What tidying the structure means
The plant structure model has four levels: view (department), building, line and location. A view is not a building. Tidying means assigning every location to the correct line, every line to a building and every building to a view, with correct geometry and level.
In plants with a large site and many kilns this structure builds up over years. Duplicates appear, locations sit outside their line, lines overlap and objects have no position. The longer this goes on, the more expensive the repair, because every change affects the risks and permits attached to those places.
Manual cost: hours that are easy to overlook
Sample calculation: tidying 800 locations by hand at 3 minutes per item means 40 hours of work. This is an assumption for your own costing, not the result of an implementation. Every plant has a different time per item, depending on how well the places are described and how far one has to search for information.
On top of the 40 hours come things that do not show in a simple multiplication: a second person's check, corrections after mistakes are found, re-checking overlapping lines and agreeing with the safety department. In a real project, manual work rarely finishes in one pass.
Errors that manual work introduces
A line without its own shape does not inherit the outline of the whole building, because it gets a separate field carved out of the building so that lines do not lie on top of each other. By hand it is easy to assign the same field to two lines. Similarly, each line must contain exactly one collective Other location, placed in the middle of the line, and line names must be unique within the building.
Hours of manual tidying by number of runs
Script cost: preparation, gates, checks
A migration script costs time for preparation and testing, but then it passes through the whole structure in one run. The condition is one: gates. Migration gates, that is overlapping lines, pins outside their line, the Other location away from the middle and duplicates, each have a red test that proves the gate really catches the error.
After the structure is regenerated, the comparison is made by content, meaning name, area, level and geometry, not by identifiers, because identifiers change on regeneration. This way the check does not depend on whether the numbering matches.
When the calculation tips toward a script
The threshold depends on three variables: the number of locations, the time per item and the number of runs. With 800 locations and 3 minutes each, every further regeneration of the structure means another 40 hours of manual work. The script pays for itself already on the second run, and in a plant that tidies its structure after every shutdown, on the first.
If the structure is to be tidied once and covers a few dozen locations, manual editing in the editor is cheaper. A script pays off where repeatability matters and where there must be proof that the result passed the gates.
How to build your own calculation
Collect three numbers: the number of locations to tidy, the average time of manual editing for one item, and the number of runs per year. Multiply the first by the second and add the checks. Compare with the cost of preparing the script and the gates. With 800 locations at 3 minutes, the reference point is 40 hours per run.
Sample calculation: manual structure tidying
| Parameter | Value |
|---|---|
| Number of locations to tidy | 800 locations |
| Time of manual editing for one item | 3 minutes |
| Number of runs | 1 run |
Result: 800 times 3 minutes is 2,400 minutes, which is 40 hours of work per run, without checks or corrections.
Note: Sample calculation: the assumptions serve your own costing and are not the result of an implementation. The price of the script is individual.
Video
Frequently asked questions
Where do I get the time per item for my own calculation?
Measure 20 items in the editor, including the search for information about the place, and take the average. The 3-minute result is a sample assumption, and your plant may have a different value.
Does the script replace human checking?
No. The script with gates detects overlapping lines, pins outside their line, Other locations away from the middle and duplicates, but decisions about the plant layout are made by a person.
How do I know a gate works?
Each gate has a red test: deliberately wrong data that the gate must reject. Without such a test the gate is only a hypothesis.
How do I compare the structure before and after regeneration?
By content: name, area, level and geometry. Identifiers change on regeneration, so they are not suitable for comparison.
Calculate the cost of tidying your structure
We will go through your numbers together: locations, time per item and number of runs. Book a conversation about migrating the structure with gates.
Let's talk