OEE Software + Infor Integration Guide (2026): Architecture, Pitfalls, and a Working Playbook

After ~40 European integrations 2022-2026, three patterns dominate. Mixing is normal, most plants run two of three. Wrong starting pattern adds 3-4 weeks.

Pattern 1, shop-order-event-driven via ION Pulse (default for CSI and M3).

Infor publishes a BOD on ION every state change. OEE subscribes, listens, updates context in near real-time. Latency 1-5 seconds end-to-end. Pros: real-time, operators never see stale data. Cons: Pulse misbehaves under high BOD volume, needs dead-letter queue handling.

Pattern 2, schedule pull via ION API Gateway (recommended for LN and conservative IT).

OEE polls API Gateway every minute or two for open shop orders. No subscriptions, no event handlers. Latency tunable 60-120 seconds. Pros: easier debug, no Pulse dependency, IT change-control friendly. Cons: 60-second blind spots, heavy call volume on shift starts.

Pattern 3, historian-pull with later ERP reconciliation (older M3, on-prem LN, plants without ION live).

OEE records autonomously against work centers, nightly job reconciles intervals to shop orders by timestamps and operator IDs. Latency hours to a day for ERP context; OEE stays real-time. Pros: zero risk to Infor, no ION licensing, 2-week go-live. Cons: operators see "unassigned" downtime live, no write-back, Pareto charts are next-day.

Decision question: does the operator need the open Infor order on the OEE screen during the shift, or is daily ERP-aligned reporting enough? Plant directors chasing real-time variance pick Pattern 1. Reliability engineers running monthly Pareto get away with Pattern 3. Most enterprise IT teams compromise on Pattern 2.

How much does Infor + OEE really cost? One site, one module, 3-5 work centers: vendor implementation €15-75k + internal hours 30-120. ION licensing usually in your Infor subscription, but confirm Pulse/API Gateway tiers.

Which OEE vendors actually support Infor?

• TrakSYS, strong Infor track record at large discrete manufacturers, full ION BOD support, expensive licensing.
• FORCAM FORCE, common in DACH automotive/aerospace on M3/LN, ION-native, long implementation cycles.
• AVEVA System Platform / Wonderware, tight when Infor sits next to AVEVA Historian, unfriendly licensing above 10 lines.
• Fabrico, supports CSI, M3, LN via ION BOD subscriptions and API Gateway, layers computer-vision OEE so no PLC tap required at pilot work center, fixed-fee under €30k.
• Many smaller cloud OEE tools claim Infor support but ship a CSV importer. Demand a live ION call before signing.

Six-test checklist:

1. BOD subscriber receives live ProcessShopOrderOperation within 5 seconds.
2. Work-center context (open order, item, planned quantity) on OEE board matches Infor exactly.
3. Downtime tagged to active shop order within 60 seconds.
4. Scrap on OEE tablet shows up as quantity-rejected on same Infor operation within 2 minutes.
5. Planned-vs-actual variance reconciles between Infor and OEE within 1%.
6. OAuth token refresh during pilot does not break the subscription.

What about Infor EAM and CMMS? Same ION pattern. Maintenance work orders flow into OEE as planned downtime. OEE-reported breakdowns trigger EAM work request via SyncWorkOrder. Most plants run Fabrico native CMMS on top of OEE, reconciled daily to Infor EAM.

Bottom Line

Infor + OEE works when treated as a real integration project. Right pattern for your module (Pulse for CSI/M3, API Gateway for LN, historian-pull when ION is not live), BODs mapped once and properly, six weeks for one line, demand a live ION call before signing. Skip those steps and you spend the next year explaining why the dashboard and ERP disagree.

Want to see a finished Infor + OEE integration? Fabrico runs working ION BOD subscriptions and API Gateway calls against live CSI, M3, LN sandboxes. Book 25 minutes.

Six weeks for one line, one Infor module, one ION instance. Crawl, walk, run.

Weeks 1-2, scoping, BOD mapping, auth.

• Confirm Infor module (CSI / M3 / LN), version, ION in production.
• Inventory BOD subscriptions and API endpoints.
• ION service account, OAuth credentials, API Gateway entitlement.
• Map every Infor reason code (downtime, scrap, rework) to OEE reason tree.
• Pick pilot work center, the highest-volume asset.

Week 3, subscriber endpoints and event handlers.

• Stand up BOD subscriber (Pattern 1) or polling job (Pattern 2). Verify ION delivers.
• Catch and parse the first live ProcessShopOrderOperation; confirm operation, work center, quantity match Infor.
• Write synthetic acknowledge back through ION, proves write-back is real.
• Dead-letter queue, retry strategy, BOD error catalogue.

Week 4, one-line pilot. Operators see open Infor order on OEE screen. Scrap from tablet writes to Infor within a minute. Run the six-test checklist (Slot 4).

Week 5, scaling. Add 3-5 work centers. Reuse subscriber. Watch for cell-specific config quirks. Train shift supervisors on the board.

Week 6, cutover, sunset legacy, document. Sunset manual downtime log. Move BOD subscription monitoring into observability stack. Write the runbook (BOD catalogue, API calls, OAuth refresh procedure). Schedule 30-day review.

Two warnings: OAuth tokens are short-lived (number-one production outage is missed refresh, bake refresh logic into Week 3). Service-pack upgrades change BOD payloads (subscribe to Infor release notes, test subscriber against sandbox before every production upgrade).

First place Infor + OEE projects fail: the version question. Three flavours show up across European manufacturers, each with different integration surfaces.

Infor CloudSuite Industrial (CSI), formerly SyteLine. Mid-market discrete cloud. Shop floor lives around shop orders, work centers, operations. Integration through ION with BODs, ProcessShopOrderOperation, AcknowledgeShopOrderOperation, SyncWorkCenter, SyncItemMaster. ION API Gateway exposes REST on top.

Infor M3. Heavy in Nordics, food, fashion, light discrete. Operates on production orders and reporting points. ION rules. Common BODs: ProcessProductionOrder, AcknowledgeProductionOrder, SyncBillOfMaterial. Older MEC / M3 APIs via H5 still used by some integrators.

Infor LN (formerly Baan). Process and discrete with long automotive/aerospace tail. ION-native, but on-prem is common. Direct DB reads at older sites, unsupported, breaks on upgrades.

When a vendor says "we integrate Infor," ask three questions:

1. Which module? CSI, M3, or LN. BODs differ.
2. Which ION pattern? Pulse subscriptions (event-driven), API Gateway (request/response), or both.
3. What gets written back? Read only is easy. Operation completions, scrap, downtime reasons, that is where ION Workflow and ProcessAcknowledge BODs come in.

What "integrated" should mean in 2026:

• Every downtime your CV OEE or sensor layer detects matches a live Infor shop or production order within 60 seconds.
• Scrap entered on the OEE tablet writes back as quantity-rejected on the same operation.
• Planned downtime (changeovers, CMMS PMs) appears on OEE timeline before it happens.
• Work-center status changes flow into OEE so the dashboard never lies.
• Operators see the open shop order, standard cycle, and takt, from Infor, on the OEE screen.

A vendor that cannot map every flow to a specific BOD or API call is not integrated.

Related deep-dives: OEE + Epicor integration · OEE software selection process · OEE software hidden-cost checklist · Closing the OEE-CMMS loop.