Help › Production › Stuck Work Tracking

Stuck Work Tracking — A Guide from Scratch

On the shop floor the most expensive job is the one left half-done and forgotten. The part sits at a machine; nobody knows "where is it, why is it stuck, who is dealing with it"; when the workload rises, small orders get lost, and the problem surfaces only on delivery day, with the customer at the door. The Stuck Work window does four things: it finds stuck work by itself, it makes you record a hold with an owner, a reason and a target resolution date, it escalates what stays unresolved to the supervisor and the administrator, and it shows the orders that will miss their delivery days in advance.

1. What this window is for

The window has five tabs: Stuck (open, resolved and cancelled holds), Due-date radar (orders whose delivery is at risk), Waiting on me (work assigned to you), Readiness check (whether an order close to its due date is ready to ship) and Pareto (which reason costs you the most time). It answers four questions:

• Which job is stuck? — the program scans the open orders every hour and flags the stuck ones itself (section 3).
• Why, and with whom? — every hold has a reason, an owner and a target resolution date; the order does not move on until the hold is resolved (section 4).
• Has it been forgotten? — once the target date passes, the hold goes first to the owner, then to the supervisor, then to the administrator (section 6).
• Will the delivery hold? — the due-date radar and the readiness check show the delay before the delivery day (sections 8 and 10).

2. The concepts you need

Hold

A hold is the record that says a work order is stuck (and, where given, one particular operation of it). It carries: the reason, the owner, who opened it, when, the target resolution date and a description; once resolved, the resolution note, who resolved it and when. If the program opened the hold itself, its source reads automatic; if a person opened it, manual. Every change goes into the hold's trail: opened, updated, resolved, cancelled, escalation.

Three statuses

Reason codes

Every company starts with these reason codes:

BILINMIYOR cannot be picked by hand; it is used only when the program finds the stuck work itself. Before such a hold can be resolved, it must be changed to the real reason — otherwise the Pareto fills up with "we don't know why" and teaches nothing.

Owner, order owner and the unowned hold

• Owner: the user in charge of resolving the hold. On a manually opened hold the owner is mandatory.
• Order owner: the foreman of the work order. When the program opens a hold itself, it picks the owner in this order: first the order owner; if none, the person who pressed "Started" on the current operation; if none either, the hold stays unowned. Make assigning the order owner a habit — most unowned holds come from here.
• Unowned hold: a hold with no owner. It is not left waiting: it goes straight to the system administrators, who are asked to assign an owner.

The "Started" stamp

When the operator takes up an operation, they press Started; the operation's start time and the person who started it are recorded. The first stamp is never overwritten — pressing again does not move the start forward. If there is no stamp, the program derives the start from that operation's first production confirmation. The start time is the only reliable way to tell that an operation "is taking longer than expected"; pressing it makes the overrun detection accurate.

3. Automatic detection — how the program finds stuck work

The program scans every hour; if you do not want to wait, press Scan now. Only open orders (open or in production) are looked at, and the order's current operation (the first unfinished one) is evaluated. Durations are counted in working days: Monday–Friday. There are four rules:

Default thresholds (the administrator can change them, section 12): multiple 2, waiting 2 working days, daily minutes 480 (used when the machine has no daily minutes of its own), escalation on, detection on.

When the program finds a stuck job, it opens the hold itself: reason No reason entered (automatic detection), target date 1 working day later. A message describing the problem goes to the owner (to the system administrators if the hold is unowned): which order, which operation, for how many working days — with the request "Enter the reason and the target date".

4. The mandatory hold gate — a stuck order does not move

While the order has an open hold, three things are stopped:

• No production confirmation can be written (no produced quantity can be reported).
• No operation can be completed. Undoing a completed operation, however, is allowed.
• "Started" cannot be pressed.

This is the 12th production-confirmation gate and its code is BLOKE_ACIK. On the screen you see a message of this form:

BLOKE_ACIK: emir 1234 takılı — bloke #56 (Malzeme yok, sorumlu ahmet, hedef 2026-09-29). Önce blokeyi çözün (Takılan İşler); sistem yöneticisi blokeOnay ile geçebilir

The message names the order, the hold number, the reason, the owner and the target date; what you have to do is resolve the hold in Stuck Work (section 5). Check (the dry run) on a production confirmation lists this gate together with the other gates.

5. Opening, updating, resolving and cancelling a hold

Opening a hold by hand

Do not wait for the program to notice: open the record the moment the work gets stuck.

1. On the Stuck tab, press Open hold and choose the work order. If you wish, also choose the operation where it is stuck.
2. Choose the reason (the list in section 2).
3. Choose the owner — the person who will solve it. It can be someone outside the team (a buyer, a quality inspector).
4. Enter the target resolution date. A date before today is not accepted.
5. If you wish, write a description, then save. An "assigned to you" message goes to the owner (not if you assigned it to yourself).

A hold cannot be opened on a closed order. If the same order and operation already has an open hold, a second one is refused: "There is already an open hold for this order/operation; update that one."

