Daily management
Shift board with safety, quality, delivery, inventory and productivity rows (SQDIP)
Shift board with safety, quality, delivery, inventory and productivity rows (SQDIP) is licensed CC BY 4.0. Attribution: Katafacts (katafacts.com).
Customise with AISkill file ↓1 · What it is
What it is
The performance grid pointed at a production area, using the five rows the letter set is named for: safety, quality, delivery, inventory and productivity. One square per shift, or per day where the line runs one shift. Green met the rule, red missed it, blue means the number is not in yet, and there is no amber because a third colour is where an uncomfortable miss goes to be called "nearly fine". It is the same method and the same engine as the performance grid — what changes is that the rows are fixed, and that fixing them buys something real on a site with many cells: anyone can walk up to any board and read it without translation, and a plant manager walking the floor sees the same five rows in the same order everywhere. That is the honest argument for the letter set, and it is a good one where the five letters each map to something the cell can actually move inside a shift. Where they do not — an inside sales team has no meaningful inventory row — forcing them teaches people the board is a form to fill in, and the plain performance grid with that team's own critical few is the better tool. The shift cadence is not a detail. On a line running two shifts, marking once a day averages the two together, and a problem that belongs to one shift disappears into a green. The worked example below is exactly that case: thirteen reds, every one of them on a night shift, which is invisible at any cadence longer than the shift itself.
2 · When to use it
When to use it — and when not to
Use it when
- A production cell, line or area with repeating work and numbers that land by the end of the shift.
- A site with several areas that wants boards anyone can read without relearning them each time.
- All five rows map to something this area can move within a shift or two — not to something decided elsewhere.
- You run more than one shift and suspect the shifts are not the same, which a daily average will never show you.
Not when
- Where a letter has no real meaning for this area. An invented inventory row is worse than four honest ones.
- Office, service and knowledge-work teams as a default — use the performance grid with that team's own critical few. The shape is identical; only the rows differ.
- As an input to appraising a shift or the people on it. The moment a square becomes a story about a crew, the reporting goes quietly green and the board stops being worth reading.
- Where the numbers only arrive the following week. A measure that cannot be marked at handover can only ever be blue.
- For the annual scorecard — a phased plan against target across a year is the performance board, reviewed monthly.
3 · How to fill it in
How to fill it in
- The measures, their rules and owners
- One row per letter, in the same order on every board on the site. Safety is almost always a judgement rather than a number — any injury, any near miss, any guard or interlock bypassed, and the square is red. Quality, delivery, inventory and productivity take a standing number the area can actually move: first pass yield, schedule attainment, work in progress at handover, good units per labour hour. Name the source precisely enough that two shifts cannot read it differently, and give every row an owner who is a person on the floor rather than a department.
- Cadence — the period a square covers
- One square per shift wherever more than one shift runs. This is the single most important choice on the board: marking once a day averages the shifts together, and the difference between them is very often the finding. A cell running one shift can mark once a day. Anything longer than a day and this stops being daily management and becomes reporting.
- The squares, and the reason behind every red
- Mark at handover, by the shift that worked it, in the two minutes before they go home. A number colours its own square against the standing goal; safety is marked by judgement. Every red gets one line of fact — what happened, never who — written by the people who were there. That line cannot be reconstructed the next morning, and it is worth more than the colour: when one changeover puts five rows red at once, only the reasons show that it was one problem rather than five.
- What is being done, by whom, by when
- Containment first, that same shift if the next shift is at risk — a setter held over, a jam-clearing procedure re-walked, a relief operator called. Then the permanent countermeasure once the cause is understood, and keep the two apart so overtime cover is never mistaken for a solved problem. Cap what is open at once. Containment that relies on a person working extra hours is a debt, and the date on the permanent fix is when you stop paying it.
- Colours, discipline checks and the reason Pareto (computed)
- Computed for you: the colours, reds with no reason, reds with nothing being done about their cause, and the Pareto of reasons. A red whose cause is already being worked counts as covered even when no countermeasure names that exact square — one bad changeover throws several reds, and a countermeasure attacks the cause rather than each square it produced. Nothing to enter here.
- What the grid says
- Drafted from the board actually marked, and grounded only in what it computes.
4 · What good looks like
What good looks like
The example below is an assembly cell on two shifts, marked per shift for a week. Fifty squares, thirteen red — and every single red is on a night shift. Not one day shift missed anything. Averaged across a day most of those misses vanish into a green, and the cell looks like it has a quality problem and a delivery problem and an inventory problem; read shift by shift it has one problem wearing four measures' clothes. Ten of the thirteen reds trace to a changeover run without a trained setter, which is why one Tuesday night shows red on all five rows at once, including a bypassed interlock. Note what the repeat-miss prompt found: nothing. Reds alternate with green day shifts, so no row ever shows two in a row — the reasons found the pattern and the streak rule could not. One red has no reason recorded, and it is also the only one with nothing being done about it. Those are the same fact.
Same example, as a downloadable xlsx workbook.
