Every plant we work with has OEE data. Almost none of them trust it enough to make decisions from it. Not because the number is wrong — but because the number doesn’t tell you enough.

The Monday meeting problem

Someone pulls up the OEE numbers. “Line 3 ran 72% last week.” Everyone accepts the number. That’s not the problem. The problem starts thirty seconds later, when someone asks: “Why?”

And then the room fractures:

• Production says it’s maintenance — the labeler keeps faulting, PM intervals are too wide.
• Maintenance says it’s operators — the crew isn’t following centerline settings, the new operator hasn’t been trained.
• Scheduling says it was a brutal product mix — 14 changeovers in five days, short runs that never let the line stabilize.

The plant manager just wants someone to tell him which of these is actually true and what to fix first. Nobody can.

The distrust in manufacturing isn’t about the OEE number. It’s about the diagnosis. Everyone knows there’s a problem. What people can’t prove is why the problem exists and who owns the fix.

The cost of unresolved diagnosis:

• Blame without proof. Wrong diagnosis = wrong intervention = wasted money and eroded trust.
• The same problems recur. Without decomposing where variance lives, plants chase symptoms. The underlying driver reasserts itself in three weeks.
• CI resources go to the wrong place. Labeler downtime caused by an untrained operator is a completely different problem than labeler downtime caused by a worn cam follower. Same Pareto bar. Different root cause. Different fix. Different owner.
• People stop trusting the system. When the data can’t resolve the argument, people fall back on politics and seniority. The MES becomes a compliance artifact.

What TRI is

TRI is a governed signal for four things:

TRI is not an operator scorecard. Not a dashboard. Not a general-purpose analytics platform. It is a decision-routing system.

The three dimensions

TRI extends OEE with the same familiar three-factor structure:

If any factor collapses, TRI collapses. A line can’t compensate for terrible reliability with strong throughput.

Four operating states

When you combine throughput, reliability, and direction, every line falls into one of four operating states:

Same OEE, different reality

Take two lines, both reporting 62% OEE:

Same OEE. Completely different operational realities. One is a scheduling anchor. The other is a planning liability generating overtime, expediting costs, and missed commitments.

In a real four-line assessment, we found two lines with nearly identical OEE (~31%) that were 27% apart in reliability. Read the full case study →

Where the variance lives

TRI doesn’t just measure instability — it decomposes it. The variance breakdown resolves the Monday meeting argument by showing exactly where the problem lives and who owns the fix:

Same Pareto bar. Different root cause. Different fix. Different owner. This is how TRI turns a 45-minute blame argument into a 15-minute review of what changed, where the variance lives, who owns the fix, and what it’s worth.

Who sees what

TRI speaks differently to each role. Same underlying truth. Different presentation. Different action vocabulary.

Give the supervisor EBITDA-at-risk and they freeze. Give the CFO machine-level decomposition and they micromanage. Governed visibility is a feature.

The money layer

Every TRI alert carries a dollar estimate of the exposure it represents. Seven loss components computed from real plant economics:

• Lost throughput — units you should have made but didn’t
• Excess labor — overtime driven by unreliable output
• Service risk — fill rate exposure from capacity uncertainty
• Emergency maintenance — reactive spend driven by instability
• Quality events — defects correlated with variance patterns
• Safety stock — excess inventory buffering against unreliable lines
• Expediting — premium freight when the schedule breaks

Interventions are ranked by return, not severity. A line at TRI 0.55 with $80K exposure and a $5K fix ranks higher than a line at TRI 0.50 with $15K exposure and a $20K fix. That reordering is how operations, the VP, and the CFO end up caring about the same system.

Where TRI fits

TRI is the decision layer between your data and your operating system:

• OEE measures loss.
• Lean defines waste and improvement routines.
• TPM protects asset reliability.
• Digital twins model expected performance.
• MES records what happened.
• TRI ranks where throughput reliability is breaking and what action should happen first.

TRI does not replace Lean, TPM, digital twins, or OEE. It tells you where to apply them first.

TRI in practice: three facilities

Every plant has a different starting point. Here’s what a Production Intelligence Assessment reveals depending on where you are.

What this isn’t

TRI is not a dashboard. It’s not a software product. It’s not a replacement for OEE.

It’s a framework applied during a Production Intelligence Assessment — using data your MES already collects — that gives your team a more complete operating picture than OEE alone can provide. The output is analysis, classification, financial exposure, and a prioritized action plan. Not a login.

Throughput Reliability Index is proprietary methodology of Crusoe Advisory LLC.