Updating

The reason, owner, target date and description of an open hold can be changed. Every change stays in the trail as "old → new". When the owner changes, the new owner gets a message. The target date can be moved forward but not back before today; moving it forward does not lower the escalation level (section 6).

Resolving

1. If the hold was opened automatically and its reason is still No reason entered, first choose the real reason. You can do this in the same step as the resolution.
2. Write the resolution note — it is mandatory. Say in one sentence what was done ("Sheet metal arrived from the supplier, lot 2026-114").
3. Save. The hold becomes Resolved, the order can move again and the stuck-clock starts again from that moment.

If you try to resolve without choosing the reason, you see this message: "Choose the real reason before resolving: the automatic detection's reason is still 'No reason entered'."

Cancelling

Cancelling is for wrong records: a false alarm, the wrong order, two records of the same problem. If the problem was real and has been removed, use resolve, not cancel — otherwise the lost time disappears from the Pareto. Only the system administrator or the person who opened the hold can cancel; a hold the program opened itself can be cancelled only by the system administrator. A reason for the cancellation is mandatory.

6. Escalation — who hears about it once the target date passes

If the target date has passed and the hold is still open, the program counts the delay on every scan (today − target, in calendar days) and moves the message up step by step:

Where is the supervisor found? In order: (1) the manager field on the owner's Human Resources personnel card; (2) if empty, the production manager in the settings; (3) if that is empty too, the system administrators. For the chain to work, link users to their personnel cards and fill in the manager field.

• An unowned hold does not wait for the middle steps; it goes to the administrators in one step.
• If the scan has not run for days (the server was down), all the missed steps go out together on the first scan.
• The target date can be updated (it stays in the trail), but the level never goes down: a hold that has reached the supervisor does not become the owner's alone again because its target was moved forward.

7. The morning summary and installation settings

Once a day, on the first scan after the morning hour, an announcement goes to everyone:

Takılan işler: 4 açık bloke, 1 sahipsiz, 2 hedefi geçmiş, 3 emrin termini 2 gün içinde. Ayrıntı: Üretim > Takılan İşler.

By default the announcement goes out only when attention is needed: when there is stuck work, an unowned hold or an order due within 2 days. An announcement saying "0 stuck jobs" every morning stops being read after the third day.

The hourly scheduler runs on the server's service company (HNR_FIRMA) with a write-enabled service session; if no service session is defined, the scan only reads, opens no holds and writes the reason into the task status. In an installation with several companies, the other companies are scanned when a production user opens Stuck Work, using that user's session. To switch automatic detection off for one company, turn detection off in the settings (section 12); manual holds, the gate and escalation keep working.

8. The due-date radar — which order will miss its delivery

The radar lists open orders that have a due date and works out an estimated finish for each: the minutes of the unfinished operations, based on the remaining quantity, are divided by the machine's daily minutes and added as working days; today counts as day 1. If the order is stuck, the estimate starts from the hold's target date.

If the company has an active capacity plan, the radar also shows the order's finish date in the plan; if the plan passes the due date, the order counts as will miss regardless of the estimate.

9. Waiting on me — who sees what

This tab is open to everyone and shows only you. It has three parts:

• Holds assigned to me — the open holds you own; those past their target are counted separately.
• Escalated to me — holds of the people you supervise that have reached the supervisor level; if you are a system administrator, also holds escalated to the administrator and unowned holds.
• My operations (for users with the production permission) — operations you marked "Started" that are not finished yet; the next job on orders you own; the next job on open orders you touched in the last 30 days through a production confirmation or an operation completion. Sorted by due date; stuck ones are marked.

Only the system administrator can see another user's waiting work. A person who has been assigned a hold but has no production module permission (purchasing, quality, shipping) sees, updates and resolves that hold here — for example, the buyer closes a "No material" hold when the material arrives.

10. Readiness check (T-7) — is it ready to ship

For orders due within 7 days (or a single order you pick), seven items are checked. Each item is OK, Missing or Does not apply to this order (not counted as missing). An order with no missing item counts as ready.

11. Pareto — what eats the most time

The Pareto groups the holds opened in the chosen date range by reason code and gives, for each reason, the working days lost, the share (%) and the cumulative share (%). The default range is the last 7 days. The loss of a hold that is still open keeps growing up to today; cancelled holds are not counted. The top one or two reasons are usually most of the loss — start improving there.

12. Settings — the stuck thresholds

The thresholds are changed in Stuck Work › Settings and are stored in the company's Production Preferences under takilma. Only the system administrator can change them; other users see the values but cannot edit them.

13. On the phone

• In the Garden + Steps interface the Stuck Work list opens in card view; each card shows the order, the reason, the owner and the target date. A new hold is opened with a step-by-step wizard: order, reason, owner and target date are asked one after another.
• In the Pulse interface, the Me › My stuck work box shows the number of open holds assigned to you and escalated to you (see the Mobile guide).

14. Frequently asked questions

Related guides: Production — warnings and gates · Production Preferences · Human Resources · Mobile · MES (hold suggestion from a downtime reason).