- Smart cities: Intelligent traffic lights that change based on traffic data
- Public transport: Trains that detect and notify anomalies for maintenance processes
- Intelligent networks: Machines that change their energy source based on cost and availability
- Agriculture: Devices that regulate crop irrigation by analyzing environmental data
- Construction: Sensors turn on the lights of a room when someone walks in
- This article by Ramya Mopidevi has more fog computing use cases across industries
Fog Computing or Edge Computing is so much more than the latest “hip term” in Silicon Valley.
It has nothing to do with misty graveyards or horror movies. It does have everything to do with the current and future viability of the Internet of Things (IoT).
Imagine you are traveling in a self-driving car at 75 mph and an obstacle appears out of nowhere. The car must make a split-second decision: pull the brakes, go forward, or avoid.
A few months ago a video emerged showing a similar situation. An autonomous Tesla predicts a highway collision two seconds before it actually happens. The action took place before the driver had time to react.
This requires hundreds of sensors in the vehicle that share data all the time. This is a ginormous amount of information. And this is only one car. Imagine the volume of data required for an entire fleet of self-driving vehicles.
Response times and connectivity are critical to the interaction between devices or machines. The term for this is Machine to Machine (M2M) communication.
And, as you can see in the video, one millisecond can make the difference between life and death.
Yes, the growth of IoT is nothing short of astounding. But infrastructure is proving to be a restriction for many applications.
On one hand IoT nodes are closer to the devices that generate data. But they don’t have the computing and storage capacity needed for M2M.
Cloud servers do have this capacity, but they are too far away to process this data and respond in time. Latency, bandwidth and uptime limitations also hinder data traffic.
Don’t Fear the Fog
The alternative is to process data closer to the source: the devices themselves. The fog exists at the point in which the cloud meets data generation.
Latency reduction is the greatest advantage of this architecture. Data transfer takes less time to start and instructions are executed faster.
Two factors are driving the development of fog computing solutions. Growing adoption in the industrial sector and increasing expenditure in smart cities. Some use cases include:
Cisco coined the term fog computing but IBM calls it edge computing. As in at the “edge of the cloud”. AT&T and Intel are also breaking ground in this promising new industry.
Gartner says IoT will include 26 billion connected units by 2020. To put that in perspective: the entire human population is 7.5 billion.
For more information on the fog you can visit OpenFog Consortium.org. This entity seeks to standardize the academic and industrial development of this technology.
Do you know any other fog computing applications? Share them in the comments section!