Security is the trade everyone forgets until the building is almost done, and then it's the trade that holds up your temporary certificate of occupancy. Access control that won't release the mag-locks on a fire alarm signal, cameras aimed at a wall because the millwork moved six inches, a head-end rack sitting in a boxed pile because the IDF room isn't painted yet — I've watched all of it stall a turnover. The frustrating part is that almost none of it is hard. It's a sequencing problem, and sequencing problems are solvable if you plan them before the drywall goes up instead of after.
This is a walk through how security system work actually lands on a jobsite — the rough-in, the pathways, the head-end, devices, cameras, access control, programming, and the commissioning gauntlet at the end — with the timing and the coordination traps that bite crews who treat it as a single line item on the schedule.
Rough-In: The Part You Can't Fix Later
Security rough-in rides along with electrical and low-voltage rough-in, right after framing and before insulation and drywall. This is the one phase where a mistake is permanent and expensive. Miss a home run, forget a back-box for a card reader, or leave out the conduit stub for an exterior camera, and you're either cutting open finished wall or running surface raceway that the architect will hate.
A few rules that save you:
- Walk the rough-in against the finish schedule, not just the framing. Device heights and locations are driven by finished floor and finished wall. A reader roughed to the stud line before you know the tile thickness ends up at the wrong height.
- Coordinate door hardware before you rough a single reader. The hardware submittal tells you strike vs. mag-lock, request-to-exit, door position switch, and whether power transfer is through the hinge or a Von Duprin EPT. Rough-in blind and you'll be back.
- Leave pull strings and slack. Coil eight to ten feet of service loop at every device and in the ceiling near the head-end. Cable is cheap during rough-in and impossible to add after the lid goes on.
Give rough-in an honest one-to-two-day buffer per area between "framing complete" and "inspection," because the framer is never quite done when the schedule says he is, and the inspector doesn't come the same afternoon you call. On a look-ahead board this is where you connect security rough-in as a successor to framing and a predecessor to insulation — if those ties are visible on a weekly work plan, the low-voltage foreman stops finding out about a closed wall the morning he shows up to pull cable.
Pathways and Conduit: Own the Ceiling Early
Structured cabling lives or dies on pathway. J-hooks, cable tray, sleeves through rated walls, and conduit stubs to inaccessible or exterior locations all have to be in before anyone pulls a foot of cable. The classic collision here is the ceiling: mechanical, plumbing, sprink, electrical, and low-voltage all want the same eighteen inches above the grid, and low-voltage is usually last in line and lowest in the pecking order.
Get into the MEP coordination early and claim your elevation. If you're doing any kind of BIM coordination, security pathway needs to be in the model, not an afterthought penciled in during install. Firestopping at every rated penetration is a real inspection item and a real schedule item — sleeve and pillow the penetrations as you go, because coming back to firestop a hundred sleeves at the end is a day you didn't budget.
Cable pulls follow pathway, and pulls want the ceiling open and the space reasonably clean but not yet finished. Sequence them after pathway is signed off and before the ceiling grid closes and the painters take the room.
Head-End and Equipment Rooms: Room-Ready Comes First
The head-end — your NVR/DVR stack, access control panels, the security appliance servers — needs a finished, conditioned, powered room before it shows up. "Room-ready" for a security or IDF space specifically means: painted, sealed concrete or VCT floor down (you don't want raw concrete dust in a rack), HVAC running and holding temperature, permanent power and UPS circuits landed, grounding bar installed, and the room lockable. Racks and gear that land before the room is ready get covered in drywall dust and become a warranty argument.
The trap is that the equipment room is almost always on the critical path for the whole security system and it's easy to let it slip because it's "just a closet." Track its completion as its own milestone. When the head-end install date on your look-ahead is tied to the room-ready milestone rather than a calendar guess, the slip in painting that closet shows up as a slip in your commissioning date three weeks out — which is exactly when you can still do something about it.
Devices, Cameras, and the Finish Trades
Trim-out devices — keypads, readers at interior doors, motion sensors, glass-breaks, indoor cameras — go in after paint and near the end of finishes, so they don't get painted over, dinged, or stolen off an open jobsite. Terminate, don't just hang. A device that's mounted but not landed reads as "done" on a walk and bites you at test.
