Robotics in Manufacturing

Manufacturing economy

Poor interoperability: $15.8 billion a year in capital facilities

Inadequate interoperability costs the US capital facilities industry $15.8 billion a year, and at least $1 billion a year in the automotive supply chain. Both figures come from studies commissioned by NIST.

The mechanism is mundane and familiar to anyone who has received a supplier's CAD file. In a survey of tool and die shops, 51 percent of jobs required the CAD data to be repaired, and in about 15 percent of cases the errors were not found until after the tooling had been cut.

This page traces both figures to their studies and keeps them apart, because they cover different industries and get quoted as though they were one number.

Data covers NIST-commissioned interoperability cost studies (capital facilities and automotive). Published 2026-08-16. Last reviewed 2026-08-16. Last updated 2026-08-16. Edited by Mike Ramsey / Reliable Media.

The figures and where they come from

Each figure is rated for how safely you can cite it today. Ratings judge current usability, not whether a number was ever correct.

FigureWhat it isSourceCitation ConfidenceNotes
$15.8 billion a yearCapital facilities industry, annual cost[A]HighThe cost of inadequate interoperability in the US CAPITAL FACILITIES industry, meaning buildings and infrastructure. This is not a manufacturing figure and is routinely miscited as one.
at least $1 billion a yearAutomotive supply chain, annual cost[B]HighA separate study covering the US automotive supply chain. Fifteen times smaller than the capital facilities figure, and about a different industry.
51 percentJobs where CAD data needed repair[B]MediumFrom a survey of tool and die shops receiving CAD/CAM data from customers, reported in the same automotive study. A concrete picture of what goes wrong in a data handoff, reported alongside the cost estimate rather than presented as its derivation.
25 percentJobs requiring complete CAD recreation[B]MediumBeyond repair, the shop had to recreate the CAD data entirely in a further quarter of cases.
15 percentErrors found only after tooling was cut[B]MediumThe expensive failure mode: the data problem surfaces after the tool has been machined.

Why the numbers disagree

The two figures are fifteen times apart and describe different industries. The $15.8 billion covers the US capital facilities industry, which is buildings and infrastructure. The $1 billion covers the automotive supply chain. Both were commissioned by NIST, which is exactly why they get merged in citation, and merging them produces a number neither study supports.

Both are annual rates, not totals. The studies state them per year, and dropping that turns a recurring cost into a one-off, which understates the case being made.

The survey figures are the part worth carrying, because they are checkable against experience. Half of tooling jobs needing CAD repair, a quarter needing full recreation, and one in seven errors surfacing only after the tool was cut, describe a workflow anyone in the supply chain recognises, in a way a billion-dollar aggregate does not. They appear in the same study as the cost estimate, though the study does not present them as the calculation behind it.

How to cite these figures

Name the industry with the number. Either $15.8 billion a year in capital facilities, or at least $1 billion a year in the automotive supply chain. Never the two combined.

Say 'a year'. Both studies report annual costs, and the time period is load-bearing.

Use the survey percentages when the audience is technical. The 51 percent CAD repair rate lands harder with practitioners than the aggregate does.

Note the vintage. These are established studies rather than current measurements, and no newer verifiable restatement was found.

Where people go wrong

Presenting $15.8 billion as a manufacturing figure. It covers the capital facilities industry, buildings and infrastructure.

Adding the two figures together, or citing one as though it superseded the other. They are separate studies of separate industries.

Dropping the annual framing. Both are costs per year, not cumulative totals.

Quoting the figures as current. They are long-standing studies, and we could not verify a more recent restatement.

How we checked

Both figures come from NIST-commissioned studies, retrieved as PDFs and confirmed to contain each figure quoted, including the survey percentages behind the automotive estimate.

We keep the two studies in separate rows and label the industry on each, because the single most common error with these numbers is treating them as one estimate of one thing.

We carry the survey mechanism alongside the aggregates. A dollar figure with no mechanism is hard to check; a CAD repair rate is something a reader can compare against their own experience.

We looked for updated versions of either study in a verifiable form and did not find one. Both should be read as established analyses of their period rather than current measurements.

Full source list

Primary sources, with live links. Every figure above traces to one of these.

  1. [A]National Institute of Standards and Technology2004

    NIST GCR 04-867, cost analysis of inadequate interoperability in the US capital facilities industry (prepared by RTI International and Logistics Management Institute)

    https://nvlpubs.nist.gov/nistpubs/gcr/2004/nist.gcr.04-867.pdf
  2. [B]RTI International for NISTAccessed August 2026

    RTI International, interoperability cost study of the US automotive supply chain (prepared for NIST)

    https://www.rti.org/sites/default/files/resources/US_Automotive.pdf

Common questions

What does poor interoperability cost?
Two NIST-commissioned studies give different answers for different industries: $15.8 billion a year in the US capital facilities industry, and at least $1 billion a year in the automotive supply chain. They are separate figures and should not be combined.
Is the $15.8 billion a manufacturing number?
No, and this is the most common misuse of it. It covers the capital facilities industry, meaning buildings and infrastructure. The automotive supply-chain figure is the manufacturing one, at about $1 billion a year.
What actually goes wrong?
A survey of tool and die shops found the CAD data had to be repaired in about 51 percent of jobs, had to be recreated entirely in a further 25 percent, and in about 15 percent of cases the errors were not discovered until after the tooling had been cut.
Are these figures current?
They are established studies rather than fresh measurements, and we could not verify a newer restatement of either. Cite them with that caveat rather than as today's cost.

More data, traced to source