Abstract
Contracted work carries every readiness risk of in-house work plus one more: the person who finds the gap has no history with the asset, no relationship to lean on, and a commercial incentive not to raise it until it is billable. This paper argues that contractor packages are the highest-value place to apply a readiness review, that the failures are almost never about competence, and that the plant's entire negotiating position is spent before anyone mobilizes.
The incentive is inverted, and nobody is behaving badly
An in-house tradesman who finds a scope gap raises it, because he will be the one standing there at midnight if he does not.
A contractor who finds the same gap has a variation order and a defensible claim.
Neither party is behaving badly. The structure produces the behaviour, and every argument in this paper follows from that sentence. If you go into a turnaround believing your contractors are adversaries, you will manage them badly and they will respond in kind. If you go in believing good intentions substitute for a complete package, you will pay for the gap at the contractor's rate, on the contractor's schedule, with the line down.
Package quality is not an administrative nicety. It is the principal commercial control the plant holds, and it is exercised entirely before mobilization. Afterwards you are negotiating from the weaker position on every discovered unknown.
A day that looked good to go
We had a planned down day to swap the camelback chains on the log decks. Three contractor crews, because the window had to be hit and one crew could not do it in the time.
On paper it was ready. Chains staged, torches out, crews onboard. Everything you could look at and count was there.
Then it came apart on everything nobody had looked at, and every single failure was multi-company.
One crane. Nobody had thought about it. Three crews and one crane means somebody is always standing around waiting to fly their chains up. In-house, crews trade off and it sorts itself out. Between competing contractors it does not, because waiting is billable and cooperating is not.
Two torches for three crews, given out on the assumption they would share. Competing contractors do not share. I watched them fight over the torches, unplug each other's welders, and deliberately tie up the crane so the other crews could not lift.
Three separate lockouts. Three companies means three sets of paperwork and the whole isolation run three times over, on a deck packed with bodies. That is not a safety inconvenience, it is hours of window consumed by an interaction nobody had planned for.
Nobody had confirmed the guys were rated to run the iron they were grabbing.
Then a crew that was behind and rushing cut the worst corner of the day. Instead of pulling a chain the right way they hooked the crane straight to it, caught it on the catwalk, and ripped thirty or forty feet of catwalk out of the building. With a camera pointed right at them. Then lied about it.
The chain was fine. The catwalk had to be cut out and replaced. That crew went home.
Total cost: the shutdown for the safety investigation, the extra down days to finish the chains, the tear-out and reinstall of the catwalk, and the production lost working around a missing catwalk until it was back. The better part of a month, out of a day.
What that day actually teaches about packages
Read the failures again and notice that not one of them is a competence failure. Every one of those crews could change a camelback chain.
The failures were all in the spaces between the packages. Shared resources, shared access, shared isolation, and shared assumptions about behaviour. A package written for a single contractor describes that contractor's work. Nothing in it describes the deck with three crews on it.
That is the specific thing a multi-contractor scope needs and almost never has. Not better individual packages. A document that says who else is on this deck, what iron is shared, how many isolations are running, and what the sequence is when two crews want the same crane.
We built an asset-reservation system after that day, and started running real down-day meetings with area leads and crews in the room rather than just planners. Reserve the iron the same way you reserve the window.
The competent stranger test
A contractor package has to answer questions from somebody with no local knowledge at all. That is a higher standard than an internal work order, and it exposes assumptions internal packages are permitted to leave implicit.
The test is simple and unforgiving. Can a competent stranger determine the boundary of the work, the isolation, the access route, the parts and their location, the specialty tooling, the coordination dependencies, and the completion criteria, without asking anyone?
Every question the package cannot answer is a variation order with a delay attached.
Internal packages get away with omissions because the millwright knows where the crib is, knows which valve that is, and knows to call the area lead before touching the conveyor. None of that travels.
Verify the paperwork, not the assurance
I was told, flat out, that all the hands on site were certified to run the telehandler. I checked one down day, and not one guy was. Not one, and I had been told they all were.
