Our website uses cookies to improve user experience, analyse website traffic and assist in our marketing efforts. By clicking “Accept”, you agree to the storing of cookies on your device. View our Privacy policy for more information. You can change your preferences at any time.
Orange cross to indicate close page icon.
male factory orker using a laptop to programme a machine

ERP/MRP shop-floor data collection: the hidden network requirements

ERP/MRP-driven shop-floor data collection fails most often because the underlying LAN/WLAN is treated like office connectivity: designed for coverage and basic access rather than reliability, segmentation, and performance under load and movement.

When scanners, tablets, line-side PCs, IoT sensors, and new applications start generating continuous operational traffic, intermittent issues become visible: sessions drop, data queues build, “systems feel slow”, and teams lose trust in the dashboards that were meant to improve performance.

TL;DR (what to do first)

  • Treat shop-floor connectivity as production infrastructure (availability and change control matter)
  • Define success in operational terms (downtime, scan success rate, latency tolerance, safety constraints)
  • Segment IT/OT and third-party access so faults and threats don’t spread laterally
  • Validate with real operational routes and workflows (not only static tests)

The trigger pattern we see

A common sequence:

  1. An ERP/MRP programme expands from offices into the shop floor (or across multiple sites).
  2. Data collection grows quickly: mobile devices, printers, machine data, quality systems, cameras, engineering laptops.
  3. The network “mostly works” — until it doesn’t. The cracks show up as intermittent faults that are hard to reproduce.

From an Ops perspective, this looks like: missed scans, delays, workarounds, and time lost. From an IT perspective, it looks like: escalating tickets, finger-pointing between Wi‑Fi/app/vendor teams, and no single source of truth.

Why “office-grade” networking breaks on the shop floor

Shop-floor data collection raises the bar in four ways:

1) Reliability beats raw speed

A fast network that fails intermittently is worse than a slower one that behaves predictably. Operational systems depend on consistency.

2) Movement changes everything

Roaming, handoff, reflections, metal, and occlusion create a different RF reality. Coverage maps aren’t the same as performance in motion.

3) IT/OT boundaries become real

Once ERP/MRP touches OT-adjacent systems, segmentation and controlled flows become mandatory — for security, resilience, and troubleshooting.

4) Visibility becomes a prerequisite

If you can’t see what’s happening end-to-end, faults will be blamed on “the system” and trust will erode.

A practical first-week diagnostic checklist

[ ]  List the operational workflows that must not fail (pick/scan/print, line-side updates, quality checks, maintenance access).
[ ]  Identify where movement is involved (routes, choke points, high-bay aisles, cranes, vehicles).
[ ]  Confirm what must be segmented (OT systems, vendors, tenants, guest, engineering).
[ ]  Check for single points of failure (core/distribution, controllers, uplinks, power).
[ ]  Validate with real routes and peak conditions (don’t rely only on static signal checks).

Common pitfalls

  • Designing for “coverage” rather than operational performance
  • Treating Wi‑Fi issues as “application issues” without correlating roaming/interference behaviour
  • Leaving east‑west access open because “it’s on the internal network”
  • Making changes without before/after validation evidence

FAQ (AEO)

Do we need to replace everything to make shop-floor data reliable?

Not always. Often the fastest gains come from measurement, targeted remediation, and prioritising change by operational impact.

Is this mainly a Wi‑Fi problem?

Sometimes, but the root cause is usually end-to-end: RF design, client behaviour, VLAN/policy, upstream capacity, and resilience.

What should we ask suppliers to prove?

Ask for evidence tied to your workflows: performance in motion, segmentation design, resilience behaviour, and before/after validation.

How do we avoid over-engineering?

Start with use cases and acceptance criteria, then build only the layers required to meet them reliably.

Next step

For an overview of the architecture patterns behind this, see: Connected Factory.

If you want to sanity-check readiness for your ERP/MRP shop-floor rollout, email solutions@oxspring.com with subject “ERP shop floor” and include your site type, critical workflows, and where issues are showing up.

If our blog post interests you and you’d like to find out more, please get in touch!
CONTACT US
Orange arrow icon for back to top link.