Costing
Machine hours per part feed labor and overhead absorption. When the hours are measured, the standard cost of a part is easier to trust and easier to defend.
Cycle time, as written and as run
Vectis Works lines the two up, per part number and per machine, so the people who cost jobs, quote work and plan capacity are working from what the floor actually did.
Sample Illustrative part HSG-0418, a machined housing on Mill 2. Made-up figures, not client data.
Standard
64sec per part
What the routing in the ERP says Mill 2 needs for one housing.
Recorded
71sec, median
What the controls logged across a week of cycles on the same machine.
The gap
+7sec per part
Small on one cycle. It rides along into every quote and every schedule that uses 64.
Most routings in a shop were written years ago, by someone timing a good run on a machine that was newer then. Nobody did anything wrong. The shop simply moved on while the numbers stayed put.
Smaller lots and more part numbers mean more setups per shift than the original standard ever assumed.
New inserts, different fixtures or a supplier change can make a part faster or slower, and the routing rarely hears about it.
A press or a lathe that has run for many years does not always hold the rate it was rated for.
Experienced operators find their own rhythm. Second shift may run the same job a little differently from first, for good reasons.
The point is not to find a culprit. It is to price and plan with numbers that describe the plant you have today.
PLCs, HMIs and servers such as Kepware hold far more than most plants look at. We connect to them read-only and pull the four things this comparison needs.
The start and finish of each cycle, so a real cycle time falls out per part.
Good parts made, per machine and per job, counted where they are made.
How long each machine was actually producing, shift by shift.
When it sat idle or stopped, and for how long, with a reason where the machine logs one.
The standard comes from your ERP or routing. The recorded value comes from the machine that ran the job. Put next to each other, the parts that need a second look tend to stand out quickly, in both directions.
Invented part numbers and times, shown to illustrate the view.
| Part number | Machine | Standard | Recorded | Gap |
|---|---|---|---|---|
| HSG-0418 | Mill 2 | 64 s | 71 s | +7 s |
| BRK-2210 | Press 3 | 9.0 s | 9.4 s | +0.4 s |
| SHF-1136 | Lathe 7 | 142 s | 128 s | −14 s |
| CLP-0752 | Press 5 | 6.5 s | 8.1 s | +1.6 s |
A part that runs faster than its standard matters as much as one that runs slower: it may be quoted too high, or holding capacity on the schedule that the floor does not need.
Machine hours per part feed labor and overhead absorption. When the hours are measured, the standard cost of a part is easier to trust and easier to defend.
A repeat part can be quoted from how it ran last time on that machine, and a new part from the closest part the plant has recorded.
Capacity plans built on recorded rates show which work centers are full and which only look full on paper.
Supervisors see which machines drift furthest from standard and can ask the floor why, with the numbers in hand.
Many plants keep this comparison alive in a workbook that one person updates by hand from exports. It works until that person is out, or the file breaks.
We move that work into a small system that fills itself from the machines and the ERP, and we build it inside your environment.
Your existing workbook shows us which parts, machines and columns matter. We keep the logic your team already relies on.
The connection to the machines is read-only. Nothing we add can change a setpoint or a program, and your own team can check that.
The code, the data and the system stay yours. Ongoing support is available if you want it, with no license tail if you do not.
Cycle times are one use of the data your machines produce. The same team builds the rest.
Tell us the part and the machine it runs on. We will look at what the standard assumes and what it would take to read the recorded cycle, and tell you plainly if this is not a fit.