Terminal Count per Cabinet

Editorial Team September 15, 2026

Terminal count per cabinet is normally one payment terminal per kiosk cabinet, and a second unit is only justified when site inputs force concurrent payments. The three inputs that matter are simultaneous users at peak, transaction duration, and the length of the unattended operating window. Size those first, then specify the cabinet.

What Is a Self-Service Payment Kiosk (and Why Terminal Count Is a Separate Question)

Self-service payment kiosk: an unattended terminal that lets a customer order, check in, register or check out and complete payment without staff assistance. It combines a touchscreen, a computing module, and one or more payment, printing or identification peripherals inside a single enclosure.

A self-service payment kiosk is a complete unit, not a component. Buyers who confuse the two end up specifying a whole fleet when the real constraint sat elsewhere in the assembly. Fleet size answers how many cabinets go on the floor. Terminal count per cabinet answers how many card readers, printers and identity readers each cabinet carries. Those are different decisions with different inputs, and the second one is usually resolved by the time you issue an RFQ for unattended kiosk payment hardware.

Start With the Baseline: One Payment Terminal Per Cabinet

The industry baseline is a single payment module per unattended cabinet, because that is how POS hardware is conventionally built. A POS terminal is generally a single employee- or customer-facing unit combining a card reader, a receipt printer and cash-register components, according to a 2026 Kiosk Industry overview of POS systems and components (^[2]). Kiosk payment terminal integration inherits that pattern, and kiosk enclosure design follows it.

Cabinet role	Typical terminal count	What forces a second unit
Restaurant self-ordering	1	Two+ simultaneous payers during a lunch peak
Retail self-checkout	1	Mixed cash and card demand at the same bay
Hotel self-check-in	1	Group arrivals and long identity-verification steps
Ticketing / queue management	1	Departure-time spikes with hard deadlines

Run the baseline until evidence says otherwise. A second terminal adds cost, PCI DSS scope and cabling to every cabinet you buy.

Four Site Inputs That Force a Second Terminal When You Add Languages, Queues or Unattended Hours

Signal you need it: two or more customers reach the payment step within the same minute.

1. Simultaneous users at peak, not daily footfall

Daily footfall tells you nothing about concurrency. Ten thousand people spread over twelve hours can run through one terminal comfortably; four hundred arriving in a fifteen-minute window cannot. Work from the busiest fifteen minutes of the week, not the monthly average. If your peak arrival rate exceeds what one terminal can clear in that window, you have a concurrency problem, not a capacity problem.

Signal you need it: average session time exceeds the interval between arrivals.

2. Transaction duration, split into session time and terminal handshake time

Split duration into two parts. Session time is how long the customer spends browsing and confirming. Terminal handshake time is everything after that: card read, authorisation, receipt print. Handshake time is the part a second terminal actually fixes, because it is the step that serialises payers. A second terminal cannot shorten menu browsing.

Signal you need it: the site is expected to run outside staffed hours.

3. Unattended operating window

An unattended cabinet has no fallback when a terminal faults. During staffed hours, a cashier can take payment on a register. At 2 a.m., the cabinet is the only path. A longer unattended window raises the cost of a single-terminal outage, which pushes some operators toward redundancy rather than throughput. Multi-language kiosk hardware requirements often appear in this window too, because unattended sites serve more language groups.

Signal you need it: every session adds language or accessibility steps.

4. Language and accessibility steps that add time to every session

Language selection, larger touch targets, audio guidance and height or reach accommodation all add seconds per session. Those seconds compound against the peak window. You will then need a higher terminal count per cabinet to hold the same throughput.

The Concurrency Model: Four Steps to a Terminal Count

Throughput and concurrency are not the same measurement. Four steps convert measured demand into a terminal count, and the result is a planning estimate only.

terminals required = ceiling( peak simultaneous users x payment-step share / 1 )

Step 1. Convert peak arrivals to simultaneous users. Take arrivals in the peak 15 minutes, divide by the number of 15-minute blocks in that period, then convert arrival rate to a concurrent count using average session length. Placeholder: 60 arrivals per peak hour, 4-minute sessions, 4 concurrent users.

