Your Records Don't Match What You See On Site: Decision Tree

Why this matters

You pull up the job and your paperwork says one thing, but the equipment in front of you says another. Different part, different location, different settings, a component you have no record of installing. This mismatch is not a nuisance to wave off; it is data. It usually means something changed after you left, and the gap between your record and reality often points straight at the fault. But it can also mean your record is wrong, and trusting the wrong source sends you chasing a ghost. This tree sorts it out.

Start here: trust the live system for safety

Whatever your records say, the physical system in front of you is what can hurt someone. If what you see includes a hazard, a defeated safety, an energized surface, a pressure or combustion risk, act on what is there, not on what your file claims should be there. Make it safe first, then reconcile the paperwork. Records never override the live hazard.

Step 1: decide which source is more likely right

Your record and the site disagree. One of them is stale. Weigh them.

• The site is current by definition; it is the system as it exists today. Its risk is that you are misreading it.
• Your record is a snapshot from install day. Its risk is that it was incomplete then, or the system changed after.

Neither is automatically the winner. A clear, photo-backed as-built beats a hazy memory; a vague invoice line loses to what you can see and measure.

Step 2: name the type of mismatch

Different mismatches mean different things.

• A component is different from what you recorded installing. Someone likely swapped it. The replacement is a suspect if the fault is near it. See related: A Later Modification Caused the Fault On Your Install.
• Something is present that you have no record of. A later addition or tie-in. Read it as a change since you left. See related: Reading What Changed Since You Left the Job.
• Settings or configuration differ from your commissioning notes. Someone adjusted them, or your notes were wrong. Verify against the equipment's correct range before assuming either.
• The whole unit or location is different. You may be looking at a full replacement someone did, not your install at all. If so, stop treating it as your work.

Step 3: reconcile with a third source

When record and site conflict and you cannot tell which is right, find a tiebreaker.

• Serial and date codes on the equipment tell you whether a unit is the one you installed or a newer replacement.
• Physical evidence of age confirms which parts are original and which are recent. See related: Separating Your Original Work From Later Changes.
• The customer's account can fill gaps, though treat it as a lead, not proof, especially if the caller is not the person who was there for the changes.

Step 4: correct the record, then diagnose

Once you know the true current state, update your file to match reality before you go further. Diagnosing off a record you already know is wrong guarantees a wrong answer. Note what changed and, if you can tell, when.

• If the mismatch is the fault, you have found it: something changed and that change is your suspect.
• If the mismatch is incidental, set it aside and diagnose the actual complaint against the corrected picture.

Recap

1. Act on the live system for any hazard, regardless of what the record says.
2. Weigh which source is more reliable; a photo-backed as-built beats memory.
3. Name the mismatch type: swapped part, unrecorded addition, changed setting, or different unit.
4. Break ties with serials, age evidence, and the customer's account.
5. Correct the record to reality, then diagnose against it.

References

• Trade-standard practice for as-built documentation and field verification
• Manufacturer serial and date-code references for equipment identification
• See related: Reading What Changed Since You Left the Job; Using Your Own Install Notes to Diagnose Faster