Glossary · IIoT, data and AI
Edge computing
Also known as: Edge
German: Edge Computing
In industrial IT, edge computing is the processing of data on or near the machine or plant where it is generated, rather than in a central data center or cloud. It reduces latency and bandwidth needs and allows functions to continue when the network connection is lost.
- IIoT
In one sentence
Edge computing processes data on or near the machine where it is generated, reducing latency, bandwidth and dependence on the cloud.
Example
An edge device next to a press reduces 10 kHz vibration signals to a few characteristic values per minute before sending them to the cloud.
How it applies
- Engineering: Edge computing suits preprocessing, local analytics, protocol conversion and AI inference close to the process. Hard real-time control remains in the controller.
- Operation: Many distributed edge nodes have to be managed: software updates, certificates, monitoring and backups. Plan central device management from the start (Patch management).
- Documentation: Describe which applications run on the edge, which data they receive and send, and how the system behaves when the edge node or its uplink fails.
Edge vs. fog vs. cloud
Edge processing happens on or next to the machine. Fog computing adds an intermediate layer between edge and cloud, such as a line or plant server. Cloud computing runs in remote data centers. The boundaries are fluid and vendors use the terms differently.