Cameras deserve their own attention because placement is where installers and designers disagree with reality:
- Field-verify every camera against actual sightlines and lighting. The design was laid out on a floor plan; the real room has a soffit, a sign, or a light fixture the drawing didn't. Aim for coverage of the thing that matters — the door, the register, the cash wrap — not a pretty symmetrical grid.
- Coordinate exterior cameras with the facade and site work. A camera stubbed for a soffit that's now EIFS, or aimed across a courtyard where the landscaper just planted a tree that'll be twelve feet tall, is a callback waiting to happen.
- Set and lock focus after the space is lit the way it'll actually be lit. Day-two lighting changes the image; commission the camera against final lighting, not work lights.
Access control is the highest-coordination item on the whole system, because it doesn't own its own doors — the hardware trade does. The reader, the electric strike or mag-lock, the request-to-exit, and the door position switch all have to agree with the door hardware that's already hung. And every access-controlled door on an egress path has to release on a fire alarm signal, which means the security integrator and the fire alarm contractor share a wire and a test. Get those two subs in the same room during coordination, not during the fire marshal's inspection.
Programming: The Invisible Two Weeks
Programming is where schedules quietly lose a week or two, because it's labor with no visible product. Nobody walks a job and says "the programming looks good." But mapping every device, building the access schedules and credential groups, defining camera recording rules and retention, and integrating with the building's directory or tenant system is real work with real duration.
Programming can start once the head-end is powered and networked, in parallel with the tail end of device install — it does not have to wait for the last device. Sequencing it to overlap installation is how you buy back the time it costs. Break it out as its own activity on the weekly work plan with its own crew and its own days; folding it into "device installation" is how it becomes the surprise that eats your float.
Integration Testing and Commissioning
This is the phase that actually decides whether you turn the building over on time, and it's the one most often crammed into the last three days. Don't. Integration testing means proving the parts talk to each other, not just that each part powers on:
- Access control to fire alarm: every egress door releases on alarm, and it's witnessed and documented. This is usually a fire marshal item — schedule it with the AHJ, not around them.
- Cameras to recorder: every camera records, retains for the specified days, and the image is usable at the resolution and frame rate specified. Pull footage and look at it.
- Readers to panels to head-end: a badge that's supposed to open a door opens that door and only the doors it should, and the event lands in the log.
- Elevator, intercom, and BMS integrations: anything that crosses to another system needs a joint test with that system's contractor present.
Point-to-point testing — confirming each individual device works — should be substantially done before you attempt integration testing, or you'll waste the integration window chasing single-device faults. Build a punch list from the tests and work it; "commissioned" means the punch list is closed, signed, and the AHJ is satisfied, not that the installer says it works.
Owner Training and Turnover
Training and turnover ride at the very end, near occupancy, and they're a scheduled deliverable, not a favor. The owner's security staff need hands-on time on the live system: adding and revoking credentials, pulling and exporting camera footage, acknowledging alarms, running reports. Hand over as-builts that reflect the actual installed device locations — not the design drawings — plus credential databases, admin passwords, and warranty documents. A system nobody on the owner's side can operate is a callback generator and a reputation problem, no matter how clean the install is.
Where the Schedule Actually Earns Its Keep
Every failure above traces back to a handoff that nobody could see coming: a wall closed early, a room not ready, a fire alarm test nobody coordinated, programming that started too late. The value of short-interval and look-ahead scheduling on security work isn't the pretty Gantt chart — it's making those handoffs visible three to six weeks out, while there's still time to move a crew or lean on a predecessor.
Location-based look-ahead tools like LookAheadWall are useful here precisely because security is a location-and-sequence trade: rough-in this wing, then close it; head-end ready in this closet, then install; test these doors with the fire alarm on this date. When the trade-flow ties between framing, hardware, fire alarm, and security are drawn on a weekly work plan the subs can actually see, the low-voltage foreman stops discovering closed walls and the super stops discovering, on turnover morning, that the mag-locks were never tested against the alarm. Plan the sequence, protect the buffers, and security stops being the trade that holds up your CO.