White paper · 11 pages · November 2026

Converting contextualized data into measurable throughput

CESMII's i3X 1.0 gives manufacturers a vendor-agnostic open API for contextualized plant information. This paper argues that the next hard problem is not technical. Between a conformant data layer and a production gain sits a decision about which asset to act on, and nothing in the architecture supplies it.

  • The attribution gap. A conformance test proves compliance. A portability test proves an application moves. Neither produces the number a plant's finance function recognizes as a result.
  • Why better data alone does not produce throughput. Contextualized data about a non-constraint is precise, well-modeled information about something that cannot move the plant.
  • What Zero Loss Analysis supplies. A baseline referenced to demonstrated physical capability rather than a negotiated standard, a decomposition tested against historian data, and a conversion into revenue or cost.
  • Why the method needs the data layer. Recomputing the binding constraint each shift as the product mix changes is impractical at scale without something like a common contextualized API.
  • A proposed pilot. Four phases with stated success criteria, including the condition under which the pilot should be judged to have failed and published as such.

The argument in brief

Conformance is not an outcome

The i3X 1.0 release is rigorous about what it can measure. The conformance suite reports Full 1.0 Compliance, 1.0 Compatible, or Not Compliant. Compliance is a property of software. A manufacturer who has completed a successful portability test still has an unanswered question, and it is the only one their CFO will ask: how many more cases did we ship?

A data layer is not self-directing

The performance of a plant is governed by a single binding constraint at any moment, and improvement applied anywhere else produces no plant-level gain. A conformant data layer makes it substantially cheaper to compute OEE consistently at every station. Improving OEE at a station that already has excess capacity produces no additional shipments. A superior data layer therefore makes it cheaper, faster and more consistent to industrialize a measurement error that is already widespread. That is not an argument against the architecture. It is an observation about where it stops, which is exactly where it was designed to stop.

What an attribution layer has to do

Zero Loss Analysis establishes Full Potential, the output an operation would produce if the binding constraint ran at its maximum demonstrated sustained capability with no downtime, no slow running and no waste. Because that reference is observed rather than negotiated, the resulting efficiency figure cannot be improved without real recovery, which makes it suitable as the baseline in a claim a third party may audit. Variance tree analysis then decomposes the gap into ranked causes tested against historian data rather than generated in a workshop.

The dependency runs both ways

Conventional practice treats the bottleneck as a location found once by a study. In discrete operations running multiple products that is wrong often enough to be dangerous, because each product runs at a different rate through each station and the binding station follows the week's mix. Recomputing it each shift requires asset identity, asset-to-process relationships, product identity, per-product rate capability and attributable loss events, continuously, across every station. Assembled from point-to-point integrations that is a bespoke engineering project at every plant. A common contextualized API is the first general route to that data shape.

A note on independence. This paper is an unsolicited proposal prepared independently by Plantworks. It has not been reviewed, sponsored or endorsed by CESMII, and no relationship between the two organizations exists as of publication. All characterizations of CESMII's technology are drawn from its own published materials and are cited in the paper. Plantworks is not affiliated with CESMII, the U.S. Department of Energy, or Manufacturing USA.