What Is Edge Computing?

What Is Edge Computing?

Edge computing means processing data at the point closest to where it is generated. In a factory, this point is a gateway, industrial PC or field panel next to the machine. Instead of moving all raw data to a central server, data is collected, filtered and summarised on the shop floor, and decisions are produced instantly when needed.

In brief

  • Edge = processing data close to its source; cloud = processing data centrally.
  • Key benefits: low latency, less bandwidth, uninterrupted data collection even if the connection drops.
  • In a factory, the edge device is usually a gateway or industrial PC that reads PLC/CNC/counter data.
  • Edge and cloud are not rivals; they are used together in a layered architecture.
  • Typical uses: cycle counting, stop detection, vibration/current anomalies, vision inspection.

How does edge computing work?

In the classic approach, every sensor reading is sent as-is to a central server. Think of a press’s current sensor producing dozens of values per second: moving all of this data creates both network and storage costs, and decisions are delayed.

In an edge architecture there is a layer in between. The field device does the following:

  • Collects: Reads data from PLCs, CNCs, counters, sensors and terminals.
  • Filters and summarises: Turns per-second raw data into meaningful events (for example “cycle complete”, “machine stopped”).
  • Makes local decisions: Triggers a light/siren when a threshold is exceeded and shows a warning to the operator.
  • Buffers: Stores data locally when the connection drops and transfers it when the connection returns.

Why does it matter in a factory?

Three constraints in the production environment make the edge approach almost mandatory:

  1. Latency: You need to catch a minor stop or a defective part within milliseconds. A request that goes to a central server and back does not work at this speed.
  2. Network reliability: Factory networks go down. In a system without an edge buffer, production data is permanently lost for the duration of the outage and the OEE calculation breaks.
  3. Data volume: When the raw data of hundreds of machines is moved, the network chokes. Summarising on the shop floor reduces the volume dramatically.

The difference between edge and cloud

CriterionEdgeCloud / central
Where processing happensNext to the machine/lineData centre
LatencyMillisecondsHundreds of milliseconds and above
When the connection dropsKeeps workingData flow stops
What it is good atInstant detection, counting, triggeringLong-term analysis, reporting, modelling
Storage capacityLimitedVery high

The right setup is usually layered: edge collects and summarises on the shop floor, while the MES and reporting layer does historical analysis centrally.

Typical use cases

  • Calculating cycle time and production count from counter pulses.
  • Catching the moment a machine stops, opening a downtime record and asking the operator for a reason code.
  • Detecting a deviation in vibration or motor current locally and generating a maintenance alert.
  • Digitising old machines through retrofit: placing a collector next to the machine without replacing it.
  • Getting data over a wireless IoT network without running cables to the panel.

Common mistakes

  • Trying to process everything on the shop floor: An edge device is not a report server. Historical analysis should be done centrally.
  • Not sizing the buffer: If an outage lasts 8 hours, how many hours of data can the device hold? This question should be asked at the start of the project.
  • Sending raw data as-is: Unsummarised data bloats both the network and the database.
  • Skipping OT security: An edge device is a bridge between the automation network and the IT network; access and segmentation rules must be clarified.

Frequently Asked Questions

What is the difference between edge computing and cloud computing?
In cloud computing, data is sent to central servers and processed there. In edge computing, processing is done on a device close to the shop floor where the data is generated. Edge provides low latency and low bandwidth; the cloud is suited to large-scale data storage and heavy analysis. The two are usually used together.
What is used as an edge device in a factory?
Usually an industrial PC, a gateway, a smart sensor concentrator or a small server placed in a field panel. This device reads data from PLCs, CNCs and counters, filters it, stores it temporarily locally and forwards summarised data to upper systems.
Does edge computing work when the internet goes down?
Yes. This is one of the most important benefits of edge architecture. When the line or the internet drops, the edge device keeps collecting data and producing local decisions, keeps the collected data in its buffer and transfers it to the central system when the connection returns.

Edge Computing in EFFIX MES

EFFIX MES designs shop-floor data collection at the edge layer. With the Inolink wireless IoT network, counter and sensor data is obtained without running cables to the machines; for CNC and PLC connections, OPC UA, MQTT, Modbus, EtherNet/IP, MTConnect, Mitsubishi MC, Omron FINS and Euromap 63/77 protocols are supported. The collected data is turned into meaningful events and fed into production monitoring, downtime records and OEE calculations. The same data is evaluated by the Proactive Alert System (PAS) detectors and linked to an action such as a notification or maintenance work order when a threshold is exceeded. To discuss a collection architecture suited to your machine park, see the EFFIX MES page or request a demo.