Skip to content
Ground Reality

Record · Supply

The site sets the order date.

Book stock knows what was delivered. The record knows what is being installed — at what rate, with what tolerance, measured at the same source the progress claim is measured at. Project that rate against what is on the ground and the reorder date stops being a habit and becomes a measurement. So does the day the pour stops.

INSTALL RATE 18.4 T/DAY ±2.1 · FIT n=14COVER 22.4 D · SAFETY STOCK 104 T · p95
Record header · sheet R-03
Record
SPECIMEN SITE 04
Material shown
REINFORCEMENT STEEL
Install rate
18.4 t/DAY ± 2.1
Stock on the ground
412 t
Days of cover
22.4 d
Lead, measured
13 d MEDIAN · n 27
Halt exposure
3.0 d
First epoch · 07 Oct 202524 epochs on the recordLatest · 21 Apr 2026
COVER · 3 MATERIALS · 6 MEASURED INPUTS

The instrument

Stock, against the pace that consumes it.

One line orders at the point the record computes. The other orders where book stock says to. Same supplier, same order quantity, same lead time — the only difference between them is the day the order goes out, and on two of these three materials that day is the difference between a pour and a stood-down gang.

Stock against measured install rate

──── Ordered at the measured reorder point- - - - Ordered on book stockSafety stock, p95

Measured install rate
18.4 t/day± 2.1 · fit over 14 working days
Days of cover
22.4 d20.1–25.3 d at the rate's tolerance
Safety stock, dynamic
104 tp95 · recomputed as the rate and the lead-time spread move
Reorder point
343 tbook-stock point 184 t · promised lead only
Order in
3.7 dat the measured rate, from today
Lead time
13 d medianpromised 10 d · measured on 27 deliveries
SPECIMEN SITE · ILLUSTRATIVE RECORDBoth policies order the same quantity from the same supplier. The only difference is the day the order goes out, and that day is a measurement.
PROJECTION · +9.4 D · FIT n=12

The forecast

Nine days late. Give or take.

A quantity divided by a measured rate is a duration, and both terms carry tolerances, so the date carries one too. The range is published rather than the single date — a forecast with no ± is an opinion wearing a calendar.

Pour 14 · L07 deck · projection from measured pace▏ PROGRAMME · ◆ PROJECTED · HATCH = TOLERANCE
Remaining quantity
1,240 m²measured at epoch 24, deck area on L07
Measured pace
58 m²/day± 6 · fitted over 12 working days
Working days needed
21.4 d19.4–23.8 d at the pace tolerance
Programme allows
12.0 dfrom the current programme revision
Slip, stated as a measurement
+9.4 d+7.4 to +11.8 d
The fit itself
σ 6 m²/dayleast squares on 12 daily placements, each measured

A forecast is a measurement with a wider tolerance, and it is published with one. The ± here belongs to the pace, not to an opinion about the crew. When a projection crosses a milestone the record prepares a directive — scope, quantity, the epochs it was measured from — and a person with the authority releases it, or does not.

PROJECTION · +9.4 D · −2.0 / +2.5 D · FIT n=12
LEAD TIME · 88 DELIVERIES · 4 SUPPLIERS

Reliability, unrequested

Nobody scored these suppliers.

Every load is weighed and captured at the gate because the mass balance needs it to be. Lead-time distributions fall out of the same evidence — promised against measured, with the number of deliveries on the end of every row.

Lead time · promised against measured▏ PROMISED · ◆ MEDIAN · p10–p90
  • Supplier A

    Reinforcement steel

    13 d medianpromised 10 d · n 27

  • Supplier B

    Cement, OPC 53

    7 d medianpromised 7 d · n 41

  • Supplier C

    Wall formwork panels

    26 d medianpromised 21 d · n 12

  • Supplier D

    Reinforcement steel

    12 d medianpromised 12 d · n 8

0 d10203040 d

A supplier with a longer median and a tight spread is cheaper to plan around than one with a short median and a wide one, because the safety stock is built out of the spread. Both figures are here, and n is the number of deliveries actually weighed — not a rating anybody typed in.

