A threshold breach, not a fix
Static control limits tell an operator they've left the spec range. They don't say how far to move a setpoint to get back inside it.
Intelligent Centerlining™
Every process has a historical best version of itself. Centerlining compares live production against that golden run and tells operators exactly how much to adjust, not just that something is off.
The problem with generic alerts
An alert tells you something’s wrong, just not what to do.
Most monitoring systems flag a deviation and stop there. The variable crossed a limit; a light turns red; an operator gets a notification. What's missing is the part that actually changes the outcome: how much to move which control, right now, to bring the process back on target.
Judgment calls under time pressure
Without a quantified benchmark, operators estimate the correction from memory and experience, and that estimate varies shift to shift.
Levers mixed with noise
Some variables that drift are within an operator's control. Others aren't. Generic alerts rarely make that distinction, so effort gets spent in the wrong place.
The golden run concept
Every process has a best version of itself.
Somewhere in your production history is a run where every variable lined up: quality held, chemistry usage was efficient, throughput was strong. That run is the golden run, and it's specific to your equipment, your product, and your process, not an industry benchmark or a vendor's assumption. When a variable drifts from it, centerlining calculates the exact adjustment needed to return, not just a warning that something is off.
Without centerlining
Drift goes unmeasured
The process wanders within the spec range with no reference point for what "good" actually looked like.
- No benchmark. "In spec" is treated as good enough, even when it's well short of your best.
- Inconsistent targets. Each shift centers the process differently, based on personal habit.
- Value left on the table. The process runs safely, but not at the level it's capable of.
With centerlining
Continuously compared to your best
Live production is checked against your own golden run, in real time, for every key process variable.
- A concrete reference. The golden run is defined from your own historical performance, not a generic spec.
- Continuous comparison. Live variables are checked against golden run targets as production runs, not on a fixed reporting cycle.
- A shared standard. Every shift centerlines to the same target, closing the gap between operators.
How it works / What you can control
Separating levers from noise
Not every variable that drifts is something an operator can act on. Centerlining distinguishes what operators can control from what they can't, so recommendations point only at levers that actually move the outcome.
- Filters out variables outside operator control before guidance is issued
- Keeps the recommendation list short enough to act on during a shift
Controllable vs. uncontrollable variables
Built for real production floors
Built to hold up on a real production floor
From your first upload of historical process data to live predictions, pinpointed root causes, and an AI assistant on the floor, without writing a line of code.
01 / Always current
The golden run, always current
Your golden run isn't a benchmark frozen in time. As models retrain with every new lab value and process condition, the golden run target reflects your process today, not a snapshot from last year.
- Targets update as equipment, feedstock, and conditions evolve
- No manual recalibration cycle required
Live vs. golden run
02 / Grade-aware
Built for grade transitions
Quality limits, upper, lower, and target, are defined per grade. As the active grade changes, golden run targets shift with it automatically, using the same grade-transition logic that manages a wide variety of grades/SKUs without manual reconfiguration.
- Golden run targets are grade-specific, not a single facility-wide average
- No manual reconfiguration needed as products change over a shift
Proactive optimization
Proof, not projection
Holding the golden run shows up in the numbers.
Keeping process variables aligned to golden run targets, continuously rather than intermittently, is a direct driver of the quality and chemistry outcomes below.
Case study Paper manufacturing line
>20%
Quality variability reduction
Driven by continuous centerlining to target
~25%
Chemistry reduction
10% conservative floor; 25% proven ceiling
$2.4M
Value per machine, per year
Customer attributed 50% to the platform
~20
Machines in autonomous control
One client, after a ~2-year trust ramp
These figures describe a single, specific ~2-year deployment journey at one client site; they are not a general timeline promise for other sites. Results vary by facility, process, and data quality; past performance is not a guarantee of future results.
Keep your process on target
See your golden run, live.
Watch a short demo of Intelligent Centerlining running against real production data and see exactly what quantified setpoint guidance looks like on the floor.
No data science degree required. Live dashboard in hours, results in less than two weeks.
Questions, answered
Frequently asked questions
What is centerlining in manufacturing?
Centerlining means holding a process on its best-known operating condition, its golden run, instead of letting it drift within a wide spec range. ProcessMiner’s Intelligent Centerlining automates this by continuously comparing live production data to your facility’s own historical golden run and telling operators exactly how much to adjust specific controls to get back on target.
How is this different from a control chart or SPC alert?
A standard SPC alert tells you a variable crossed a static control limit. Centerlining tells you the specific adjustment needed to bring the process back to your own historical golden run, calculated from your real performance data rather than a fixed boundary.
Does the golden run change with product grade?
Yes. Golden run targets are defined per grade, and the platform automatically adjusts which targets apply as the active grade changes, using the same grade-transition logic that manages 5 to 40+ grades without manual reconfiguration.
Can operators override the guidance?
Yes. Operators retain full control. Centerlining recommends an adjustment; it does not act on its own unless a site has deliberately moved to closed-loop autonomous control for that specific variable.
Does centerlining work outside pulp and paper?
Yes. The golden run concept is not specific to paper. It applies to any continuous manufacturing process with quality targets and process variables, including nonwovens and plastics, because centerlining is built from each facility’s own data rather than an industry template.
Get in touch
Have a question?
Whether it's the golden run concept, your process, or a pilot you'd like to see in action, let us know.