Drive-By Cx, Defined
The term isn't in any standard, so let's fix that. Drive-by commissioning is commissioning performed as documentation rather than verification — and it is sold under the same name as the real thing.
There is a version of this job where a commissioning authority shows up once, at 95 percent complete, walks the mechanical room for two hours with a phone camera, collects the contractor's startup sheets, and goes home. Six weeks later the owner gets a 180-page PDF. It has a cover page with the project rendering on it. It has an issues log with eleven items, nine of which are "provide O&M manuals" and "label piping." It has a signature. It gets the certificate of occupancy released.
Nobody tested anything.
That is drive-by commissioning, and I want to be precise about what makes it that, because the failure is not laziness. Plenty of drive-by Cx is performed by people working hard. The failure is definitional: drive-by Cx treats the report as the deliverable and the building as incidental. Real commissioning inverts that. The building is the deliverable. The report is a record of what you did to it.
Once you see that inversion, every other symptom follows from it.
The tells
You can usually spot it from the documents alone, without ever setting foot on site.
1. There is no Owner's Project Requirements document, or it was written by the CxA in week 40. The OPR is the whole basis of the process. It is the thing you test against. If the OPR appeared after the systems were installed, it was reverse-engineered from the submittals, which means it cannot possibly be a statement of what the owner wanted. It is a description of what the contractor built.
2. The commissioning plan is a template with the project name find-and-replaced in. Look for the giveaway: a scope section that lists systems the project does not have. Chilled beams on a project with rooftop units. A cooling tower in a building with air-cooled condensers.
3. The functional test procedures are generic. A real functional test script is written against this project's sequence of operations, pulled from this project's controls submittal. If the test procedure says "verify economizer operates per sequence," it is not a test procedure. It is a note to self.
4. The issues log has no failures in it. This is the loudest tell. Every real building has failures. Every one. A 300,000-square-foot project with a 14-item issues log and no controls findings did not have a good contractor; it had an absent CxA. My rough heuristic: a genuinely commissioned new building surfaces somewhere between 0.5 and 2 issues per 1,000 square feet, and a meaningful share of them are sequence and controls problems, not labeling.
5. Issues are logged but never closed. Or they're closed with "contractor states corrected." Nobody retested. A finding that was never verified as fixed is not a finding, it's a rumor.
6. Testing happened once, in one season. Heating tested in July with the boilers forced on by overriding the outdoor air sensor, and nobody came back in January. Seasonal testing gets dropped first because it falls outside the construction schedule and therefore outside the fee.
7. The CxA was hired by the contractor, or by the design engineer who designed the systems. Then the CxA is being asked to grade work their own client performed. Some firms manage this honorably. The structure is still wrong, and under pressure the structure wins.
8. The training deliverable is a video of a startup tech talking for forty minutes. Facilities staff who cannot explain the sequence of operations six months after turnover were not trained; they were present.
Why this is a different product, not just bad work
An owner comparing three commissioning proposals at $18,000, $34,000, and $71,000 for the same building reasonably assumes they are buying the same thing at three price points. They are not. The $18,000 proposal is not a discounted version of the $71,000 one. It is a different service with the same name on the invoice.
This is what makes drive-by Cx corrosive rather than merely disappointing. It is not competing on quality within a market. It is arbitraging the fact that the buyer cannot tell the difference until years later, if ever — because the whole point of commissioning is to catch problems the owner would otherwise never know existed. The counterfactual is invisible. When a building runs badly, nobody says "the commissioning failed." They say "the design was bad" or "the controls contractor was bad" or, most often, they don't say anything, because a building that quietly burns 25 percent more energy than it should, with three air handlers fighting each other and an economizer that has been locked out since turnover, does not announce itself.
Markets where the buyer cannot evaluate the product converge on the cheapest offer. That is not a moral claim about anyone in this industry; it's just what happens. It is happening to us right now.
What this site is for
The term "drive-by commissioning" does not appear in ASHRAE Guideline 0, in Standard 202, in the BCxA best practices, or in any LEED reference guide. There is no definition to point at. So the industry has no shared word for the specific thing that is eating it, which makes it very hard to argue about.
I'm proposing the word, and I'm going to use it precisely: commissioning performed as documentation rather than verification. Not "commissioning I disagreed with." Not "commissioning done by a competitor." Documentation instead of verification, sold to a buyer who cannot tell.
The rest of this site is the argument.