Performance grid · Cell 4, assembly, Plant 2
Cell 4 assembly — per-shift grid
Marcus Adeyemi, Cell Leader · 2026-09-17 · one square per shift
Team: Ana Ruiz, Day shift, Tomasz Nowak, Night shift, Freda Osei, Quality, Lionel Barrow, Maintenance
| Measure | Goal | Mon D | Mon N | Tue D | Tue N | Wed D | Wed N | Thu D | Thu N | Fri D | Fri N |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Safe shiftJudgement · Safety | Judgement | Met | Met | Met | Miss | Met | Met | Met | Met | Met | — |
| First pass yield≥ 98% · Quality | ≥ 98% | 99% | 96% | 99% | 95% | 98% | 97% | 99% | 99% | 98% | — |
| Schedule attainment≥ 95% · Delivery | ≥ 95% | 97% | 91% | 96% | 88% | 98% | 99% | 96% | 93% | 97% | — |
| Work in progress at shift end≤ 120 · Inventory | ≤ 120 | 110 | 138 | 115 | 142 | 118 | 120 | 112 | 131 | 109 | — |
| Units per labour hour≥ 42 · Productivity | ≥ 42 | 44 | 39 | 43 | 38 | 44 | 42 | 43 | 40 | 45 | — |
Red — missed the ruleGreen — met itBlue — not measured yet
The measures, their rules and owners
Safe shift
Red if there was any injury or near miss, or if a guard or interlock was bypassed for any reason.
Safety · Source: Shift leader, at handover · Marcus Adeyemi
First pass yield
Red if fewer than 98% of units passed end-of-line test first time.
Quality · Source: End-of-line test station, shift report · Freda Osei
Schedule attainment
Red if the shift finished fewer than 95% of the units on its schedule.
Delivery · Source: Production system, shift close · Ana Ruiz
Work in progress at shift end
Red if more than 120 units were left between stations at the end of the shift.
Inventory · Source: Physical count at handover · Tomasz Nowak
Units per labour hour
Red if the shift produced fewer than 42 good units per labour hour worked.
Productivity · Source: Production system, good units divided by hours booked · Marcus Adeyemi
Cadence — the period a square covers
One square per shift
The period is whatever the work’s own rhythm is. Most commercial, service and knowledge-work teams run this weekly; a line running shifts runs it per shift.
The squares, and the reason behind every red
Safe shift — Tue N
Interlock bypassed to clear a jam during the 02:00 changeover
Code: Changeover without a trained setter
First pass yield — Mon N
Changeover at 02:00 run without a trained setter; first 40 units out of tolerance
Code: Changeover without a trained setter
First pass yield — Tue N
Same changeover pattern — setter not on shift, machine dialled in by trial and error
Code: Changeover without a trained setter
First pass yield — Wed N
Second changeover of the week on nights, same scrap at the start of the run
Code: Changeover without a trained setter
Schedule attainment — Mon N
Changeover took 70 minutes against a 25 minute standard
Code: Changeover without a trained setter
Schedule attainment — Tue N
Changeover plus the jam that followed it
Code: Changeover without a trained setter
Schedule attainment — Thu N
Two operators off sick, no relief called in
Code: No relief operator on nights
Work in progress at shift end — Mon N
Line stopped for the changeover, work stacked up behind station 3
Code: Changeover without a trained setter
Work in progress at shift end — Tue N
Same — long changeover, buffer never cleared before handover
Code: Changeover without a trained setter
Work in progress at shift end — Thu N
No reason recorded — nobody can learn anything from this square now.
Units per labour hour — Mon N
Seventy minutes of the shift lost to the changeover
Code: Changeover without a trained setter
Units per labour hour — Tue N
Changeover and the jam after it
Code: Changeover without a trained setter
Units per labour hour — Thu N
Ran two operators short all shift
Code: No relief operator on nights
What is being done, by whom, by when
Day-shift setter stays on for the 22:00 changeover, paid as overtime, until night-shift setters are certified.
First pass yield — Mon N · Containment · Marcus Adeyemi · Due 2026-09-15 · Closed 2026-09-15
Should do: No changeover-related red on nights while this is in place — first pass yield back above 98% and schedule attainment above 95%.
Evidence: Covered the Wednesday and Thursday night changeovers. Thursday night hit 99% first pass yield; the Wednesday red was the run that started before cover began.
Train and certify two night-shift setters against the changeover standard work, with sign-off by the day-shift setter.
First pass yield — Tue N · Permanent · Lionel Barrow · Due 2026-10-10
Should do: Nights run their own changeovers inside the 25-minute standard, and the changeover cause disappears from the grid for a full month.
Jam-clearing procedure reviewed with both shifts; the interlock override key moved to the shift leader and signed in and out.
Safe shift — Tue N · Containment · Freda Osei · Due 2026-09-16 · Closed 2026-09-16
Should do: No bypassed guard or interlock recorded on any shift from here on.
Evidence: Walked both shifts through it at handover; key now signed out twice, both times by the shift leader.
