Put an aerial decontamination pass and a ground crew on the same surface and you have not built redundancy. You have built an attribution problem. Each subsystem can be individually competent and individually tested, and the programme can still fail at acceptance, because the question that decides the money is not did it work but whose residual is this.
Where the line actually falls
Combined programmes are usually specified as a coverage split. Air handles the exposed, the elevated, the wide; ground handles the enclosed, the delicate, the confined. Drawn on a map, the split looks clean. In execution it never is, for three reasons that are structural rather than operational.
First, the boundary is not a line but a band. Aerial coverage falls off at the edge of the pattern rather than stopping, so the transition is a gradient. Second, sequencing changes chemistry and physics: a surface treated from the air and then treated again from the ground has a different residual profile than either alone, and the difference is not always in the direction of cleaner. Third, the two subsystems almost never share a measurement plan, because they were procured separately, tested separately, and accepted separately.
The result is an overlap band in which a failed reading has three possible owners and no assigned one. Every party can produce a test report showing its own subsystem met specification. All the reports are true. The surface is still out of criterion.
Three questions an interface contract has to answer
The remedy is unglamorous and cheap if done early: an interface contract, agreed before the first combined trial, that answers three questions in writing.
1. Which test governs acceptance of the overlap?
Not which subsystem is better — which single test, run by whom, using which criterion, decides whether the band passes. A joint test that neither party owns will not be run. Name an owner, and give that owner the authority to stop the trial.
2. What counts as a defect versus a residual?
A defect is a failure of a subsystem against its own specification. A residual is contamination remaining after both subsystems performed to specification. These require different remedies — rework at supplier cost versus an additional pass at programme cost — and if the contract does not distinguish them, every argument becomes a negotiation.
3. Who pays for the second pass, and how many are budgeted?
Programmes routinely budget one pass per subsystem and then discover that the overlap needs a third. Where that third pass is unfunded, the pressure lands on the acceptance criterion, and the criterion moves. Budgeting the pass protects the criterion.
Sequencing is a specification, not a preference
Air-then-ground and ground-then-air are different systems. The order changes wetting, penetration, and the surface state the second pass encounters, and it changes which subsystem is exposed to the other subsystem failure. Programmes tend to leave sequencing to the field, because it feels like a tactical decision. It is not: it is the variable that determines what the acceptance test is testing. Fix it in the specification, and record any field deviation as a deviation, so the test result remains interpretable.
The measurement plan is the real interface
Two subsystems can share a boundary and still not share a language. Where the aerial subsystem reports coverage rate and the ground subsystem reports surface readings, no arithmetic converts one into the other, and the programme has no way to say whether the combined effect met a single criterion. The interface contract should therefore specify one measurement plan for the combined effect, with sampling points that deliberately straddle the band, chosen before the trial and recorded. If sampling points are chosen after the pass, the trial has measured convenience.
Liability follows the record, not the intent
Where an incident is later reviewed, the reviewer will not have access to the good intentions of either supplier. They will have the specification, the test plan, the readings, and the closure artefact. A programme that can produce a written interface contract and a joint measurement plan is in a materially different position from one that can produce two separate, individually excellent, mutually silent test reports.
This is the same discipline that governs any multi-supplier system, and CBRN adds only one feature: the consequence of an unattributed residual is a person, not a warranty claim. That raises the stakes without changing the method.
An audit you can run on an existing programme
- Is there a written interface contract, or only two subsystem specifications?
- Does one named party own the combined acceptance test?
- Is sequencing specified, and are deviations recorded?
- Is there a single measurement plan whose sampling points cross the boundary?
- Is a contingency pass funded, and by whom?
Boundary
This column addresses contractual and test-design structure. It does not describe any specific national programme, does not recommend chemical agents or application parameters, and is not legal advice; liability allocation is jurisdiction- and contract-specific and should be confirmed with counsel. References to airborne dry decontamination concepts stay at role level; system design detail is outside the scope of this site.
Source frame. The role, test and liability triad used here is the question structure from Drawing the Line: Roles, Tests and Liability in a Combined Air-Ground Decontamination Programme (Kindle, ASIN B0HJ6XDNVC). The overlap-band argument, the audit and the examples were written for this column. View on Amazon
Verification. The structural claims about multi-supplier acceptance are author analysis grounded in general systems-acceptance practice (L3), not measurements from a fielded combined programme. Falsification. A combined programme that funds a contingency pass and runs one joint measurement plan across the boundary would not exhibit the failure described here.
Leave a Reply