, so drivers end up with a phone full of parking apps. And each one of those apps is a honeypot.
When ParkMobile got breached in 2021, the records of roughly 21 million users — license plates included — ended up for sale on a crime forum. The company later settled for $32.8 million.
Two years later, EasyPark disclosed a breach of its own. EasyPark, for context, had already rolled up ParkMobile, the UK's RingGo, and others into a group spanning some 20 countries.
Consolidation shrank the number of apps. It didn't remove the underlying exposure.
The Fix Everyone Is Already Building
The obvious fix is a single account that works everywhere and bills only for time actually used. Plenty of well-funded players are chasing exactly that — and mostly, there's no blockchain in sight.
Mercedes-Benz now lets drivers start, stop, and pay for parking straight from the dashboard, through its Mercedes pay system. It pulls space data from Parkopedia and settles payment via Mastercard.
Stellantis has gone a step further. Working with tolling firm Verra Mobility (no relation to the carbon registry of the same name), it's building a platform that gives each vehicle a verified digital identity — one that unifies parking, tolls, charging, and fuel under a single system.
Telecom operators and card networks have shown off their own in-car versions too.
The vision itself isn't really in question: one identity, many operators, pay only for what you use.
The open question is who ends up building the plumbing underneath it.
Where A Ledger Actually Helps
This is where the case for on-chain settlement gets specific — and needs to be made carefully.
Billing a driver for the exact minutes they parked is already possible today. Plenty of apps, including Mercedes's dashboard system, simply charge the card once the session ends. So overpayment, on its own, isn't really a blockchain problem.
What card rails can't do economically is settle continuously, in tiny increments, while the session is still running. A fee of around twenty cents plus a percentage per charge would dwarf a one-cent slice of parking time.
But the comparison only holds if the alternative's costs are counted too, and they usually aren't. Opening and closing a channel means two separate on-chain transactions, and on Ethereum's base layer, a simple transfer commonly runs $0.10 to $0.25 each — so open-plus-close alone can land in the same range as the card fee it's meant to beat, before layer-2 discounts are applied. On top of that, collateral has to sit locked for the length of the session, which is itself a cost, even if a small one for something as short as a parking stay. None of this rules out channel-based settlement — it just means the honest pitch is "cheaper once you're on cheap-enough rails and running enough sessions to spread the fixed costs," not "basically free."
That narrower, more defensible gap — between a single charge at the end and a live, co-signed running meter — is what an off-chain channel is meant to fill. But it's worth asking plainly: if end-of-session billing already works, as the Mercedes example shows, what is continuous settlement actually buying you?
The honest answer is narrower than the pitch usually makes it sound. For a driver with a card already in their wallet, in their home market, end-of-session billing is not a hard problem — it's already solved, cheaply, today. The case for something more continuous is real but specific: a driver with no card to hold against, whether unbanked or paying from a crypto-native wallet; a cross-border parking session where a foreign card transaction adds currency conversion and authorization friction on top of the parking fee itself; or a one-off session between a driver and an operator with no prior relationship, where a card hold is the only thing standing in for trust. Outside those cases, for most drivers in most markets, a channel is solving a problem that a card on file already solves.
State channels are designed for exactly this: two parties transacting frequently, off the base chain, with only the opening and closing state written on it. It's a known technique, and a handful of projects are building on it for different markets; Yellow Network is one, proposing to apply it to parking as well.
It's worth being precise about what stage that's at. Applying state channels to parking, as described here, is a conceptual design. not a shipped product or an announced pilot with an operator or city.
When a car parks, the driver's wallet opens a state channel with the operator. The running cost updates through signed messages, minute by minute. When the car leaves, the channel closes and the exact amount settles — with only the final figure written to a public chain.
Because each update is co-signed, the operator can't inflate the duration, and the driver can't deny having parked. A dispute becomes evidence, not argument.
One wallet could hold channels with many operators at once. But whether it actually does depends entirely on getting operators to adopt the same rails.
But a ledger, however well designed, can only make settlement exact and tamper-evident. What it can't do is establish the fact that actually matters: that a specific car occupied a specific bay for a specific length of time. That ground truth comes from a sensor or a number-plate camera, and a reading that's wrong, spoofed, or simply missing is still wrong — even after it's been signed. The integrity problem sits in the hardware, upstream of the channel, and no amount of cryptographic elegance downstream fixes a bad sensor.
Crypto has reached for this before, and mostly stalled at the demo stage. Volkswagen ran a car-wallet pilot with the fee-less ledger IOTA, years ago. Hobbyist proofs-of-concept — a car paying for its own parking — have circulated since.
None of it became real infrastructure. Yellow Network faces the same test. Nothing about this design has proven yet that it clears the bar where earlier attempts stalled.
And the biggest barrier isn't technical at all. It's persuading thousands of operators and city authorities onto shared rails — while carmakers and card networks attack the same problem with account-based systems that ask the driver to install nothing new.
The Version Worth Building
So the credible version of on-chain parking is a modest one.
Not a crypto wallet the average driver ever sees. Just invisible settlement plumbing beneath a familiar card-on-file experience — the thing that finally makes true per-minute billing economic, and turns a duration dispute into a matter of signatures rather than trust.
That's a real and useful role.
But it lives or dies on adoption. And adoption in parking has always been a coordination story: get the sensors accurate, get the operators aligned, get a city to say yes.
Manage that, and the payoff is small but universal — you pay for the minutes you actually parked, in any city, and never think about the app again.
Disclosure:
Yellow Network is a project of Yellow Group, which publishes this site. The piece reflects only the author’s view.