Biometric Access Control Installation in Houston, TX 77027

Which openings justify a finger

Biometrics are not a building-wide product. In a leased suite they belong on a small number of openings where you need to know which individual went through, and badges alone belong everywhere else.

The thinking behind the table is straightforward: a badge proves someone holds a credential, a fingerprint proves who is holding it. You pay for that distinction only where credential sharing would actually cost you something.

Building rules, the work letter and the engineer

The low-rise and mid-rise stock across this ZIP is largely multi-tenant, professionally managed, and governed by a work letter that spells out what a tenant contractor may and may not do. Skipping that step is the fastest way to have a job stopped on day two.

• Approvals before tools. Most buildings want a scope narrative, a drawing marked with each affected opening, and a certificate of insurance naming the owner and manager as additional insured before anyone badges in.
• Core drilling and penetrations. Slab cores, header penetrations and anything through a demising wall are almost always landlord-approved and after-hours work, and fire-stopping of every rated penetration is inspected.
• Plenum space. Most suspended ceilings here are return-air plenums, which means plenum-rated cable and no exceptions.
• Freight and loading. Elevator reservations, dock windows and quiet hours for clinical suites all shape the schedule more than the wiring does.
• Base-building systems stay base-building. The lobby card system, the elevator and the fire alarm panel are the owner’s. We integrate at the interface the building allows and leave their equipment to their contractors.
• Make-good at lease end. Worth deciding on day one what comes out and what stays, because it changes whether hardware is surface-mounted or cut into a frame.

Wiegand is finished; specify OSDP

For decades readers talked to controllers over Wiegand: one-way, unencrypted, unsupervised, and practically limited to a few hundred feet. Anyone who can reach the wiring behind a reader can capture credentials off it and replay them. On a badge-only door that is bad. On a door protected by biometrics it defeats the entire point of installing them.

The replacement is OSDP, running over an RS-485 pair. It is two-way, so the controller knows in real time whether a reader has stopped answering or been tampered with. It supports an encrypted secure channel, so what crosses the wire is not replayable. It reaches thousands of feet rather than hundreds, and multiple readers share one run instead of each needing its own home run.

Practical points we hold ourselves to on a fit-out here:

• RS-485 is a daisy chain, not a star. Stub a reader off the middle of a run and you get intermittent faults nobody can trace later.
• Terminate the ends of the bus, and keep the shield grounded at one end only.
• Turn the secure channel on at commissioning. A reader left in its default install mode is an OSDP reader with Wiegand-grade security.
• If a door controller is powered over the network, the power budget has to include lock current, not just the controller. A magnetic lock drawing meaningfully more than a strike is exactly where that gets missed.

One-to-one matching and why list size matters

There are two ways to use a fingerprint, and the difference decides how well the system ages.

Identification asks the reader to search every enrolled template and work out who you are. It is the convenient mode — walk up, touch, go — and it degrades as the list grows. Search time climbs, and the statistical chance of a wrong match climbs with every template added.

Verification asks a narrower question. You present a badge, the system retrieves the one template tied to that badge, and the finger only has to confirm it. The comparison is one against one no matter how many people are enrolled, so accuracy does not drift as the company grows from twelve people to ninety.

This is why we specify card plus finger on the highest-consequence doors in a suite rather than fingerprint alone. It scales, it survives an acquisition or a merged tenant roster, and it gives an auditor a clean answer to the only question they ever ask: can two people use one credential? Enrolment quality is the other half. We capture two fingers per person and check the quality score at the desk, because a marginal enrolment does not fail on the day it is taken — it fails six weeks later at the server room door with a visiting auditor watching.

Egress, fire alarm interface and ADA

These are the parts of the job that are not negotiable and not a matter of preference.

Getting out. People must always be able to leave without a credential, without special knowledge and without power. On most suite doors that is satisfied by the lever on the inside working mechanically at all times. Where electrified locking is used on a door in an egress path, the release arrangements are prescriptive, and a magnetic lock in particular brings requirements that a card reader alone does not satisfy.

