Where a paper machine loses its year
Mills are among the best-instrumented plants in industry. The problem is almost never missing data — it is that the systems holding it never talk to each other.
Breaks get argued about after the fact
Wet end, press section and dryer causes blur together in the log. Without the seconds before the break laid out across every relevant tag, the cause becomes a matter of seniority rather than evidence.
Giveaway is a permanent, invisible cost
Running at 78.4 gsm to protect a 77 gsm spec is fibre you bought and gave away. Over a year it is measured in thousands of tonnes, and it never appears as a loss on any report.
Specific energy varies by shift with no accountability
Steam per tonne and kWh per tonne move between crews and grades. If nobody sees it inside the shift, nobody owns it.
Chemical dosing is set by habit
Retention aid, sizing, starch and dye rates get set to a number that worked once. Overdosing is normal, expensive and rarely challenged.
Grade changes take longer than they need to
Every crew has its own route from grade A to grade B, and the fastest one is not written down anywhere.
Reel quality is never correlated with the process
DCS, QCS, lab and energy data sit in four systems. So the question of what the machine was doing when a bad reel was made stays unanswerable.
What we connect to in your plant
Your DCS and QCS already hold most of this. We read across them, plus the drives, utilities and lab, and put them on a single timeline.
Utilities and cogeneration
Business and lab systems
The signals we read, and what each one tells you
On a paper machine the useful unit is not the tag, it is the correlation — what the wet end was doing when the dryer section lost the sheet.
| Signal | What it tells you |
|---|---|
| Machine speed with break events tagged by section | Cause attribution based on the seconds before the break, across every section, rather than on the shift log. |
| GSM, moisture, ash and caliper from the QCS scan | Giveaway quantified in tonnes and currency, not just as a control chart that stays inside spec. |
| Refiner load and specific energy | What fibre development is costing per tonne, and whether it is buying you any strength. |
| Stock consistency and freeness | The upstream conditions that decide runnability two hours later. |
| Headbox pressure, vacuum levels and press loads | Wet-end stability, which is where most break sequences actually begin. |
| Steam flow and pressure by dryer section | Drying cost and profile, and the single biggest lever on cost per tonne after fibre. |
| Chemical dosing rates against production | Consumption against standard by grade, which is where habitual overdosing shows up. |
| Broke and winder waste | Yield loss traced to the reel and the conditions that produced it. |
| kWh and water per tonne | Utilities cost per tonne by grade, shift and season. |
The mill's numbers, on the same timeline
Production, quality, energy and maintenance stop reconciling four reports and start reading one.
Machine efficiency
%, by grade
Tonnes per day vs plan
tonnes
Breaks per day and MTBB
count and hours
Recovery time per break
minutes
GSM giveaway
% above target
Moisture variability
2σ
Specific energy
kWh per tonne
Steam ratio
tonnes steam per tonne paper
Water consumption
m³ per tonne
Chemical cost per tonne
currency
Grade change time
minutes and off-spec tonnes
Reel quality yield
% saleable
AI trained on your process. Not a generic factory.
Models built on continuous web processes, where the useful warning arrives in minutes and the cost of a false alarm is a slowdown, not a shutdown.
Break prediction
Learns the minutes-before signature across tension, consistency, vacuum and dryer profile, and warns the desk with enough lead time to slow down or intervene rather than lose the sheet and thirty minutes of production.
Giveaway optimiser
Shows how close to the lower spec limit the machine has genuinely run before, by grade and by crew, and holds the target there instead of at whatever margin feels comfortable.
Specific-energy benchmarking
Ranks shifts, grades and seasons on kWh and steam per tonne, then points at the setpoint combinations behind the best runs. On most machines this is the fastest payback available.
Grade-change playbook
Reconstructs the fastest historical path from grade A to grade B and gives the desk the sequence that produced the least off-spec, instead of leaving it to whoever is on shift.
Your first thirty days
One machine, read-only, with twelve months of history rebuilt in the first fortnight.
Connect one machine
DCS, QCS and sectional drives on a single paper machine, read-only. No changes to control, no interruption to production
2
Rebuild twelve months of history
Every break and every reel from the last year placed on one timeline. Break causes get ranked by evidence, often for the first time.
3
Add steam, power and water
Utilities join the timeline so a tonne can be costed properly by grade, shift and season.
4
Run live on one grade
Break watch and giveaway targeting go live for a single grade. The desk gets a number to hold and a warning worth acting on.
What plants aim for in the first quarter
10–25%
fewer sheet breaks on the connected machine
0.5–1.5%
GSM giveaway recovered against target
4–10%
reduction in specific energy per tonne
20–40%
faster grade changes with less off-spec
Target ranges for scoping a pilot. Replace with your own validated customer results before this page goes live.
