Internet do Futuro 2026.2 - Edge e Fog Computing - Seminário 1

Summary

Daniel Dionísio explains how edge, fog, and cloud computing work together, focusing on whether moving computation closer to users reduces latency and what reliability trade-offs it introduces.

• Cloud computing provides abundant processing and storage resources, but its distance from users can create latency. Sending time-sensitive data to the cloud and waiting for a response may be unsuitable for real-time applications.
• Edge computing processes data locally, at or very near the device. It offers the lowest latency and can continue operating despite network interruptions, but edge devices have limited resources and storage.
• Fog computing sits between edge devices and the cloud. It provides intermediate latency and resources, processing data closer to users than the cloud without being as local as the edge.

The autonomous-vehicle example illustrates the need for local processing: a car traveling at 100 km/h covers about 28 metres per second. The speaker’s example appears to refer to a response delay of roughly 100 milliseconds—the subtitle’s “100 minutes” is likely an automatic-transcription error—which would mean the car travels nearly 3 metres before reacting. Braking decisions therefore should not depend solely on a remote cloud response.

The speaker also notes that autonomous vehicles can generate data at rates of up to 2 GB per second. Filtering or processing some of it locally can reduce network traffic and avoid sending all raw data to the cloud. In the example, vehicle systems handle immediate actions, roadside or nearby infrastructure can provide warnings to other vehicles, and the cloud supports heavier tasks such as mapping, analysis, and model training.

A referenced comparative study is described as finding lower average latency for edge and fog computing than for cloud computing, though the study’s title is unclear in the subtitles. Technologies mentioned include Kubernetes, IoT, Cloud Edge, and Kafka; the subtitles’ transcription of some technology names may be imperfect.

Conclusion

Edge and fog computing can reduce latency, especially for applications requiring fast responses. However, distributing computation creates more points that may fail and makes monitoring, troubleshooting, redundancy, and orchestration more important. The three layers are presented as complementary rather than competing.

Reviews, Guides, or Tutorials

None; this is an explanatory seminar and analysis.

Main Speaker/Source

Daniel Dionísio.