Named relief operator added to the night rota, with a second name when the first is on leave.
Schedule attainment — Thu N · Permanent · Ana Ruiz · Due 2026-09-26
Should do: No shift runs short-handed again; schedule attainment on nights stays above 95% through the next absence.
Colours, discipline checks and the reason Pareto (computed)
Red
13 / 50
Reds with no reason
1
Reds with no action
1
Open
2
Past its date
0
Which causes keep coming back
- Changeover without a trained setter10 · 76.9%
- No relief operator on nights2 · 15.4%
- Not recorded1 · 7.7%
This is what earns the next problem-solving cycle — not whichever red was loudest or most recent.
What the grid says
Fifty squares, thirteen red — and every single red is on a night shift. Not one day shift missed anything all week. That is the whole case for marking a square per shift rather than per day: averaged across a day, most of these misses disappear into a green, and the cell looks like it has a quality problem and a delivery problem and an inventory problem. Read shift by shift, it has one problem wearing four measures' clothes. The Pareto says so plainly — ten of the thirteen reds trace to a changeover run without a trained setter, which is why the same Tuesday night shows red on safety, quality, delivery, inventory and productivity at once. An interlock got bypassed clearing a jam during that changeover, and a safety red is never a statistic: it is the one row where the rule is a judgement and the answer is always to go and look. Two more reds are an unrelieved short-handed Thursday night. Worth noticing what the repeat-miss prompt found here: nothing. Reds alternate with green day shifts, so no measure ever shows two in a row, and a team watching only for consecutive misses would have seen five separate rows misbehaving and no pattern at all. The reasons found it; the streak rule could not. Containment is in place and held — the day setter covered Wednesday and Thursday nights, and Thursday came back to 99% first pass yield — but containment is a person working overtime, so the permanent fix is certifying night-shift setters, and that is open until the tenth. One red has no reason recorded at all: the Thursday night inventory miss. It is also the only red with nothing being done about it, and those are the same fact — nobody can work a cause nobody wrote down. Friday night is blue across all five rows because the shift is still running, which is not the same as a good shift.
5 · Common mistakes
Common mistakes
Marking once a day when the line runs two shifts
It averages the shifts together and hides the commonest finding on a multi-shift line: that the shifts are not the same. A problem belonging to nights disappears into a green day, and the board quietly stops being able to see the thing most worth seeing.
Forcing a letter that does not fit
An inventory row on a team with no inventory teaches everyone that the board is a form to be completed rather than a tool to be used, and that lesson spreads to the rows that do matter. Four honest rows beat five with a passenger.
Treating the safety row as a statistic
It is the one row where the rule is a judgement and the right response to a red is always to go and look, that shift. A safety square that gets discussed at the end of the week alongside productivity has already been mis-framed.
Calling overtime cover a fix
Holding a setter over is good containment and a person paying for it with their evening. Log it as containment with a date on the permanent countermeasure, or the cell runs on goodwill until the goodwill runs out.
The board becoming a report to the plant manager
Once its audience is upward the numbers get managed instead of the work, and the shift that marks honestly looks worse than the shift that does not. The crew marks its own squares for its own use; anyone above reads the same board rather than asking for a version of it.
Five reds from one cause read as five problems
A bad changeover can put every row red in the same shift. Counting colours makes that look like a cell falling apart; reading reasons shows one problem with one countermeasure. This is exactly why the reason line is worth more than the square.
6 · What it connects to
What it connects to
upstream
The same method with the rows left open. Use that one wherever the five letters do not each map to something the team can move — which is most office, service and knowledge-work teams.
downstream
The handover conversation this board exists for. On a shift pattern the huddle is where the outgoing crew tells the incoming one what went red and what has been done about it.
Where a repeat cause goes to be removed, and where you find out whether overtime containment ever turned into a permanent fix.
What travels with the board between crews. A red marked at handover and not explained to the incoming shift is a red that gets re-discovered rather than worked.
Which reason keeps repeating across a longer window than one week of shifts, and across several cells at once.
The cycle a repeating cause has earned — a changeover that needs trained setters and revised standard work is not something a shift can contain its way out of.
7 · Where AI helps
Where AI helps
Judgement — stays yours
- Deciding whether all five letters genuinely fit this area, or whether four honest rows would be better
- Setting a goal a shift can actually meet with the equipment and people it has
- Judging whether a red is worth stopping for now or watching another shift
- Deciding when containment paid for in overtime has gone on long enough to stop being acceptable
Analysis — AI helps
- Reading the board by shift rather than by day, and saying plainly when one shift carries the reds
- Spotting that one cause has put several rows red in the same shift, so it is one problem and not five
- Turning handover notes into consistent reason codes, so a cause stops fragmenting across shifts and crews
- Saying when there is too little history to claim a shift pattern at all
Drudgery — automated
- Colouring every square from the numbers and the standing goals
- Counting reds with no reason and reds whose cause nobody is working
- Building and ranking the Pareto of reasons with its cumulative share
- Exporting the board to xlsx in the house format
9 · Rate this kata
