Commercial Access Control in Stafford, TX 77477

In Stafford the trouble usually starts at door number six

One door is a product decision. Six doors across two tilt-wall buildings, with a dock, a yard gate, three shifts and a rotating contractor list, is an engineering decision – and the mistakes are structural rather than cosmetic. They are also expensive to undo, because they are buried in wall and conduit.

The typical 77477 site is a tilt-wall multi-tenant building: concrete panels, bar joists, an open warehouse behind a small office build-out, a dock or grade-level door, and a man-door in the panel beside it. Leased, usually, which adds a landlord to every decision. What makes it different from an office is the mix of openings – front entry, office-to-warehouse, dock man-door, yard gate, server room – each with a different population allowed through it at a different hour.

So the first design document is not a product list. It is a table of openings against groups of people against times. If a proposal skips that, the system gets redesigned in year two.

Where the intelligence lives: panel, edge device, or hosted

Three architectures dominate, and the building decides which fits.

Two constraints bite here specifically. Network cable is good for roughly a hundred metres of channel, and a long tilt-wall building plus a routed path through the office exceeds that easily – so a second equipment location, or fibre between them. And a warehouse is electrically noisy: trays shared with line-voltage feeders and drive-controlled motors cause genuinely intermittent faults, so separation matters more here than in an office fit-out.

OSDP or Wiegand – the wiring decision you make only once

Between the reader and whatever makes the decision there is a protocol, and it is one of the few choices that is genuinely hard to change later, because it is in the wall.

Wiegand is the legacy one: one-way, unencrypted, sending the credential number down two data wires in the clear. Anyone with access to the reader – which is by definition the unsecured side of the door – can fit a small device behind it, collect numbers and replay them. It also has a modest distance limit and gives the reader no way to report tamper.

OSDP is the current answer: a two-wire RS-485 bus, multiple readers on one run, bidirectional so the reader reports tamper and takes configuration, and with Secure Channel enabled it encrypts the link to the controller. It also runs considerably further on the same gauge of wire.

Our position is simple: on any new pull we run the pair OSDP needs and enable Secure Channel, even when today’s reader could live with less. Wire is the cheap part on install day and the expensive part on the day it must be replaced.

The badge in your pocket may be copyable in the parking lot

This is the part of the industry that has aged worst, and most sites in the area are still living in it.

Older 125 kHz proximity credentials – the thick white cards standard since the nineties – transmit a fixed number with no authentication. Handheld cloners that copy one in seconds are sold openly. If your badge is one of those, treat the number as public information. The same goes for the first generation of 13.56 MHz smart cards, whose encryption was broken publicly years ago.

Use instead a modern 13.56 MHz credential with mutual authentication and a diversified key, or a mobile credential on a phone. Both make the card and reader prove themselves to each other, so capturing the exchange yields nothing reusable. Keep the site key under your control rather than accepting a supplier default, and stop printing the badge number on the card face.

Migration need not be a forklift: dual-technology readers read both formats, so new credentials go out as people are onboarded and the old format is switched off on a date. That is how we stage it for a plant that cannot stop for a weekend.

Power supplies, batteries, and the inrush nobody budgets for

Multi-door systems fail on power far more often than on software, and the causes are arithmetic.

Start with voltage drop. Eighteen-gauge conductor runs about six and a half ohms per thousand feet, so a lock two hundred feet from the supply sits at the end of roughly four hundred feet of copper – near two and a half ohms. Half an amp of holding current loses more than a volt before it arrives. Twelve-volt hardware does not have that volt spare; twenty-four does, which is why we specify it where the hardware allows and size conductor by run length rather than habit.

Then inrush. Motorised latch retraction on an exit device draws far more at the instant of pull-in than at rest. Put four on one undersized supply and they work fine one at a time, then brown out when a shift change hits three at once. Those get a dedicated supply sized to simultaneous operation.

