Buying the wrong biometric terminal is an expensive mistake. A face reader in direct UAE sunlight fails through the summer. A fingerprint reader on a construction workforce reads maybe two thirds of the workforce reliably. A low-throughput terminal at a 200-person gate causes a ten-minute queue every morning.

These are not exotic failures. They are the four ways most bad specifications happen. Here are the four questions that avoid them.

Question 1: How many people, how fast?

The single biggest specification decision. A terminal that clears 200 people through one door in five minutes costs about three times a terminal that handles 30 people in the same window. Paying for that speed on a 30-person office is wasted budget that would buy a second door.

Rough guide:

  • Under 30 people arriving together: Any modern terminal is fine. Choose on hygiene, cost, credential.
  • 30–100 people: Mid-range terminal, SpeedFace-V5L or FaceStation F2 comfortably. Standard fingerprint terminals will queue at the door but manage.
  • 100–200 people: ProFace X or a second mid-range terminal. Fingerprint queues become measurable.
  • Over 200 people: ProFace X, FaceStation F2 or multiple terminals. Fingerprint alone is not the answer.

The trap: assuming your total headcount is the number. What matters is how many people arrive within about ten minutes at the busiest door. A 500-person factory with three staggered shifts is a different problem to a 200-person warehouse with one 06:30 start.

Question 2: Indoors or exposed?

A terminal on an indoor reception wall is a different specification to one on an outdoor site gate. Confuse the two and the outdoor unit fails within a season.

UAE-specific factors:

  • Temperature. 45–48°C in July and August, before you add solar loading. An indoor terminal specced for 45°C will run above rating on any west-facing outdoor mount.
  • Dust. Site environments, near-industrial locations and highway-side installations kill non-sealed units through dust ingress at the sensor.
  • Humidity. Coastal locations (Dubai Marina, Corniche) run 70–90% humidity in summer. Condensation at night is real.
  • Wash-down. Compound gates and warehouse entrances get rinsed with pressure washers. IP65 or higher survives that; IP54 does not.

For anything outdoors, the honest specification is an IP65-rated unit. Suprema BioEntry W3 for fingerprint/card outdoors; ZKTeco's outdoor face and card readers for other cases. We install neither indoors, the outdoor units cost more than needed for a covered reception.

Question 3: What are your workers' hands like?

The question buyers most often skip, and the one that causes the most avoidable failures.

Fingerprint readers work well on hands that:

  • Have not been worn smooth by manual labour.
  • Are not wet, oily or gloved.
  • Are not frequently damaged (workshop, kitchen, factory floor).

Fingerprint readers fail on hands that:

  • Have been doing concrete, welding, warehousing or kitchen work for years.
  • Are constantly gloved (food handling, medical, cold storage).
  • Are wet or oily at the time of the punch.

For workforces in those failure categories, the honest answer is:

  • Face recognition if it can be positioned away from direct sunlight and the enrolment area is well lit.
  • Palm recognition (ZKTeco ProFace X) if the environment defeats face too, heavy dust, poor lighting, or where a face terminal cannot be mounted well.

A common trap: enrolling everyone on fingerprint at go-live, then discovering three months in that a fifth of the workforce cannot be read reliably. If more than about 20% of the workforce have hands like the failure list above, do not attempt fingerprint as the primary method.

Question 4: Does the machine also need to open the door?

Attendance-only is cheaper. Attendance + access control on the same unit costs more but replaces a separate door reader.

The trap is buying attendance now and deciding to add door control later. That means re-cabling the door (adding a Wiegand or OSDP line, plus a strike or maglock), which costs almost as much as the second unit would have.

Decide at the survey:

  • Attendance only: Standard biometric terminal, no relay wiring.
  • Attendance + access on this door: Terminal with door relay, cabled to the strike or maglock at installation, request-to-exit and door contact wired if you need audit-quality access records.
  • Attendance here, access control on other doors: Attendance terminal at the entrance, plus separate access controllers on the sensitive doors (server room, executive floor). One central software.

If the door is a compliance door, server room, cash office, executive floor, do not skip the request-to-exit and door contact wiring. Without them, an audit sees door opens without knowing whether someone walked in or out.

Putting it together

The four questions in order:

  1. Throughput: How many, how fast, at the busiest door?
  2. Environment: Indoor, outdoor, or between the two?
  3. Hands: Clean office workforce or worn manual workforce?
  4. Access control: Attendance only, or does this door also need locking?

Answer these four and the specification writes itself. Most bad terminal choices trace back to one of them being skipped or answered wrong.

Related reading

Or skip the analysis and let us do the specification: free UAE-wide site survey, fixed written quote, honest recommendation.