Manufacturing PLM and Acumatica
Acumatica already carries bill-of-material revisions with effective start and end dates. That is the whole shape a PLM needs, so the connector uses it directly instead of inventing a parallel structure in custom fields.
Field map — object by object
| Manufacturing PLM object | Lands as | Master | Direction | On conflict |
|---|---|---|---|---|
| Part | Stock Item / Non-Stock Item | Manufacturing PLM | Manufacturing PLM → Acumatica on release | Manufacturing PLM wins; Acumatica fields read-only here |
| Released EBOM | Bill of Material revision | Manufacturing PLM | Manufacturing PLM → Acumatica on release | Reconciliation task naming both |
| Effectivity | BOM effective start / end date | Manufacturing PLM | Manufacturing PLM → Acumatica | Dates are ours; edits are flagged |
| Operations | BOM operations | TENANT | Decided per tenant | Usually Acumatica's; named in the config |
| Cost | Item cost / standard cost | ERP | Acumatica → Manufacturing PLM | Display only |
| Inventory | Item quantities by warehouse | ERP | Acumatica → Manufacturing PLM | Cached with an as-of stamp |
| Vendor | Vendor + vendor item | ERP | Acumatica → Manufacturing PLM | Manufacturing PLM adds AVL approval state only |
What this connector is, and what it is not
It publishes released engineering structures into Acumatica as bill-of-material revisions with effective dates, and reads cost, inventory and vendor context back the other way.
Acumatica is the connector with the least impedance mismatch of the six, because its BOM revision already carries an effective start and end date and its planning engine already reads them. Manufacturing PLM's effectivity model maps onto that almost directly, which means fewer translation rules and fewer places for the two systems to drift.
It is not a replacement for Acumatica's own engineering change order module. If you use it, Manufacturing PLM feeds it the released outcome — the review, the redlines and the affected-object analysis happen upstream, where the relationship graph lives.
Nor does it try to own the item master outright. Acumatica will always hold items that originate in purchasing or sales, and the connector's job is to make the division legible rather than to win it.
The mechanism: dates land where planning reads them
On release, Manufacturing PLM resolves the structure at the change's effectivity and creates a new BOM revision with matching effective start and end dates. The previous revision keeps its own window rather than being edited, so the record of what was buildable when survives in Acumatica as well as in Manufacturing PLM.
This is worth dwelling on because it is where most integrations quietly fail. An effectivity date written into a custom field, a note, or a description is invisible to the planning run. An effective start date on the revision is not — it is the field material requirements planning already consults, which means the change takes effect on the date engineering chose without anybody intervening.
Transport is Acumatica's contract-based REST interface, with credentials held per tenant. Writes are idempotent, publication is all-or-nothing per release, and a retried run does not create a duplicate revision.
The failure it prevents
A change is approved in August to take effect on 1 October. Somebody publishes it to Acumatica in August, because that is when the approval happened and the integration fires on approval rather than on effectivity.
For six weeks the planning run believes the new structure is current. It nets requirements against the new component, releases purchase orders for it, and stops consuming the old one — which is still sitting on the shelf in the quantity that determined the October date in the first place.
The mechanism that prevents this is not a better sync schedule. It is publishing the effectivity itself into the field planning reads, so that approval and effect stay two different events in both systems, exactly as engineering intended.
How it meets the rest of the product
What publishes is a resolved structure: effectivity, variant selection, revision policy and plant scope evaluated together into a single configuration. Acumatica receives a concrete bill, because a BOM revision holds a structure and cannot evaluate conditions.
Warehouse-level inventory reads back and appears on impact analysis, so an engineer scoping a change sees where the affected stock actually sits. A component with three months of coverage in one warehouse and none in another is a different disposition problem from one with a single balance.
Cost and vendor fields render read-only in Manufacturing PLM, labelled with Acumatica as owner, and feed cost rollups through the resolved structure so design decisions can be priced without either system writing the other's figures.
Component risk works the same way. Lifecycle health, end-of-life notices and lead-time exposure come from the component data feeds and attach to the manufacturer parts on your approved list, while Acumatica keeps the vendor relationship and the purchasing history. Neither system has to model the other's half to answer its own question.
How it meets your ERP
Item identity matching runs on manufacturer part number first, internal part number second, then a configurable tie-break. Acumatica will hold items Manufacturing PLM never created — packaging, freight, service items, non-stock — and those keep their own identity. Anything unmatched is surfaced for a person rather than merged on a similarity score.
Every sync is an attributed audit row naming the connector, the run and the direction, and sync history is queryable with bounded retries. Publication is all-or-nothing, because a half-written revision looks complete to the planning run that reads it next.
If a published revision is edited in Acumatica, the next run raises a reconciliation task naming the item, the revision, the field, both values and the owning system. It does not overwrite. Operations lines in particular are frequently and legitimately adjusted downstream, which is exactly why that row of the matrix defaults to tenant-decided.
Where the boundary is
No purchase orders, production orders, inventory transactions, invoices or ledger entries, in either direction. Manufacturing PLM reads cost and inventory as context and writes engineering definition; everything transactional stays in Acumatica.
BOM operations are not published by default. Where manufacturing owns routings and work centres inside Acumatica, Manufacturing PLM stops at the material structure rather than publishing an opinion about a process it does not own.
Facts
| Publishes as | BOM revision with effective start and end dates |
| Effectivity | Lands on the field planning already reads |
| Trigger | On change release, dated to effectivity — not to approval |
| Prior revisions | Left intact; windows are not edited |
| Transport | Contract-based REST, credentials per tenant |
| Reads back | Cost · inventory by warehouse · vendor |
| Operations | Not published by default — tenant decides |
| On conflict | Reconciliation task naming item, revision and field |
Frequently asked
How does effectivity reach Acumatica planning?
As the effective start and end dates on the BOM revision — the fields material requirements planning already consults. Written into a custom field or a note, effectivity is invisible to the planning run, which is how a change dated for October starts driving purchase orders in August.
Does publication happen on approval or on effectivity?
Publication happens on release, but it carries the effectivity date rather than today's date. Approval and effect stay two separate events in both systems, exactly as engineering intended. Collapsing them is the single most common way a phased cutover turns into a stock problem.
Are previous BOM revisions modified?
No. Each release creates a new revision with its own window and leaves earlier ones untouched, so Acumatica keeps the record of what was buildable when. That history is what a warranty investigation depends on, and editing it in place to save a step is a poor trade.
Who owns BOM operations?
Usually Acumatica, and that is the default. Operations sit under the tenant-decided row of the matrix because manufacturing legitimately adjusts routings downstream. Manufacturing PLM publishes the material structure and stops, rather than publishing an opinion about a process it does not own.
What about items created by our buyers?
They keep their Acumatica identity. Packaging, freight, service and non-stock items are not pulled into engineering's namespace. Matching runs on manufacturer then internal part number, and anything that does not match cleanly is surfaced for a person rather than merged automatically.
Does inventory show per warehouse?
Yes, and it appears on impact analysis. A component with three months of coverage in one warehouse and none in another is a different disposition problem from one showing a single balance, and the engineer scoping the change should be able to see which situation they are in.
Is anything transactional written back?
Nothing at all. No purchase orders, no production orders, no inventory transactions, no invoices and no ledger entries, in either direction. Manufacturing PLM reads cost and inventory purely as context for impact analysis, and writes engineering definition. Everything transactional stays inside Acumatica where it belongs.