Contractors will actively lie about certifications, or pass along a lie somebody told them, because "yeah, he's ticketed" is the answer that gets them on site. The day you find out otherwise is the day the machine is on its side or a man is under it.
A ticket is a piece of paper. Ask to see the paper.
I want to be fair about this, because it is not simple dishonesty. The man giving you that assurance may genuinely believe it. He was told by a supervisor who was told by a scheduler who was working from a list that was accurate two years ago. The assurance is sincere and the document does not exist. That is why the answer is to look at the paper rather than to judge the character of the person speaking.
The shortcut that bills later
Contractors slammed a coupler on using a forklift counterweight as a press. I found out later they had not seated it all the way, and it caused problems downstream.
A shortcut that skips the right tool does not save time. It moves the cost downstream into a period where nothing will connect it back to the decision that caused it.
That is a package failure too, and a specific one. The package said change the coupler. It did not say how, it did not specify the tooling, and it did not define what "seated" looks like or who verifies it. A completion criterion that a stranger cannot check is not a completion criterion.
Loops that never get closed
Contractors were in one of our plants and unplugged a pond temperature sensor. They never plugged it back in, and nobody verified their work.
The sensor fed back a comfortable number for months while the pond sat near boiling. We burned a pump seal, wasted a day of my time chasing a collapsing suction hose, and heated millions of litres of water to ninety-six degrees for several months before anyone noticed.
Nothing about that is exotic. It is the most ordinary contractor failure there is: work that touches something adjacent to the scope, and a handback that nobody checks.
Every contractor package needs a section that does not exist in most of them. What did you disturb that was not in scope, and what state did you leave it in? Isolations restored, sensors reconnected, guards refitted, temporary supports removed, settings returned. Signed by name.
The reason it is missing is structural. The package describes work to be done to an asset, and disturbance to adjacent systems is not work done to the asset. So it has no field, and what has no field does not get recorded, and what does not get recorded does not get checked.
Where the review pays first
Run the readiness gates against contractor packages before any internal backlog.
The work is bounded, the packages are already written, the failure cost is measured in variation orders rather than goodwill, and the finding is defensible in commercial terms that a general manager already understands. "Eleven of our forty packages have no completion criteria a stranger could verify" is a sentence that survives contact with a finance meeting.
It is also the cleanest early application for automated gap detection. The input is a document, the output is a list of questions. No write access, no actuation, no authority question to settle first.
Vendors are already selling into this ground. IAMTech markets AI-generated work packs built from work descriptions, drawings, photographs and engineering documents. Prometheus promotes planning automation on time-to-produce.
Treat both as vendor claims rather than independent validation, and notice what the claims are actually about: the speed of producing a package, not the completeness of its content. A pack generated in four minutes that still cannot tell a stranger the access route has automated the half that was never the bottleneck.
The question to put to any of them is the one from the story above. Show me where the generated package says what was disturbed outside the scope and who signed that it was restored.
Conclusion
Most plants apply their least rigorous readiness process to the work where they hold the weakest negotiating position and pay the highest hourly rate. Reversing that is a matter of sequence rather than investment.
The chain day did not fail because three contractors were bad at changing chains. It failed because one crane, two torches and three lockouts were nobody's package, and because the only document describing how three companies share a deck was the one nobody wrote.
References
- IAMTech, “iPlanSTO / AI-powered planning & work pack generation.” https://go.iamtech.com/iplan-turnaroundinsurance/
- Prometheus Group, “From Hours to Minutes: How AI Is Transforming Maintenance Planning and Scheduling,” June 26, 2026. https://www.prometheusgroup.com/resources/posts/from-hours-to-minutes-how-ai-is-transforming-maintenance-planning-and-scheduling
- Misra, Vyas, Gutta & Mercangöz, “ADMITBench: A Safety-Governed Reference Framework for Evaluating the Admissibility of Industrial LLM Advisories,” 2026. https://arxiv.org/abs/2608.03866
- Oracle, “AI Agent: Maintenance Work Order Builder.” https://docs.oracle.com/en/cloud/saas/readiness/scm/25d/maint25d/25D-maintenance-wn-f40777.htm