Two airports announced “gateless” parking in 2026 and built substantially different systems. El Paso International moved to a ticketless model in April in which drivers enter and exit without stopping at a booth or kiosk, with license plate recognition and mobile payment carrying the transaction. Raleigh-Durham also markets a stop-free experience — but at RDU the entrance gate opens on a QR code and a plate reader vends the exit gate automatically. The arm is still there. It just no longer waits for a human.

That distinction is the single most useful thing an operations team can carry into a 2026 access-control procurement. “Gateless” has become a marketing word covering two projects with different equipment lists, different failure modes, and very different revenue exposure.

Gate-automated is a control-system upgrade

In the RDU pattern, the physical barrier stays. LPR becomes the credential that triggers the vend instead of a ticket, a validated card, or an attendant. The barrier arm continues to do the job it has always done: it is the enforcement mechanism of last resort. A vehicle that cannot be matched to a paid session does not leave, because a piece of aluminum is in the way.

This is the conservative path, and for high-transient airport lots it is often the right one. The capital work is concentrated in cameras, lane controllers, and the PARCS integration rather than in civil work. The existing gate housings, loops, and arms stay in service. Lane throughput improves because the transaction happens while the vehicle is approaching rather than while it is stopped at the arm.

The trade-off is that everything still depends on the gate operating reliably. Automating the credential does not reduce the maintenance load on the barrier itself — if anything, a lane that now cycles faster cycles more often, and arm, motor, and flange wear track cycle count rather than calendar age.

Truly gateless moves enforcement into the back office

Killeen Regional and Redding took the other path: no arm at all, with occupancy and sessions monitored through fixed and handheld plate reading. Killeen’s premium lot uses fixed LPR cameras plus handheld devices for inventory. Redding’s lot is monitored entirely by plate reading.

Removing the barrier removes the backstop. There is no longer a physical event that prevents an unpaid vehicle from leaving, so collection moves entirely to post-hoc enforcement: matching plates to registrations, issuing notices, and pursuing payment. Walker Consultants published an assessment of this in January 2026 arguing that revenue loss in gateless parking still requires human enforcement — the technology identifies the vehicle, but recovering the money remains a staffed process.

The practical obstacles are not exotic. Out-of-state registrations, leased vehicles, temporary tags, and simple data mismatches all break the chain between a captured plate and a collectable account. Each break is revenue that the system correctly observed and could not recover. In an airport environment, where the share of out-of-state and rental plates is far higher than in a downtown garage, that failure rate is structurally worse than the vendor case studies drawn from municipal surface lots.

Where each model fits

The variable that matters most is not lot size — it is the composition of the traffic and the cost of a failed collection.

Gate-automated suits long-duration lots with high average transaction values, heavy out-of-state and rental mix, and any facility where a single unrecovered session represents multiple days of parking. Economy and long-term airport lots sit squarely here. The arm is cheap insurance against a high-value leak.

Truly gateless suits short-duration, low-ticket, high-turnover environments where the cost of collecting a small unpaid session exceeds its value anyway, and where the throughput gain from eliminating the lane constraint is worth more than the leakage. Cell-phone lots, employee lots with a known plate population, and small regional facilities fit this profile — which is consistent with the airports actually choosing it.

A mixed deployment across a single campus is normal and not a sign of indecision. Different lots have different economics.

What to specify before the demo

Anyone evaluating either model in 2026 should force the following into the specification rather than the sales conversation:

  • Read accuracy stated by plate condition, not in aggregate. A blended accuracy figure is dominated by clean, front-plate, daylight reads. Ask for performance on rear-only states, obscured plates, temporary tags, and night approaches, because those are the reads that generate the exceptions.
  • The exception workflow, costed in staff hours. Every unmatched read becomes work. Whoever owns that work — and what it costs per exception — determines whether the model actually saves money.
  • Registration lookup coverage by state. Back-office enforcement is only as good as the operator’s ability to resolve a plate to a billable party. Coverage gaps are the leak.
  • Behaviour on system failure. In a gate-automated lane, what happens when the reader is down: does the arm fail open, fail closed, or fall back to a ticket? In a gateless lot, there is no fallback at all, which is itself the answer.
  • Cycle-count-based maintenance terms where an arm remains. Faster lanes age the mechanism faster, and a service agreement written against calendar intervals will under-serve it.

The airports moving fastest in 2026 are not the ones that removed the most hardware. They are the ones that were specific about which lot was getting which model, and why.