Editorial Article
The Sensor That Isn't Installed: Where the Money in Parking Sensing Actually Goes
The standard picture of a smart garage is a ceiling of sensors. Every bay knows whether it is empty, a light turns green or red, a sign at the ramp counts down, and the driver finds a space without circling. It is a good picture, and it is a false description of where the money goes. Across the USD 3.15 billion that parking occupancy sensing and space guidance will transact in 2026, the largest and fastest-growing source of new value is not the sensor above the bay. It is the camera at the barrier reading a number plate — a device that touches no parking space at all.
That distinction is the argument of this piece and of the Pheonix Research report behind it. The category is sold as one market and behaves as two. One half instruments the floor: bay sensors, in-ground pucks, overhead cameras, counters at the ramp, and the signage that reports the result. The other half identifies the vehicle: plate recognition at entry and exit, the ticketing and permit layer it feeds, and the software that reconciles a session to a payment. The first half answers where may this vehicle go. The second answers may this vehicle enter, and on what terms. Only one of them is really about parking spaces.
The two halves, in numbers
Pheonix models the market at USD 2,872 million in 2025 — a modelled baseline set at the report’s stated rate rather than an independently observed figure — rising to USD 3,150 million in 2026 and USD 6,020 million by 2033, a 9.69 per cent compound rate across the forecast window. That aggregate conceals the split. Broken into revenue lines, recognition grows at 12.25 per cent against 7.67 per cent for instrumentation, and lifts its share of total value from 32.4 to 38.0 per cent. Bay detection, still the largest single line, moves from USD 1,180 million to USD 1,980 million. Recognition moves from USD 1,020 million to USD 2,290 million and overtakes it. Signage and counters go from USD 395 million to USD 610 million; software from USD 555 million to USD 1,140 million at 10.83 per cent.
The reallocation is the number to hold on to. Over the forecast window the market adds USD 2,870 million. Recognition takes USD 1,270 million of that, bay detection USD 800 million, software USD 585 million and signage the remaining USD 215 million. In a category that describes itself as being about sensing occupancy, 44 per cent of the growth accrues to a technology that never senses occupancy.
Eight units against twelve hundred
The arithmetic behind that is structural, and it is not going to change. A mid-sized urban garage might have 1,200 bays and eight entry and exit lanes. Instrumenting the bays means 1,200 devices, 1,200 power or battery decisions, 1,200 installation hours, and a maintenance obligation that scales with the floor plate. Instrumenting the lanes means eight cameras, eight mounting points and eight network drops. Both tell an operator something useful. Only one of them is bought in single digits.
That ratio also explains margin, which is the part of the story usually left out of vendor decks. A bay sensor is a low-power radio with a magnetometer or an ultrasonic transducer, increasingly sold per unit into a tender and competed on price by manufacturers reading the same reference designs. A plate-recognition installation is a site-specific engineering job: camera geometry, a lighting survey, plate formats across jurisdictions, integration into the operator’s tariff and payment stack, and now often a payment terminal carrying its own certification burden. It is harder to compare on a spreadsheet, and therefore harder to commoditise.
Whoever owns the transaction owns the value
Follow the money one step further and the split becomes a question of who owns the customer. Bay sensing produces a data stream for the operator’s own use. Recognition produces a transaction with a name, a vehicle, an account and a payment instrument attached. The first is a capital expense justified by labour saved and revenue recovered. The second is a revenue system that can be priced per session, attached to a monthly permit, or bundled into a consumer app. Capital markets have already taken a view: Metropolis’s acquisition of SP+ created a combined business valued at roughly USD 1.5 billion, and the price was paid for sites and the transactions running through them, not for a fleet of ceiling sensors.
Concentration follows from that. The top eight vendors in Pheonix’s assessment hold 45 to 50 per cent of market value on under a fifth of the units shipped — the signature of a category where the valuable layer is a small number of integrated installations and the volume layer is a large number of cheap devices. Two competitive sets, then, rather than one. Metropolis and SP+, Verra Mobility and T2 Systems, SKIDATA (Wikidata Q1168035) and Flowbird compete at the barrier; Nedap, Amano (Q11284890) and Park Assist compete in the bay. They meet the same customers weekly and rarely bid for the same purchase order.
The constraint is a job description, not a price
If bay sensing is cheaper per unit and more obviously useful to a driver circling a level, why does it grow more slowly? The most persuasive answer — and the one that keeps surfacing in operator disclosures and industry conference proceedings — is that the barrier to instrumentation is rarely the sensor. It is the absence of a person whose job is to act on the result.
Occupancy data only pays back when somebody uses it: to reprice a permit, reallocate a section, dispatch an enforcement officer to a named bay, release capacity to a pre-book channel, or route a driver at the ramp. In organisations where parking is run as a property overhead — a cost line attached to an asset, staffed by a manager who also runs security — that person does not exist. The sensor is bought anyway, because the development’s design standard specifies it, and it produces a dashboard nobody opens. Recognition survives that organisational gap because it bills somebody. It creates its own revenue line and therefore its own constituency.
