Manufacturing OT
Andon boards, scanners, machine data and dashboards that all report the same numbers.
Reliable shop floor data comes from capturing it where the work happens — scanners at receiving and picking, operation reporting at the machine, SCADA and MES signals from equipment — and posting it into ERP once, so andon boards, production reporting and Power BI dashboards all read the same source instead of separate spreadsheets that disagree by shift.
An andon board is only as good as the signal behind it. If the PLC cannot tell the system why it stopped, the board turns red and everyone stands around guessing. We design the data capture so the reason is logged when the line stops, not remembered ten minutes later.
The real value is not the display; it is the pattern. When the same downtime reason appears twice a week, or one line consistently underperforms, the data becomes a conversation that leads to a fix.
Scanners, labels, bins and Wi-Fi — the layer where inventory accuracy is either created or destroyed.
A working scanning workflow needs four things aligned: labels that encode the right data and survive the environment, handheld devices managed and charged, warehouse Wi-Fi that holds while the operator moves through aisles, and ERP bin and location design that matches how material physically flows. Weakness in any one of them shows up as inventory variance.
Handhelds get dropped, batteries degrade, and firmware drifts. A spare pool, an imaging standard and battery replacement on a schedule prevent the slow decline that ends with three of eight scanners working.
Move production counts, downtime reasons and scrap into Business Central and Power BI without exposing the control network.
Data flows one way from the SCADA or MES layer, through a broker or historian in the industrial DMZ, to ERP and Power BI. The control network stays isolated; ERP never polls a PLC directly. Production counts, downtime reasons, scrap and machine state are normalized, validated and posted at a cadence that matches the business process — not a manual spreadsheet at the end of the shift.
The biggest risk in integrating OT and ERP is giving the business system a path back into the control network. We design the integration so the data moves out, not in. If ERP needs to send a schedule or job order, it is staged and verified through a separate, controlled path.
Most manufacturers still have someone walking the floor with a clipboard, entering counts into ERP at the end of the shift. That data is late, error-prone and hard to reconcile. A governed data flow gives you the same numbers, continuously, with an audit trail.
Production, inventory and financial reporting that refreshes itself instead of being rebuilt every month.
Start with the numbers leadership already argues about: production output against plan, scrap and rework, on-time shipping, inventory accuracy and turns, purchasing lead times, and margin by product line. Build those from ERP data with consistent definitions and automatic refresh before adding anything more elaborate.
Two departments will define 'shipped', 'complete' and 'scrap' differently, and a dashboard makes that disagreement visible immediately. Settling definitions is the first deliverable.
Common questions
No. Many manufacturers start with a few andon boards and downtime reason capture, then expand as the data proves useful.
Yes. Industrial displays, sealed enclosures and proper mounting are standard for us. A consumer TV on a plant floor will fail.
Yes — device class, ruggedization, scan engine range and printer type, chosen for the environment rather than the catalog price.
Only alongside process discipline. Scanning removes transcription errors; it does not fix material moved without being recorded.
No. The SCADA or MES buffer is local. ERP receives summaries at a business cadence, not every PLC cycle.
That is normal. OPC UA and MQTT are vendor-neutral, and REST APIs from ERP can consume the data once it is normalized. We design the broker layer to be the bridge.
Yes, with row-level security or dedicated reports, and often on a wall-mounted display rather than a login-based dashboard.
It depends on the connection and refresh schedule. Scheduled refresh several times a day is typical; near-real-time is possible with the right architecture.
We review your network, security posture, cloud tenant and recovery plan, then hand you a plain-language report of what we found and what it means for the business.
