System Commissioning: The Phase That Separates Professionals from Installers

Installation is the easy part. Any crew can mount cameras, pull cable, and rack equipment. Commissioning is where the gap between a professional integrator and a cable-pulling outfit becomes undeniable. It is the systematic, documented process of verifying that every device, every connection, every software configuration, and every integration point functions exactly as designed. It is also the phase that most projects either skip entirely or rush through with a clipboard and a prayer. The result is a system that works on demo day and falls apart within six months.

This case study walks through the commissioning process for a real-world project: a 200-door, 400-camera mixed-use campus comprising two office towers, a parking structure, and a retail concourse. The project included access control (200 doors with online readers), IP video surveillance (400 cameras on a segmented network), intrusion detection (85 zones across retail and office spaces), and a unified management platform tying all three systems together. What follows is not theory. It is a proven process that passed acceptance testing on the first attempt.

Acceptance Test Procedures: Defining "Done"

Before a single test is performed, both parties, the integrator and the client, must agree on what constitutes successful commissioning. This agreement is the Acceptance Test Procedure (ATP), a document that defines every test, its acceptance criteria, and the method for recording results. Without an ATP, commissioning devolves into subjective arguments about whether the system "works."

For this campus project, the ATP was a 47-page document developed during the design phase and refined during submittals. It covered 1,400+ individual test points organized into six categories: infrastructure verification, access control, video surveillance, intrusion detection, integration and automation, and network and cybersecurity. Each test point had a pass/fail criterion, a designated tester, and a sign-off field for the client's representative.

Point-to-Point Verification: Testing Every Wire

Point-to-point verification confirms that every cable run terminates where the drawings say it does, passes performance testing, and is labeled correctly at both ends. On this project, that meant certifying 400 Cat6A drops for cameras, 200 multi-conductor runs for door hardware (locks, REX devices, position switches), 85 home runs for intrusion zones, and approximately 150 additional runs for intercoms, elevators, and ancillary devices.

Every Cat6A cable was tested with a Fluke DSX-8000 certifier to TIA-568.2-D Cat6A permanent link standards. Cables that failed were retested at the patch panel; if they still failed, the termination was reworked and retested. We maintained a 98.4% first-pass rate across 400 drops, with the 1.6% failure rate almost entirely attributable to termination defects at the camera end (field conditions with tight junction boxes). Every door control cable was continuity-tested conductor by conductor and meggered at 500V to verify insulation integrity. This is the work that prevents intermittent lock failures and phantom alarms months after go-live.

Camera Aiming and Scene Optimization

Mounting a camera is not the same as commissioning one. Camera commissioning involves verifying that each camera delivers the image quality and coverage specified in the design. For this project, every camera was tested against five criteria:

• Pixels Per Foot (PPF) verification: Using a test chart or known-distance reference, we confirmed that identification cameras delivered 80+ PPF (per SIA/ANSI ONVIF Profile M guidelines), recognition cameras delivered 40-80 PPF, and detection cameras delivered 20-40 PPF at the target distances specified in the design drawings.
• Wide Dynamic Range (WDR) tuning: Cameras facing windows, loading docks, and building entrances were adjusted for backlight compensation. WDR was enabled and the strength was tuned until both the bright and dark areas of the scene were properly exposed. This is a per-camera adjustment that cannot be set globally.
• IR illumination adjustment: Night-mode performance was verified between 10 PM and 2 AM during commissioning. IR intensity and focus were adjusted to eliminate hotspots and ensure uniform illumination across the field of view. Exterior cameras with nearby reflective surfaces (glass, metal) were repositioned or fitted with IR hoods to prevent whiteout.
• Stream configuration: Primary stream (recording) and secondary stream (live viewing) were configured per the VMS recording profile: primary at full resolution, 20 fps, H.265 CBR; secondary at 720p, 7 fps, H.265 VBR with a max bitrate cap. Audio was enabled on cameras in lobbies and interview rooms per the client's policy.
• Analytics configuration: Cameras with edge analytics (line crossing, loitering, people counting) were configured and tested with live walk-throughs to verify detection zones, sensitivity thresholds, and alarm routing to the VMS event engine.

Access Control: Door-by-Door Verification

Access control commissioning is the most labor-intensive phase because every door has multiple components that must be tested in combination: reader, lock (electric strike, mag lock, or electrified hardware), request-to-exit device (REX), door position switch (DPS), and the controller I/O point mapping. A single wiring error on one conductor can result in a door that unlocks on invalid credentials, fails to relock, or does not report its state correctly to the management platform.

