WIP and Plan vs Actual — Work in Progress, Bottlenecks, Overtime and the Weekly Plan vs Actual

1. What the window is for

It answers three questions in one place: how much work in progress (WIP) is waiting between operations and in front of which machine work is piling up; is capacity enough this week, is overtime or an extra shift needed; how much of last week's plan finished on time and why late orders were late. Everything is derived from existing data: MES start/finish events, operation start and completion dates, confirmations, the APS schedule and machine calendar, Stuck Work holds and maintenance breakdowns.

Tabs: WIP and bottleneck · Overtime and extra shifts · Weekly plan vs actual · Delay reasons. The window is in the Production group of the dock.

2. WIP: work in progress waiting between operations

Quantity done: for each operation, good + rework quantity in MES finish events; without MES a completed operation counts as the order quantity and the last operation is combined with confirmations. An earlier operation can never appear to have done less than a later one.
Waiting = quantity done by an operation − quantity done by the next. WIP quantity = done by the first operation − done by the last.
WIP value = waiting × unit cost × completion ratio. Unit cost comes first from the standard cost (nearest period); completion ratio = material share + labour/overhead share × share of routing time up to that operation. Without a standard the actual cost is used as a full unit and the row is marked with * (upper bound).
Wait time: time from one operation finishing to the next one starting; if the next has not started the wait is still running (…). Colour: under 1 day green, up to 3 days yellow, above red.

Clicking an order row opens the operation chain: what each operation has done and, on the arrow between two operations, the waiting quantity and wait time. Open in Production shows the order in the Production window.

3. Queue in front of the bottleneck

The work in progress piled up in front of an operation (what the previous operation made and this operation has not processed yet, partial lots included) is that machine's queue; even if the operation has started, the remaining parts are in the queue. The first operation's input is raw material, so it is not counted as queue. Per machine you see the number of queued orders, queue hours (operation time × quantity piled up in front; + setup if the operation has not started), the oldest wait and the queue days (queue minutes ÷ the machine's daily capacity this week, from the APS calendar). The daily capacity of work center boxes is the APS work center capacity (efficiency included). The queue in front of a machine with unknown capacity (an undefined machine) is red because it can never be processed. Queue days under 1 are green, up to 3 yellow, 3 or more red — the worst machine is shown as the bottleneck. Open in APS goes to the Bottleneck tab of the Production Schedule, where the "Add 1 shift" scenario calculates the effect.

4. Little's law cross-check

From orders completed in the last 28 days the throughput (λ, orders/day) and the average flow time (W, days from the first operation start — or the order opening date — to the last confirmation) are calculated; L = λ × W is the expected number of orders in the system. If the number of orders started but not finished right now deviates from this estimate by more than 25%, a "flow is not balanced" warning appears: above means WIP is piling up, below means the line is emptying. The check uses order counts because quantities of different products cannot be added.

The verdict depends on the absolute difference. If the difference between observed and estimated is less than 3 work orders, the flow counts as balanced. If it is 3 or more: when the estimate is at least 1 order the percentage deviation must also exceed 25%; when the estimate is under 1 order (a workshop with few orders, or data where confirmations were entered in bulk) the percentage is meaningless and is not shown, but a build-up is still reported with the "not balanced" warning.

5. Overtime and extra shift need

The grid shows work center × week load (a machine without a work center has its own row). Capacity comes from the APS machine calendar (shifts, breaks, planned maintenance, downtime, day overrides). Load comes from open operations in the official APS schedule; the remaining time of open operations not in the schedule is added to the week of the operation plan, order plan or due date (past dates go to this week). Cells are green up to 85%, yellow up to 100%, red above.

Overload = load − capacity (member machines of a work center share each other's load).
OT capacity = crew × weekly overtime cap per person; the cap is read from the active HR regulation (daily × working days, weekly cap, the remaining yearly allowance spread over the remaining weeks — Türkiye 3 h/day and 270 h/year, Indonesia 4 h/day and 18 h/week). Overtime already used this year is read from timekeeping. It is also limited by the machine's free calendar time.
Extra shifts = ⌈overload ÷ shift length⌉ (shift length from the calendar).
Suggestion: Overtime is enough · OT not enough: extra shift · OT and extra shifts not enough: subcontract or move load · No capacity: move the load to another machine (machine under maintenance or inactive).

