Leading the Intelligent Factory:  Chapter 11

The first 90 days, and making the case

Data you already have, a named owner, a daily review. Then savings potential, what it takes, and what waiting costs.

Spread 1

What it takes to start

Data you already have. Your historian already stores what this work needs. You do not need new sensors, a new control system or a data science team to begin.

A named owner. One engineer or superintendent for each area receives the findings and answers each one: fixed, planned, or not a problem. A finding without a name on it does not get done.

A daily review. Fifteen minutes in the morning meeting. What changed overnight, what is on the list, what moved off it. The habit matters more than the tool.

The right order. See first, then set, then study. That is reliability, then quality, then cost, the same order as this book.

Day 0Day 30Day 60Day 90RELIABILITY FIRSTSeeDays 1 to 30Connect the historianyou already haveScan every tagevery nightA daily short list, witha name on each findingTHEN QUALITYSetDays 31 to 60Targets and limits fromyour own run historyFirst compliance scoreTop tags: run to target,or reset the targetTHEN COSTStudyDays 61 to 90First cost orvariation studySavings potential, withwhat it takes to get itFirst targets movedEvery day: a 15-minute review. One named owner for each area.
Figure 11.1. A first 90 days. Each step builds on the one before it.

Spread 2

Making the case

Your leadership will ask three things: what it is worth, what it takes, and what happens if you wait.

What it is worth. Five studies found $141,000 to $270,000 a year each in savings potential. The paper machine had two studies, worth about $424,000 a year together. In our experience, back-end areas such as recovery, recaust and bleaching typically show $100,000 to $200,000 a year, and paper machines $300,000 a year and more.

What it takes. Every figure should come with two more lines: what it takes to get it, and what would take it to zero. Changes you have run before come first. Trials on new ground are listed apart.

What waiting costs. The lime kiln in Chapter 8 ran outside its best window for eight months. At about $141,000 a year, that is roughly $90,000, and nothing broke to announce it.

The test for the first 90 days

Did the first findings cover the first year?

Start in the area where that is most likely.

Savings potential found, per yearLime kilnHold stack O2 in a 7.0 to 7.2% window$141,000Paper machine, variationRemove one 13 to 14 minute cycle$154,000Recovery boilerRatio air to firing, lower blowdown$233,000Bleach plant and washingCap one chemical ratio, lower one dose$246,000Paper machine, costA refining trial and a moisture target trial$270,000Under each study: the main thing it takes. Each figure also states what would take it to zero.
Figure 11.2. Five studies in pulp and paper. Several figures use placeholder prices where the mill had not shared its own.

Questions for Monday morning

  1. Who would own the morning list in each of our areas?
  2. What data do we already store and never look at?
  3. Which of our problems would pay back fastest if it were solved?
  4. What does one month of running off target cost us?
  5. What would we need to see in 90 days to keep going?

In practice. Envoy connects to the historian you already have. Our process engineers review each finding before it reaches yours.