For each of the 200 doors on this project, we performed the following tests: valid credential presentation (grant access, lock releases, DPS reports open/closed, event logged), invalid credential presentation (deny access, lock remains engaged, event logged), REX activation (lock releases, no alarm, re-locks after timeout), door held open beyond the timeout threshold (alarm triggers in software), door forced open (immediate alarm), lock power failure behavior (fail-safe doors unlock, fail-secure doors remain locked per life-safety requirements), and fire alarm relay integration (all doors on the fire alarm release relay unlock simultaneously during alarm). Every test was recorded in the ATP with the door number, test result, tester initials, and timestamp.

Intrusion Detection: Zone Walk Tests and Timing Verification

Intrusion zones were tested using systematic walk tests. For each of the 85 zones, a technician armed the area and physically triggered every sensor while a second technician monitored the receiver at the central station connection. PIR motion detectors were tested by walking through the detection pattern at normal pace from multiple angles. Door contacts were tested by opening the protected door. Glass break detectors were tested using an approved simulator (such as Honeywell FG-701 or a calibrated clap tester) at the rated distances.

Entry/exit delays were configured per the client's operational requirements: 30-second entry delay for main lobbies, 15-second entry delay for secondary entrances, and zero delay (instant) for interior perimeter zones. Exit delays were set to 60 seconds for all armed areas. Cross-zone verification was enabled on high-security areas (server rooms, executive suites) requiring two zones to trip within a 60-second window before alarming, which virtually eliminates false dispatches while maintaining detection integrity.

Network Validation and Failover Testing

Network commissioning verified the logical configuration of the surveillance and access control network: VLAN assignments, inter-VLAN routing, QoS policies, IGMP snooping, and redundancy mechanisms. We performed bandwidth saturation tests using iPerf3 to confirm that 10GbE uplinks between distribution switches and the NVR cluster sustained 9.2+ Gbps without packet loss. VLAN isolation was verified by attempting to ping camera IPs from the corporate VLAN (all attempts blocked as expected). SNMP polling was confirmed operational for all managed switches, providing real-time port status, PoE draw, and bandwidth utilization to the network monitoring system.

Redundancy testing involved physically disconnecting primary uplinks and verifying RSTP convergence within 2 seconds, VRRP gateway failover within 3 seconds, and NVR recording continuity (no frames dropped) during network path changes. This testing was performed during off-hours with client authorization and documented with packet captures showing convergence timing.

Commissioning Checklist Summary

As-Built Documentation: The Deliverable That Matters Most

When commissioning is complete and the ATP is signed, the final deliverable is the as-built documentation package. This is not a photocopy of the design drawings with a red "as-built" stamp. It is a comprehensive reference that enables the client's team and any future integrator to understand, maintain, and expand the system without starting from scratch.

For this project, the as-built package included: updated floor plans showing every device location with cable labels and IP addresses, riser diagrams showing the complete network topology from core through access layer, rack elevation drawings for every IDF and MDF showing equipment placement and power connections, a complete IP address schedule with VLANs and device assignments, door hardware schedules with lock types and fail-safe/fail-secure designations, camera schedules with model numbers and stream configurations, credential database with card format specifications and access level definitions, default credential changes log confirming all factory passwords were replaced, and the signed ATP with all test results.

End-User Training and Warranty Handover

Training is not a courtesy; it is a contractual deliverable that directly impacts system effectiveness and callback frequency. Untrained operators misuse software, ignore alerts, and call in service requests for issues that are user error. For this campus, two levels of training were delivered: operator training (8 hours covering daily use, alarm response, credential management, video search and export) and administrator training (16 hours covering system configuration, report generation, firmware updates, and troubleshooting procedures). Both sessions were recorded and provided to the client for onboarding future staff.

Warranty terms were clearly documented: 1-year parts and labor warranty covering all installed equipment and workmanship, with defined SLA response times (4-hour response for system-down events, next-business-day for non-critical issues). Punch list items identified during commissioning (11 items on this project, including one camera repositioning and two door closer adjustments) were resolved within 5 business days of the ATP sign-off, at which point the warranty period officially commenced.

The Standard You Should Demand

This commissioning process added approximately 12% to the total project cost compared to a walkthrough-and-handover approach. It also delivered a system that passed acceptance testing on the first attempt, generated only 3 warranty service calls in the first year (all related to environmental factors, not system defects), and received the client's approval for a multi-year service contract extension. The ROI on commissioning is not abstract; it is measured in reduced callbacks, reduced liability, and a client relationship that generates recurring revenue.

At Zimy Electronics, formal commissioning with documented Acceptance Test Procedures is standard on every project, not an optional add-on. A rigorous commissioning process like the one described here should be standard on every project regardless of size, from a 10-door office build-out to a 500-camera campus. It is the single practice that most consistently separates reliable deployments from systems that erode trust within months of handover.