ERP Integration
Your plan knows what you intended. Your machines know what happened.
Those two halves rarely meet. The plan lives in one system, the truth lives on the control, and somebody in the middle retypes one into the other a shift later. Integration closes the loop both ways: the part, the target and the standard rate coming down to the floor, and the measured count, downtime and OEE going back up.
The problem
Why this matters
- Production counts, downtime, and OEE already live in ThingConnect, measured from the controller, but planning and costing decisions still run on numbers typed in by hand, often a shift or more behind.
- A missed or late downtime entry means a planner is scheduling against a number that was already wrong when it was entered.
- Standard cycle times and setup times sit wherever you keep your routings, and a second copy of them sits in ThingConnect so it knows what good looks like. Two copies of the same number is one copy too many.
- Closing the loop between what the machine actually did and what the plan assumed it did is the same "measured, not estimated" problem the rest of ThingConnect already solves on the shop floor.
Whatever you run
What counts as an ERP here
The useful question isn't which ERP you run. It's where the plan lives today, before the machine sees it.
A full ERP suite
Routings, work orders and costing already in one system. The connection has the most to draw on here, and the most to give back.
An accounting package that grew
It holds your part masters and your job numbers because that is where they ended up, not because anyone designed it that way. It's still the system of record, and it still counts.
A planning spreadsheet
One person maintains it, and a copy of it lives in four inboxes. Everything on this page applies, and the drift below applies more.
We say ERP throughout this page because it’s the shortest word for it. Read it as whichever of these you actually have.
The loop
What moves, and which way
Integration isn’t a one-way feed of numbers upward. Half the value is the plan arriving on the floor from the system that already holds it.
Your plan already knows
Down to the floor
Planning data ThingConnect keeps today, maintained by hand, that your own system of record already holds.
- Part number and revision
- Which part this machine is making, and which version of it
- Target quantity, by shift
- What the plan actually asked for, rather than a number recalled later
- Standard cycle time
- The engineered rate every performance figure is measured against
- Setup and changeover time
- The CNC cannot report this at all. It is a routing number or it is nothing.
The floor already knows
Back up to the plan
Measured from the control as it happens, not typed in at the end of a shift.
- Actual count against that target
- Parts produced, read from the machine, next to what was planned
- Downtime with its reason
- Every stop, classified by the operator in two taps, with real hours on it
- Cycle and run time
- How long the job really took, against how long it was quoted to take
- OEE rollups
- Availability, performance and quality on numbers nobody had to estimate
The up-flow is what production tracking already reports against plan, and what OEE monitoring rolls up. If a particular ERP is what your plant runs on, tell us which. It’s useful input on what we wire up first.
Two copies of one number
The number nobody notices is wrong
ThingConnect needs a standard cycle time per part to know what good looks like, and it keeps one. So does your routing, wherever that lives. Whenever the same engineered number is maintained in two places by hand, the copies drift, and a part whose cycle time never got filled in doesn't announce itself. It just quietly reports a performance figure that is too high, on a machine that looks fine.
If that routing is a spreadsheet, this is sharper still, not softer. A file has no validation to catch a blank rate, no history to show when it changed, and no single copy. The one on the floor is a download from three weeks ago, and nobody can tell by looking.
Setup time makes the case on its own: there is no reading on any control that reports it. It comes from your routing or it stays at zero. That is the clearest argument for the plan coming down to the floor rather than being retyped into it.
Scope
What this isn't
Worth being plain about, because plenty of things in this category quietly grow into a second system to maintain.
Not a second ERP
ThingConnect doesn't schedule work, hold inventory, raise purchase orders or cost a job. Your ERP does those things and keeps doing them.
Not an MES
No routings to maintain here, no work-in-progress to move between operations. ThingConnect measures what the machines did and reports it back.
Not more work for operators
Counts and stops come from the control. The only thing an operator adds is the reason for a stop, on the screen already in front of them.
Questions
Before you book a demo
Alongside it. ThingConnect isn't a planning or costing system. It pushes shop-floor numbers into the ERP you already use rather than replacing it.
No, and the direction people forget is the useful one. The part, the shift target and the standard cycle time come down from the ERP, which is already the system of record for them; the measured count, downtime and OEE go back up. See "What moves, and which way" above.
That's exactly what integration is for. ThingConnect needs a standard cycle time per part to calculate performance, and today that's kept in its own configuration. Sourcing it from the routing you already maintain removes the second copy, and with it the risk that the two drift apart without anyone noticing.
Production counts, downtime events, and OEE rollups, all measured from the controller and landing in your planning system instead of being entered by hand. See "What moves, and which way" above.
Yes, and often more so. What matters is that the plan lives somewhere before the machine sees it: a full ERP, an accounting package that ended up holding your part masters, or a spreadsheet one person maintains. The loop is the same in each case; only the mechanics of connecting differ. A spreadsheet is also where standard times drift fastest, because nothing in a file catches a blank rate or shows when it changed.
Tell us where your plan and your routings live today and we'll walk through what the connection looks like for it on a demo call. What ThingConnect measures and sends is the same either way.
No. Production counts and machine stops are detected from the controller, and reasons are classified by the operator in two taps on the screen they already use. The ERP side is downstream of that, with no extra data entry on the floor.
Talk to us
Tell us which ERP data would help you most
A 30-minute call to talk through your ERP setup, which numbers your planners actually need from the floor, where your standard times live today, and how downtime, OEE, and utilization tracking would read your own machines.