Processing near the data source cuts latency and bandwidth.
Edge nodes process data locally to cut latency, bandwidth and cloud cost and to keep working offline; they also improve privacy. Tasks: filtering, anomaly detection, inference, protocol translation. Orchestration of containers and updates is a challenge.
A camera detects people on-device and sends only counts.
Edge computing processes data close to the source, which saves bandwidth and reduces delay for time-sensitive decisions.
Processing near the data source cuts latency and bandwidth.
Edge computing processes data close to the source, which saves bandwidth and reduces delay for time-sensitive decisions.
Edge processing reduces latency, network use and dependence on connectivity, and it can keep private data local.
Edge devices have less computing power and storage than the cloud, so complex models and long history may still need central systems.
Typical applications include detecting a fault on a motor locally rather than sending raw vibration data to the cloud, running a camera-based safety check on a factory robot and filtering sensor data in a gateway before it is uploaded.
Study "MQTT" first (topic 87 of 104 in Electronics & Embedded). It covers the groundwork this topic builds on.
Fog computing. Latency in ms. Resource-constrained hardware.
A camera detects people on-device and sends only counts.
The next topic is "TinyML", which builds directly on this one.
It is classed as Advanced in Electronics & Embedded (topic 88 of 104). It is an advanced topic. It assumes you are comfortable with the intermediate topics in Electronics & Embedded.
Want to practise? Open the interactive course for quick checks, quizzes, simulations and progress tracking.