The current week counts from today. This week's capacity and overtime allowance are calculated only from the days between today and the end of the week; capacity of days already past can no longer be used, while overdue load is written into this week. If a row has no capacity at all in the remaining days of the week (for example at the weekend), the load falling into that week is moved to the first week with capacity; if no row has remaining capacity, the Boss Dashboard and the weekly page show the overtime need for next week. The "No capacity" suggestion stays only on a row that has no capacity anywhere in the horizon. A work center row's capacity is the APS work center capacity (efficiency included). The Company reports › Machine utilization and OEE button above the grid opens the actual machine utilization for the same weeks.

6. Weekly plan vs actual page

Planned finish, in order: the APS operation plan → the official APS schedule / capacity plan → the order's planned start + planned duration → the due date. The bottom of the page says how many orders came from each source.
Planned: orders whose planned finish falls in this week (count, quantity, total operation time). Actual: orders completed in the week, confirmed quantity and working hours in the week (MES net run time + confirmation time not linked to MES).
On time % = completed on or before the planned finish ÷ evaluated (orders whose planned day has not come yet count as "waiting").
Late: orders planned for this week that finished late or are still open + carried over open orders from earlier weeks; delay days are counted from the planned finish.
The delay reason Pareto, this week's overtime / extra shift need, the WIP summary and next week's risks (Stuck Work due-date radar) are on the same page.

Print opens the page in A4 layout; the PDF is produced on the server (print from the browser if the message service is off). Numbers are written in the format of the document's language. Save snapshot freezes the week's figures — saving the same week again updates the same record, no new record is created; the version goes up only if the figures changed. Send: an in-app message to the chosen users, an e-mail with the PDF attached to the e-mail addresses, a short summary to the WhatsApp numbers.

Drilling into the company report. The Company reports › Planned and actual production button at the top of the page opens, in the Reports window, the planned / actual comparison of the same week by work order and operation (permission: general reports).

Past weeks. On the page of a past week the WIP and overtime figures are not calculated from today; they are shown only from a snapshot taken during that week (the weekly job takes the new week's snapshot every Monday). If there is no such snapshot these two boxes stay empty and a note appears on the page.

7. Delay reason: one catalogue and priority

The catalogue has eight system reasons: Material shortage, Machine breakdown, Staff shortage, Quality problem, Plan change, Supplier delay, Customer change, Other; new ones can be added, system reasons are never deleted (they can be deactivated, except Other). Source code mapping links the Stuck Work hold reason, the MES downtime reason (or its category) and the maintenance failure code (or its category) to this catalogue; an unmapped code falls to Other.

The reason for a late order is found in this order: correction (a reason overridden manually with plan_gercek_yonet) > manual assignment (an assigned reason comes before all automatic sources) > Stuck Work hold (with a known reason) > MES downtime (a downtime event linked to the order) > maintenance breakdown (a breakdown work order on the order's machine overlapping the order's time) > "unknown" hold (a hold the Stuck Work job opens on its own for late orders, carrying no reason — used only when there is no other evidence, falls to Other). Within a source the longest record wins. If none exists the row shows "No reason assigned" in yellow; choose one with Assign reason. Each assignment deactivates the previous one (nothing is deleted) and stays in the log.

8. On the phone

On a small screen the window switches to the Garden layout: Summary (the week's KPI cards, busiest work center, reason Pareto), Late (card list; tapping one opens a three-step wizard: 1 Order · 2 Reason · 3 Result) and WIP (machine queue cards). Touch targets are at least 44 pixels.

9. Boss Dashboard, Pulse and the weekly job

The Boss Dashboard's attention box shows money tied up in WIP and this week's overtime / extra shift need (red if overtime and extra shifts are not enough; at the weekend for next week).
In the Pulse mobile interface a work center over capacity this week appears as an attention row.
The weekly job is off by default (PLAN_GERCEK_GOREV=1). When on, on Monday morning (PLAN_GERCEK_GUN, PLAN_GERCEK_SAAT) it saves the snapshot of last week and of the new week — at most once a day per week, and the version goes up only when the figures change; the new week's record keeps that week's WIP and overtime figures as measured during the week. Once a week it writes an in-app summary message to managers and users with plan_gercek_yonet or production planning permission. E-mail and WhatsApp are not sent by the job; use Send on the page.

10. Permissions

View: plan_gercek, plan_gercek_yonet or one of the production permissions.
Assign a reason to a late order: plan_gercek (or production planning).
Correct the automatic reason, catalogue, mapping, saving and sending snapshots: plan_gercek_yonet.
Run the job manually: system administrator. Writing also requires the write gate to be open and the user not to be read-only.