Step 2. Split average dwell time. Separate session time from terminal handshake time. Placeholder: 3 minutes of session time plus 45 seconds of handshake time.

Step 3. Apply the payment-step share. Only some sessions reach the payment terminal; the rest are browsing, cancelled, or routed to a counter. Placeholder: 70% of sessions pay at the cabinet.

Step 4. Subtract re-routable steps. Language selection, loyalty lookup and identity verification can sometimes move off the payment terminal to a separate reader or an app. Whatever you remove lowers the terminal count per cabinet.

Worked example: 4 concurrent users x 0.70 payment-step share = 2.8, so ceiling gives 3 payment steps needing simultaneous service. At 45 seconds of handshake time each, that is roughly 135 seconds of serialised occupancy per cycle. If sessions overlap at that rate, one terminal is the bottleneck.

Same site	One terminal	Two terminals
Payment steps per peak cycle	3 serialised	3 across two lanes
Handshake occupancy	~135 s on one reader	~90 s worst lane
Queue at the payment step	forms	absorbed
Cabinet bays required	1	1 (second terminal) or 2 (added bay)

Right-sizing a kiosk fleet stops at cabinet count. This worksheet goes one layer down the assembly.

One Terminal With Re-Routed Flow, or Two Terminals and an Added Cabinet Bay?

Both options solve concurrency. They differ in what they change and who they suit.

Option	What changes in the cabinet	What changes on site	Best fit
One terminal, re-routed flow	Peripheral layout and cable routing only; payment module unchanged	Staffed fallback, app-based payment, or a second lane nearby	Sites with staff nearby and moderate peaks
Two terminals, one cabinet	kiosk enclosure design payment module layout splits into two bays, with separate printers and cabling	Higher unit cost and wider PCI DSS scope per cabinet	Peaks where no staff fallback exists
Two cabinets, one terminal each	No change per unit	Floor space, power and network drops double	High overall volume, not short bursts

Decision rule: if the peak is short and staff can absorb overflow, re-route the flow and keep one terminal; if the peak is short and nobody can absorb it, add the second terminal. Most self-checkout kiosk hardware requirements published by suppliers assume one reader per unit, so screen this section against your self-service kiosk configuration guide before approving a two-terminal cabinet.

How Terminal Count Ties Back to Compute and Peripherals

More terminals, printers and identity readers on one cabinet means more simultaneous I/O and more local processing. A cabinet running two payment modules, a receipt printer and an ID scanner is handling overlapping input streams that a single reader never generates. That pushes up CPU headroom, peripheral controller count and power budget, and shifts the on-device versus cloud split for anything latency-sensitive.

More payment modules means more concurrent I/O and authorisation traffic to manage locally.
Kiosk peripheral integration across two readers usually needs separate controller paths, not a shared bus.
Edge computing headroom should be sized after terminal count, not before.

Sizing that compute is a separate exercise. See right-sizing edge AI NPU compute and the companion notes on NPU sizing when the edge AI workload is uneven rather than repeating the budget here. Single-unit pilots such as edge AI tablet procurement and AI digital human display compute are the right places to validate these numbers.

Where the Demand Is Coming From: Evidence Versus Inference

Keep the two separated, because only one of them changes your terminal count.

Cited evidence (dated)

Self-service kiosks and self-checkout systems dominate the unattended terminals market, attributed to high adoption across retail and hospitality segments, per Strategic Market Research’s market report (^[1]). Self-service kiosks represent 15–20% of US POS hardware value, with growth driven by replacement and upgrade cycles alongside labour-cost pressure, according to a QTenboard payment kiosk guide published in 2026 (^[5]). Supplier guidance states that kiosk hardware should provide flexible installation space, mounting and cable routing for different payment terminals (^[6]). Older general kiosk surveys, like Frank Mayer’s 2022 overview of the seven common kiosk types, are historical context rather than current conditions (^[3]).

Our sizing inference

Two of those inputs translate directly into concurrency pressure: adoption growth in self-checkout and self-ordering means more bays competing for the same peak window, and labour-cost pressure pushes hours toward unattended operation, where a single terminal has no fallback. Flexible mounting and cable routing does not raise concurrency; it only makes a two-terminal cabinet easier to build. The realistic terminal count per cabinet is therefore driven by measured arrival patterns at your site, not by market growth figures.

