How to connect Power BI to DELMIAworks (IQMS)
Short answer
Connect Power BI to DELMIAworks through Oracle views on a reporting replica, a nightly warehouse, or the vendor API, using an on-premises data gateway for scheduled refresh. Build dashboards for OEE, on-time delivery, scrap and inventory, then add natural-language Q&A on top of the same model.
Most DELMIAworks plants already have report writers and Excel exports, and many want Power BI because it sits in the Microsoft stack they already license. The question is what to connect to. Pointing refreshes at the live production Oracle database is the first mistake nearly everyone makes.
A good Power BI setup separates analytics load from transaction load, models the plant in terms planners recognize, and refreshes on a schedule that matches how often decisions change.
This page walks the architecture options for an IQMS-based plant and the dashboards that tend to earn their keep.
What teams use it for
OEE by press and shift
Availability, performance and quality from RealTime MES counts and downtime, sliced by press, mold, shift and part.
On-time delivery
Compare promise date with ship date by customer and part family, with drill-through to late order lines.
Scrap and regrind analysis
Scrap quantity and cost by reason code, mold and material lot, trended over weeks.
Inventory and coverage
Days of cover by resin and component, with open PO receipts overlaid against forecast demand.
Purchase price variance
Standard versus actual cost on receipts by supplier and item, for buyers and finance.
Backlog and capacity
Open order hours against press capacity by week, highlighting overloaded work centers.
DELMIAworks (IQMS) integration surface
- Oracle database: Power BI can use an Oracle connector, ideally against read-only views or a reporting replica, with the Oracle client installed on the gateway machine.
- Reporting replica or standby: a read-only copy, such as one maintained with Oracle Data Guard where licensed, protects production from heavy queries.
- Staging warehouse: a nightly ETL into SQL Server or another store gives a stable star schema and faster dashboards.
- Vendor web services: where available, useful for smaller datasets or cloud tenants where direct database access is not offered.
- Version and license details vary, so confirm what your agreement allows for data access.
How it works
- 1
Choose the data path
Prefer a reporting replica or nightly warehouse. Use direct views on production only for small, indexed queries.
- 2
Model the plant
Create dimension views for item, customer, supplier, press, mold and date, and fact views for orders, shipments, production and scrap.
- 3
Install the gateway
Set up the on-premises data gateway on a server that can reach Oracle, with the Oracle client libraries and a read-only credential.
- 4
Build the semantic model
Define relationships, measures such as OEE and OTD in DAX, and row-level security by plant or customer.
- 5
Design the dashboards
Start with three pages for the roles that will use them, keeping each tied to a decision.
- 6
Schedule refresh
Refresh MES-driven data more often, such as every 15 to 60 minutes, and financial data nightly.
- 7
Add Q&A
Curate synonyms and featured questions, or add an on-prem language layer that generates governed queries over the same model.
Example flows
Weekly operations review
Trigger: Monday morning review meeting.
- Dashboard shows last week's OEE by press with the bottom five highlighted.
- Team drills into one press to see downtime reasons by shift.
- They filter to the top reason and view affected molds and parts.
- Actions are logged back in the meeting notes.
Result: The meeting works from the same numbers everyone can reproduce.
Ask a question in plain language
Trigger: A planner types 'scrap by mold last month'.
- The Q&A layer maps the phrase to the scrap measure and mold dimension.
- It generates the visual from the semantic model.
- The planner pins the result to a dashboard.
Result: Ad hoc questions no longer wait in the report writer's queue.
Keeping data on-prem
Power BI Desktop and the on-premises data gateway let you keep the DELMIAworks Oracle database inside the plant while publishing to the Power BI service, which means aggregated model data is stored in Microsoft's cloud. If policy forbids that, Power BI Report Server runs on-premises and can read the same views, with fewer cloud-only features. For natural-language questions without sending data out, add a locally hosted open-weight model that translates questions into governed queries on your semantic layer, and keep the warehouse and model on your network.
Security checklist
- Use a read-only database account scoped to reporting views and store its credential in the gateway.
- Apply row-level security by plant, customer or cost visibility in the semantic model.
- Keep the gateway on a patched, monitored server with limited outbound rules.
- Review which datasets are shared externally, especially customer-specific ones.
- Decide whether cost and margin fields belong in the model at all.
Ways to get it done
Vendor reporting tools
Effort: None to low.
If existing DELMIAworks reports and analytics cover your needs, adding Power BI may be unnecessary.
Build in-house with warehouse
Effort: Four to eight weeks for the first set of dashboards.
Plants with a BI analyst and an Oracle DBA can model and publish themselves.
Partner such as Netray
Effort: A few weeks beyond the base dashboards.
When you also want natural-language Q&A running on-prem over the same model.
Pitfalls
- !Refreshing heavy queries against production Oracle during shifts, slowing the shop floor.
- !Joining directly across many raw tables in Power Query instead of in views.
- !Defining OEE differently in each report, so numbers never match.
- !Ignoring time zones and shift calendars when slicing MES events.
- !Forgetting gateway capacity and failing refreshes during month-end.
Frequently asked questions
Can Power BI connect directly to DELMIAworks?
Power BI can connect to the underlying Oracle database with its Oracle connector, subject to your license and policy. Use read-only views and preferably a reporting replica or warehouse rather than the live production instance.
Do I need a data warehouse?
Not always. Small deployments run well on curated views. A warehouse helps with history, speed and combining DELMIAworks with other sources, and it protects production from analytic load.
How often should data refresh?
Match the decision. MES and downtime views are useful at 15 to 60 minute intervals, while finance and backlog views usually need only nightly refresh. More frequent refresh adds load for little benefit.
Can I ask questions in plain English?
Power BI offers built-in Q&A features depending on your plan and configuration, and they work best on a clean semantic model. For stricter privacy, a local model can generate governed queries without sending data out.
Is a gateway required?
If your database is on-premises and you publish to the Power BI service, yes, an on-premises data gateway is the standard way to refresh. Power BI Report Server on your own network is an alternative that avoids it.
More for DELMIAworks (IQMS)
Want Power BI working on your DELMIAworks (IQMS)?
A free 30 minute assessment: we map the integration path for your version and setup, and where it should run.
Get a free assessment