A self-driving car braking for a pedestrian doesn't have time to wait for a cloud server's reply. That's where fog computing comes in — a processing layer sitting between your devices and the cloud, built for moments when milliseconds actually matter.
What It Is
Cisco coined the term in 2012 — fittingly, since fog is just cloud computing closer to the ground. Instead of sending everything to a distant data center, fog computing processes data at nearby nodes (routers, gateways, local servers), sending only results upward when deeper analysis is needed.
Why It Exists
Cloud-only systems hit a wall with billions of IoT devices streaming constant data — latency breaks real-time apps, bandwidth chokes under the volume, and everything fails if connectivity drops. Fog fixes this by keeping decisions local.
Fog vs. Edge
People use these interchangeably, but they're different. Edge computing means the device itself does the processing. Fog sits one layer back, coordinating multiple devices at once — think a floor manager overseeing several desks before escalating to head office (the cloud).
Where It's Used
Smart traffic lights reacting instantly to local sensors, wearables flagging irregular heartbeats without cloud delay, factory floors catching equipment faults before costly downtime, and remote oil rigs staying operational despite spotty internet.
The Catch
More distributed nodes mean more complexity, more cost, and more points to secure individually. Standardization across vendors is still a work in progress too.
The real story isn't fog replacing the cloud — it's a layered system where cloud handles heavy long-term computation, fog handles near-real-time coordination, and edge handles instant, device-level reactions. Fog is what makes the cloud actually workable for a world that can't afford to wait.
Read full article here: https://www.weejix.com/technology-articles/fog-computing
Top comments (0)