LEAD TIME · 88 DELIVERIES MEASURED · 4 SUPPLIERS
DECISION · CARRY vs HALT · 2.2×

The halted pour

Which mistake would you rather make?

Cover costs money to hold. A stopped pour costs more, or less, depending entirely on the site. The arithmetic is not hard — it just needs two measured quantities that most sites have never had.

The two ways to be wrong · reinforcement steelRATES ARE THE SITE’S OWN FIGURES
Carrying cover nobody needed56,160

104 t × 18 per tonne-day × 30 days

Standing the pour down126,000

3.0 days × 42,000 per day

Reading · 2.2× · at these rates the cover is the cheaper mistake

Change either rate and the reading changes; both belong to the site, not to us. What the record supplies is the pair of measured quantities the arithmetic needs — a dynamic safety stock computed from the install rate and the supplier’s measured spread, and a halt exposure in days, projected before it happens rather than counted afterwards.

DECISION · 2 INPUTS · 2 MEASURED QUANTITIES · ILLUSTRATIVE RATES
BYPRODUCTS · 3 · SAME EVIDENCE

What else the gate produces

Measure once. It pays in three places.

Nothing on this page asks a site to do extra work. It asks the evidence already crossing the gate to be typed, registered and kept.

IDENTITY · BATCH → POSITION

Batch, bound to position

Heat numbers and batch references are read at the gate and carried through on the elements they end up in. Query a batch years later and the answer is a list of positions in the building, not a folder of dockets.

TELEMETRY · CONTINUOUS · EPOCH-AUDITED

The machine between epochs

Epochs are periodic; a machine is not. The machines already know what they moved, and the record takes that too — so the consumed and excavated terms keep moving between captures, and the next epoch checks them rather than being replaced by them.

DIRECTIVE · PREPARED · RELEASED UNDER DoA

The order, prepared not placed

When stock crosses the computed reorder point the record drafts the purchase directive — material, quantity, the rate it was computed from, the epochs behind that rate — and stops there. Financial actions exist only as proposals; the release lever is human.

NOTES · 7 · STATED AS SPECIFICATION

Questions of record

The awkward questions, answered on the sheet.

Prerequisites, cadence and limits belong in the same ink as the claims.

01How is material delivered reconciled against material used?
Bought, delivered and placed are put into one balance per epoch: the load is measured at the gate, the installed quantity is measured on the same capture the claim is measured on, and anything left over is stated as a residual rather than absorbed into an average.
02What if the site has no weighbridge?
The delivered term is measured at the gate from capture and ticket instead, at a wider tolerance, and the record states which of the two methods it used for every load rather than averaging them into one number.
03How long before an install rate is usable?
A fit needs working days, not calendar days. Ten to fourteen give a usable rate for bulk trades. The tolerance is published from the first day and narrows as the fit lengthens — it is never withheld until it looks good.
04Does it place orders?
No. It prepares them. Release is a human act under delegated authority, and the record keeps the proposal, the release or the refusal, and the epochs the quantity came from.
05What about materials with no measurable install rate?
They stay on book stock and are marked as such. A record that quietly mixed measured and reported quantities would be less useful than one that draws the boundary and says which side each figure is on.
06Does the safety stock ever come down?
Yes — that is what dynamic means. A supplier whose spread tightens over twenty deliveries needs less cover behind it, and the record lowers the cover rather than waiting for someone to notice.
07Which figures does a supplier see?
Only what is shared with them explicitly. Access goes down to the part of the site, so a supplier's view is a scope on the record rather than a report somebody emails out of it.
SUPPLY · PREREQUISITES STATED · LIMITS STATED
SUPPLY · MEASURED PACE · DYNAMIC COVER

Order dates should be computed, not remembered.

The install rate is measured at the same source as the claimSafety stock comes from the supplier's own measured spreadHalt exposure is stated in days, before the days happenPROCUREMENT · FROM MEASURED PRODUCTIVITY
CONTACT · SALES · REPLY 1 BUS. DAY

Buy at the rate you build.

Talk to our team about your materials, your gates, and the install rates your sites are already producing without anyone recording them.