That is the finding that travels furthest beyond parking. A technology that generates a transaction is adopted faster than a technology that generates information, at equal technical merit, because adopting the second requires an internal decision about who owns the consequences. Hardware vendors who cannot answer that question will keep losing tenders to software vendors who answer it by sending an invoice.
Where the bay cannot be skipped
None of which makes bay sensing optional everywhere. There are contexts where knowing which space a specific vehicle occupies is the obligation rather than the optimisation: reserved and permit bays, accessible bays under the ADA in the United States and the Equality Act 2010 in Britain, EV-charging bays that have to be blocked against unauthorised dwell, airport pre-book products where the space is sold before it is occupied, and enforcement evidence that must prove a vehicle was in a place at a time.
Airports deserve their own line. They account for 26 per cent of the market’s value on a fraction of its floor area, because a pre-booked airport space is a product with a price on it and an operator who loses money every time it is given away. That is the pattern in miniature: instrumentation is funded where a space is individually sold or individually enforced, and starved where it is merely counted.
Geography follows enforcement authority
Europe holds 38 per cent of the market at USD 1,197 million, North America 27 per cent at USD 851 million, Asia-Pacific 24 per cent at USD 756 million and the fastest-growing of the five, the Middle East and Africa 6 per cent at USD 189 million on the highest value per bay, with Latin America at 5 per cent and USD 157 million.
What separates the European position from the Asian one is not read accuracy or sensor cost. It is whether a municipality or an operator has the legal authority to enforce against a space, and a commercial reason to exercise it. Where the penalty or the fee can be levied, collected and defended on appeal, instrumentation pays for itself against recovered revenue. Where enforcement is nominal, the same hardware has no business case and the market grows only where new capacity is built. The jurisdictions worth watching through 2033 are the ones changing enforcement powers, not the ones changing sensor specifications.
Regulation is doing the consolidating
Compliance is shaping the supply side more than innovation is. The GDPR pushes plate recognition into a specific architecture — a lawful basis, retention limits, and a processing agreement between the operator and every vendor in the chain. Britain’s Product Security and Telecommunications Infrastructure Act and the EU Cyber Resilience Act shift security obligations onto the device maker rather than the site, favouring suppliers who can carry that documentation across a product line. The Energy Performance of Buildings Directive recast and AFIR turn the EV bay from a marketing feature into a regulated asset with its own reporting duty. Each is a compliance exercise, and compliance exercises consolidate: they raise the fixed cost of being a supplier, which suits the eight firms already at scale.
What would reprice this market rather than re-mix it
One development would change the size of the market rather than its composition. Occupancy data is increasingly a by-product of vehicles rather than of buildings: connected cars already report where they parked, and fleet telematics knows precisely when a vehicle arrived and left. If that data becomes accurate, timely and licensable, the bay sensor’s remit narrows to the cases where a vehicle cannot be trusted to report on itself — enforcement, accessible bays, EV blocking, permits. That would not delete instrumentation. It would cap it, and move the value decisively to whoever holds the identity of the vehicle and the right to its data. Any operator signing a seven-year bay-sensing contract should ask whether their capital plan assumes the floor plate is the only place occupancy can be observed.
Until then the market behaves as it has: a growing category in which the most valuable layer is the least discussed. The ceiling of sensors makes the photograph. The camera at the barrier makes the money.
Related research
- Global Parking Occupancy Sensing & Space-Guidance Market Report, Size, Share and Forecast 2026–2033 — the full model, chapter set and regional tables behind this article.
- Global Road Safety & Enforcement Systems Market Report 2026–2033 — the enforcement economics that decide whether a citation, and therefore a camera, pays for itself.
- Global Fuel Card & Fleet Payment Solutions Market 2026–2033 — the fleet-side transaction layer that already knows where a vehicle is.
- Global Automotive Market Report 2026–2033 — the 99.8 million vehicle market that sets the installed base for connected-car data.
Sources and method
Market values, revenue-line splits, regional shares and vendor concentration in this article are drawn from Pheonix Research’s model of the global parking occupancy sensing and space-guidance market (base 2025, forecast to 2033). All figures are Pheonix estimates and are stated in USD.
Regulatory and programme references are to public instruments and official programme documents: the General Data Protection Regulation; the UK Product Security and Telecommunications Infrastructure Act 2022; the EU Cyber Resilience Act; the Energy Performance of Buildings Directive recast; the Alternative Fuels Infrastructure Regulation; the Americans with Disabilities Act; the Equality Act 2010; the USD 5 billion Airport Terminal Program and the USD 5 billion National Electric Vehicle Infrastructure programme. Company identification uses Wikidata entities (SKIDATA Q1168035, Amano Q11284890).
The 30 per cent of this analysis that is not document-level review is Pheonix assessment and is labelled as such: it is an interpretive view drawn from operator disclosures, procurement documents and industry conference proceedings, not a survey panel and not a commissioned interview programme.
