« Back to Glossary Index
Edge computing means running data processing on a device at the data source rather than sending raw readings to a remote server. A factory gateway that filters 1,000 sensor readings per minute and only forwards the five that exceeded a threshold is doing edge computing.
The practical benefits are lower bandwidth costs (send results, not raw data), lower latency (milliseconds locally vs hundreds of milliseconds to the cloud), and continued operation when the WAN link fails. Edge platforms vary by capability:
- Node-RED – visual flow editor, runs on industrial routers and gateways, no coding required
- Python scripting – available on Linux-based IoT gateways for custom analytics
- Containerised apps – Docker containers on industrial computers for complex edge workloads
- Local buffering – store-and-forward during connectivity loss, retransmit when link recovers