Industrial networking and IT/OT service

Industrial Ethernet and Factory Network Design

We map switches, cables and connected devices from machine cabinets to the plant backbone. For line renewals, we prepare backup paths, the IP/VLAN plan and the migration sequence according to the production schedule.

OT switchesRing / redundancyIP and VLAN planFiber backbonePost-implementation documents

Your situation

What problem does this service solve?

A network that has grown over the years and that nobody has a complete map of

Unmanaged switches added to cabinets, daisy-chained connections and undocumented cables lengthen the search for root causes during faults. Which machine is connected to which port is not recorded.

Office and production networks intertwined

If office computers, printers and PLCs share the same broadcast domain, a loop or broadcast storm can also cut off the line’s communication.

Lines that fail at a single point

Without a backup path in the backbone, a single cable, switch or power supply failure cuts off communication for several machines at once. During the survey, we list which uplinks, switches and power supplies are common points of failure for more than one machine.

Office-grade equipment used in field conditions

In cabinets with dust, vibration, heat and electromagnetic noise, office switches and unprotected cabling may not be enough for stable operation.

Technical framework

What we cover, and within what limits

Industrial Ethernet uses standard Ethernet (IEEE 802.3) technology at the physical and data link layers. The difference shows in the environment where the network is installed and in the traffic it carries. In a production network, time-critical control traffic such as PROFINET may share the same physical infrastructure with operator panel, camera and data collection traffic. The design is therefore driven by the following: the environmental class and mounting of switching equipment (DIN rail, wide temperature range, dual power), topology and redundancy protocol, traffic prioritisation, addressing discipline and the maintenance team’s ability to read the network. Industrial Ethernet is not a protocol. PROFINET, EtherNet/IP or Modbus TCP run on top of this infrastructure.

Technical scope

Current state and topology map

Inventory of switches, cabinets, cable routes, ports and connected devices; LLDP/SNMP data from managed devices and one-by-one identification on site for unmanaged ones.

Target topology and redundancy

Cell/line-based structure, backbone, ring or star. MRP and RSTP give different recovery times; the recovery time of the chosen protocol is compared with the application’s tolerance, and its support with the documentation of the existing devices.

IP address and VLAN plan

Control, HMI, engineering, data collection, camera and management networks are separated; naming and address blocks are documented.

Equipment and cabling specification

OT switch class, port type, fiber/copper choice, connector and cable category, in-cabinet layout and labelling rules; vendor-neutral technical criteria.

Phased migration plan

Change steps sequenced according to downtime windows; a rollback procedure and test list for each step.

Acceptance testing and handover

Connectivity, port error counter and management access tests. The backup path test (pulling a cable) is carried out in a planned verification window approved by the plant. Up-to-date topology and connection documents (as-built) are then handed over.

Choosing a topology: what to use when?

The topology is determined by the protocols the devices support, the need for redundancy and the maintenance team’s operating habits. The table is a general framework. The final decision is verified against the technical documentation of the devices on site.

StructureWhen it fitsWhat to watch out for
Star (cell switch)Few devices in the cell, easy fault isolationThe cell switch and its power supply are a common point of failure for all devices in the cell
Line (daisy-chain)Cable savings with integrated 2-port devicesRemoving one device in the chain also disconnects the devices after it
Ring + MRP / RSTPA backup path against cable breaks in the backbone or along the lineAll ring participants must support the same protocol. MRP and RSTP do not give the same recovery time; the time must be compared with the application’s tolerance
Fiber backboneLong distances, high EMI, connections between buildingsMatching module and fiber type (single-mode/multimode), splice and termination quality

Prerequisites: what we need to get started

We list missing information and access together in the first meeting and complete them during the survey.

  • A sketch or layout plan showing the line and cabinet layout (if available)
  • Existing switch models and management access (read-only is sufficient)
  • A list of PLCs, HMIs and drives; if possible, the network view from the TIA Portal / engineering project
  • Planned downtime and maintenance windows
  • The decision-makers on the automation and IT sides
  • IT’s IP addressing and security policies, if any

Deliverables

IP address and VLAN plan

Topology and per-cabinet connection diagrams

Device, port and cable inventory (matched to labels)

Equipment and cabling specification

Phased migration and rollback plan

Acceptance test record and as-built documentation

Sample scenario

Representative scenario: an assembly line that has grown increasingly complex

Imagine an assembly line daisy-chained through unmanaged switches in its cabinets and sharing the same broadcast domain as the office network. The first step is to compile an inventory from existing records; if port mirroring or a TAP is needed for traffic observation, its installation is planned with the plant team. The line is then split into control and data collection VLANs, a backup path is added to the backbone and a pilot is implemented in one cell. Success criteria are written down before work begins: port error counters do not increase during the monitored period, control connections recover within the time the application tolerates during the planned backup path test, and the relevant switch/port can be found from the records during a fault.

This scenario describes a typical project flow; it is not a completed customer reference.

Service summary

Related services

PROFINET and PROFIBUS Communication

We investigate intermittent dropouts between PLCs, remote I/O and drives using device diagnostic logs, port counters and the field topology.

View details →

network.eretek.com for office and campus networks (LAN/WAN) ↗

Scope limit

Field work such as cable pulling and cabinet installation is planned once the scope and occupational safety conditions are clear. We do not modify PLC application software. We assess the impact of network changes on the control software together with the automation lead.

Project approach

How do we proceed?

No step that touches production is carried out until the maintenance window, configuration backup and rollback plan are clear.

01 / STAGE

Discovery

Inventory from existing records, a site walk and identification of critical lines; an initial risk list.

02 / STAGE

Design

Target topology, IP/VLAN plan, redundancy and equipment specification.

03 / STAGE

Pilot

Implementation in one cell or line; measurements and operator feedback.

04 / STAGE

Rollout

A phased migration spread across downtime windows, with a rollback plan at each step.

STAGE 05

Handover

Acceptance tests, up-to-date post-implementation documents (as-built), training and maintenance recommendations.

Frequently asked questions

Questions about Industrial Ethernet and Factory Networks

01

What is the difference between an office switch and an industrial switch?

The basic Ethernet function is the same. The difference lies in environmental resistance, mounting, power redundancy, supported redundancy and industrial protocol features, and management/diagnostic capabilities. Not every cabinet needs an industrial switch; environmental conditions and how critical the line is decide.

02

Do we have to stop production while renewing the network?

Reviewing existing records and design work can usually be done while production continues. Physical changes affecting control connections are split across planned downtime windows, and each step is carried out with a rollback plan.

03

Is a ring topology always necessary?

No. A ring provides a backup path if a cable or device fails, but it requires configuration discipline. How critical the line is, device support and the maintenance team’s operating capacity are assessed together.

04

Do you work with a specific switch brand?

No. We evaluate industrial switches from different manufacturers, including Siemens SCALANCE, against technical criteria. We do not claim to be an official reseller or partner of any manufacturer.

Technical survey

Let's plan a technical survey for industrial Ethernet and factory networks

Briefly describe your line, your PLC and network equipment and your goal. We will respond with a written proposal covering scope, assumptions and deliverables.