Core service
A Power BI integration that ends the spreadsheet era on the floor
We do not sell a dashboard. We build the whole path of the data — from the controller signal to the decision at the morning stand-up — and leave you a model you can develop without us.
Scope of work
Six layers every project has to have
Most failed industrial BI projects stop at the third layer — a good-looking report on bad data. Here we go the whole way, to the last layer.
Data acquisition
Controllers and sensors connected over OPC UA, Modbus/TCP and MQTT. Readings of vibration, temperature, cycle times and machine states.
Integration layer
ETL, validation and normalisation of data from many sources. A warehouse in Microsoft Fabric or on your existing infrastructure.
Semantic model
DAX measures for OEE, MTBF, yield and takt. One definition of each metric across the plant — the end of arguing about numbers.
Reports for people
Separate views for the board, production, maintenance and quality. Everyone gets what they can actually influence.
Automatic refresh
The data updates itself. No manual exports, no copied spreadsheets, no “final_v3”.
Security and compliance
Row-level security (RLS), data in your own Microsoft 365 tenant, pipelines built to meet GxP requirements.
The metric we start from
OEE broken down by cause, not reduced to one number
“We are at 64%” on its own changes nothing. The value appears once you can see how much was taken by changeovers, how much by micro-stops and how much by quality rework — and on which shift.
We calculate the same model for a line, a cell and a single machine, so the conversation about the bottleneck stops being a conversation about opinions.
Where the data comes from
We start from what is already connected
A robot, an in-process inspection sensor and an operator panel feed the same stream. An Andon event and a drop in performance appear in the report as one story, not two unrelated facts.
See the MES- PLC controllers (Siemens, Allen-Bradley, Beckhoff)
- SCADA and HMI panels
- IoT sensors retrofitted on older machines
- ERP and APS systems
- CMMS and maintenance tickets
- Weighing and quality control systems
- Utility meters: electricity, compressed air, water
- Spreadsheets and files that are created anyway
Pricing
Three Power BI implementation packages
Prices are net of VAT. Pilot has an open entry threshold; Standard and Scale are quoted individually — the amount depends on the number of data sources and on how readily they can be accessed — and is settled after the workshop, before the contract is signed.
Pilot
One area, one report
from 8,000 PLN
net · 2–3 weeks
Proof of value on real data from the shop floor.
- Analysis of a single production area
- One data source connected (PLC / SCADA / ERP / file)
- One Power BI report with the key KPIs
- Kick-off workshop and handover of the report
- Recommendations for the next steps
Standard
most often chosenSeveral areas of the plant
Quoted individually
after a call with the sales team · 4–6 weeks
A consistent reporting layer for production and the board.
- Several plant areas in one model
- One semantic model (single source of truth)
- 5 reports: production + management
- OEE, performance, scrap and availability reports
- Automatic refresh and row-level security (RLS)
- Team training on working with the reports
Scale
The whole plant, many systems
Quoted individually
after a call with the sales team · 8–12 weeks
A full enterprise-class production data architecture.
- The whole plant and multiple source systems
- Warehouse architecture / Microsoft Fabric
- Predictive models and anomaly detection
- Integrations with ERP / APS / CMMS
- Post-implementation support and report development
- Technical documentation and know-how handover
Common questions
The data questions that come up at the first meeting
- Where does the data sit — in the cloud or on our servers?
- What if our data is incomplete or messy?
- We have an old machine park with no network — can anything be done?
- Who sees which data?
Next step
Let’s start with a pilot on one area
Two to three weeks, one report on your data and a clear answer on whether it is worth going further.