Frequently Asked Questions

What hardware does a restaurant self-ordering kiosk need?

A restaurant self-ordering kiosk needs a touchscreen, a computing module, one payment terminal, a receipt or order-number printer, and an enclosure with mounting and cable routing that accepts different payment terminals. Card, contactless and NFC options are commonly offered as standard configurations (^[4]). Add a barcode scanner or cash module only if your order flow requires it.

How many kiosks does my restaurant need, and where should they be placed?

Size kiosk count from peak arrival rate and average session time, not daily covers, then place units where the queue already forms: lobby entry, waiting area, and near pickup. Building kiosks for high-traffic environments means planning for overflow at the busiest moment, not the average day (^[7]).

Can one cabinet run two payment terminals?

Yes. A single cabinet can carry two payment modules if the enclosure allocates separate bays, separate printers, independent cable routing and enough power and controller capacity. The constraint is usually mechanical and compliance-related rather than electronic, because adding a second reader widens PCI DSS scope and needs confirmation from your payment processor.

How do languages and unattended hours change the count?

Language selection, audio guidance and accessibility accommodation add seconds to every session, and unattended hours remove the staffed fallback that would otherwise absorb overflow. Both raise concurrency pressure at the payment step. Treat the result as a planning estimate and replace every assumption with your own measured peak-hour data.

Common Mistakes to Avoid

Sizing from daily footfall instead of the peak fifteen minutes.
Adding a second terminal to fix slow menu browsing, which the reader cannot solve.
Ignoring the terminal handshake step, which is the only part a second reader actually speeds up.
Forgetting that unattended hours raise the cost of a single-terminal outage, not just the queue.
Ordering a two-terminal cabinet before confirming PCI DSS scope with the payment processor, and before checking accessibility requirements for the payment interface against ADA or local equivalents. These are constraint checks, not legal advice; confirm with the processor and a qualified adviser.

Verification and Next Steps

Before approving a cabinet design, confirm four measured inputs: peak arrivals per fifteen minutes, average session time, average handshake time, and the share of sessions that reach the payment step. If they are not measured yet, instrument the current counter flow for two weeks. Then run the four steps, compare the one-terminal and two-terminal rows, and hand the resulting numbers to your OEM or ODM partner. For OEM ODM self-service kiosk projects, specifying terminal count per cabinet early avoids redesigning the enclosure after the compute module, printers and cash handling and cashless payment kiosks modules are already placed.

Content reviewed: 2026-09-15.

Evidence confidence

Confidence: Medium. This rating reflects cross-checking 7 sources across 6 independent domains. It measures evidence coverage, not certainty; verify safety-critical work against manufacturer instructions and local requirements.

References

APA 7th edition

↑Strategicmarketresearch. (n.d.). Unattended Terminals Market 2026: Expert-Crafted Insights You. Retrieved September 15, 2026, from https://www.strategicmarketresearch.com/market-report/unattended-terminals-market.
↑Kiosk Industry. (2026). POS Systems, Terminal and Components 2025. https://kioskindustry.org/pos-terminal-kiosk/.
↑Frank Mayer. (2022). 7 Common Types of Self-Service Kiosks. https://www.frankmayer.com/blog/7-common-types-of-self-service-kiosks/.
↑Self-Ordering Kiosks - Diebold Nixdorf. (n.d.). DN Series® EASY MAX. Retrieved September 15, 2026, from https://www.dieboldnixdorf.com/en-us/retail/portfolio/systems/self-service/self-ordering/.
↑Qtenboard. (n.d.). What Is a Payment Kiosk? Definition, Business Models & Application. Retrieved September 15, 2026, from https://www.qtenboard.com/kiosk-guide-653.html.
↑Westarpos. (2026). What Hardware Does a Restaurant Self-Ordering Kiosk Need? A. https://westarpos.com/2026/07/22/what-hardware-does-a-restaurant-self-ordering-kiosk-need-a-complete-buyers-guide/.
↑Frankmayer. (n.d.). Building Kiosks for High-Traffic Environments. Retrieved September 15, 2026, from https://www.frankmayer.com/blog/building-kiosks-for-high-traffic-environments.