Fire alarm release. Access-controlled doors in the egress path must release on alarm. That interface is a supervised connection from the building’s fire alarm system, and on a leased suite it is made by the building’s fire alarm contractor with our controller providing the input. We coordinate it and test it with them; we do not open their panel.

Accessibility. A reader is an operable part, so it has to sit within the reach range and be usable without tight grasping, pinching or twisting of the wrist. In practice that means mounting height, approach clearance and where the reader sits relative to the door swing get decided during the survey, not when the installer is already on a ladder.

Clinical suites carry the extra constraint that doors cannot be taken out of service during patient hours, which is a scheduling problem rather than a technical one but sinks more small projects than either.

Employee notice, retention and the audit trail

Enrolling staff fingerprints is a workforce decision before it is a hardware decision. Texas regulates the capture and use of biometric identifiers for commercial purposes, and the structure of the law is notice and consent before capture, use limited to the stated purpose, reasonable care in storage, and destruction once that purpose has ended. Your employment counsel should own the policy language. Our job is to deliver a system that can actually do what the policy promises:

• Templates stored on the controller or reader, never as retrievable images, and never shipped off-site by default.
• An enrolment record with a date, so retention can be measured from something real.
• Single-action deletion at offboarding that removes the template itself, not merely the person’s access rights.
• A non-biometric path for any employee who declines — the same door, opened by badge and PIN — so consent is genuinely a choice.
• Reports that show door, credential, time and result, with no biometric data in the log at all.

Two-thirds of the value of the system turns up during an incident review. A log that reliably distinguishes the badge that was presented from the person who was verified is the difference between an answer and a shrug.

How EVOTECH runs a suite installation

1. Walkthrough. Every candidate opening measured: frame material, leaf preparation, existing hardware, swing and whether it sits in an egress path.
2. Pathway assessment. Ceiling access, plenum status, existing tenant conduit, IDF location and distance to the furthest reader.
3. Building compliance package. Scope narrative, marked drawing and insurance certificate submitted to management, and after-hours windows booked.
4. Design and itemised quote following the free on-site estimate, separating hardware, cable, door hardware, programming and coordination.
5. Rough-in. Plenum-rated cable pulled, penetrations fire-stopped, controller and battery-backed supply mounted in the IDF.
6. Hardware and trim-out. Strikes and electrified locks fitted, readers set at surveyed heights, RS-485 bus terminated and tested end to end.
7. Fire alarm coordination. Interface scheduled, wired to the building contractor’s supervised output, and witness-tested.
8. Enrolment and cutover. Two fingers per person with quality scores verified, schedules and holiday calendars built, secure-channel encryption enabled, then a documented test of every door in both directions.
9. Handover. As-built drawing, credential list export, and training for whoever will add and remove staff.

What the estimate separates, and what moves it

We quote this work only after seeing the suite, because two offices with identical square footage can differ by a wide margin on identical door counts. These are the real variables:

• Door count and door type. A hollow metal frame accepting a strike is routine; a frame that needs a full electrified lock, or a pair of doors with a vertical rod device, is a different scope entirely.
• Cable distance and ceiling access. An open accessible ceiling is fast. Hard lid, ductwork and a long run back to the IDF are not.
• After-hours and access constraints. Clinical hours, quiet hours and freight windows lengthen the schedule.
• Landlord requirements — core drilling approvals, fire-stopping details and make-good obligations.
• Reader grade. A sensor that reliably reads dry or worn fingertips costs more than one that does not, and it is worth it on a door people use daily.
• Integration depth. Standalone door control, versus tying into a video system, a visitor process or an existing badge population from another site.

Every EVOTECH quote is itemised so you can see hardware, cable, labour and coordination separately and cut scope deliberately rather than guessing. We do not give a figure over the phone for access control, and we do not bid a suite we have not walked.

Related services

Frequently asked questions