Then standby. Batteries keep doors, controllers and the event log alive through an outage, and they are consumables – sealed lead-acid cells lose capacity steadily and want replacing every few years. We date-label every battery, size the bank to the real connected load, and put the interval in the handover document, because a bank nobody has touched in six years is decorative.

What the inspector will ask you to demonstrate

Expect to be asked to prove the system releases, not to describe it. That test is easy to pass when the design was right and impossible to talk your way through when it was not.

Fail-secure hardware – electric strikes, electrified exit devices – stays locked from outside on loss of power while the inside lever or panic bar still opens it by hand. Mechanical egress is never interrupted, which is why it is our default in any exit path. Fail-safe hardware, principally magnetic locks, unlocks when power stops, so it must be tied to the fire alarm to release on alarm, fitted with a release device on the egress side and a labelled button that cuts lock power directly.

In a warehouse there is one more habit to design against: the dock man-door propped open with a pallet. No lock addresses that. A door position switch with a held-open timer, an annunciator audible over the floor and a camera at that opening do – and the timer must be realistic, because an alarm that cries wolf is ignored within a week.

Shift schedules, contractors, and the five in the morning problem

The office model – unlock at eight, lock at five – does not survive contact with a plant. What we build instead is schedules against groups: production on their shift window plus a grace margin, office staff on business hours, maintenance wider, a short list unrestricted, and a contractor group whose credentials carry an expiry date set at issue rather than a reminder nobody acts on.

Two refinements earn their keep at this scale. First-credential unlock means a scheduled door does not open itself at five in the morning before anyone authorised has arrived – the first valid badge starts the schedule. Anti-passback on a dock or turnstile stops one credential walking a group in. But the highest-value change is one nobody buys: a process where the event that ends someone’s employment also ends their credential the same day. The commonest serious audit finding is not a hacked reader – it is a live badge held by someone who left months ago.

Six things that bring a truck back to an industrial park

• Lock and controller sharing a power supply. The lock releases, dumps a transient on the shared rail, the panel reboots – intermittently, at the worst times.
• Reader cable in a tray with motor feeders. Reads fail only when a particular machine runs, which nobody connects to the door for months.
• Cable shield grounded at both ends. That creates a loop and injects the exact noise the shield was bought to stop. One end only.
• Door contacts on the hinge side, where the gap barely changes – so a wide-open door reports closed.
• No held-open timer, or one set to five seconds. Either the propped door is invisible or the alarm becomes background noise.
• A battery bank never replaced, discovered during the first real outage.

What a Stafford install week actually looks like

We plan multi-door work around your production. Pathway and rough-in come first and disturb nobody: conduit, sleeves and firestopping through rated walls, cable pulled and labelled at both ends, backboard mounted and powered.

Door hardware goes in opening by opening, each finished and tested before the next starts, so the building is never left with several half-converted doors. Mechanical precedes electrical every time – closer, latch alignment, hinges – because a door that does not close reliably will not lock reliably.

Cutover is scheduled: credentials loaded and tested in advance, schedules reviewed with whoever owns the shift calendar, old keys retired on a defined date. Then the demonstrations, with you present – power pulled, alarm interface tested with the alarm company, a credential revoked and re-tested, forced and held-open alarms triggered on purpose – followed by drawings, labels, the credential list and the part numbers you will want in two years.

What moves a multi-door number

No dollar figures here – an honest multi-door estimate is impossible without walking the site. What drives it:

• Opening count and type. A man-door in a tilt panel, a glass front entry and a yard gate are three different jobs.
• Topology and distance – one equipment room, or a second location and fibre.
• Existing infrastructure. Reusable conduit helps; a previous installer’s undocumented cable rarely does.
• Credential strategy – staying put, or migrating to encrypted credentials or phones, and for how many people.
• Power and standby – supply count, 24-volt hardware, conductor sizing, battery capacity.
• Fire alarm interface work, where the lock selection requires it.
• Schedule. Cutting over around three shifts costs more than a single-shift building – and we would rather charge for off-hours than stop your line.

Related services

Frequently asked questions