Fog Computing: Connecting The Dots Through The Fog
In the beginning there was The Cloud.
And to begin with the cloud was able to meet our processing needs. But with an ever-growing network of IOT devices and increased demand for instantaneous, actionable insights, the cloud has started to struggle under the terabytes of data it’s being asked to process. Empowered Edge devices offloaded some of the processing towards 'The Edge', enabling some real-time processing to occur, but deeper analysis was still undertaken by the cloud. 'Fog Computing' is the next step; bringing the capabilities of the cloud closer to those Edge devices, allowing for deeper analysis to be conducted in real-time.
What is Fog computing or Fogging?
But what is the difference between Fog Computing (Fogging) and The Cloud, and where do Edge devices fit in?
Fog Computing is so called because unlike clouds, which float high above our heads, fog hovers much closer to the ground and is more dispersed. The idea behind Fog Computing is to physically move data processing from one centralised cloud to many Fog Nodes (data processors), which are closer to where the data is collected, The Edge. The main advantage is that this reduces the distance data has to travel along the network before it can be analysed. Although internet speeds have improved significantly over the years, the speeds and bandwidth is still not enough in certain applications for real-time analysis to be performed solely in the cloud. This is especially true when relying on cellular networks, even when taking the roll-out of 5G into consideration. To overcome this, these Fog Nodes can be placed strategically in remote areas where even a 2G signal is difficult, or in areas with a high concentration of data collectors, which would otherwise overload the mobile network if data was sent individually to the cloud. With the cloud, all data is funnelled towards one centralised location, so there is only one access point a hacker needs to infiltrate. Whereas, the added benefit with Fogging is that smaller batches of data are stored within many Fog Nodes. If a Fog Node is compromised, firstly, the actual data loss is significantly reduced, and secondly, that Node can be cut off from the rest of the Fog network and processing is rerouted to other Fog Nodes, reducing disruption.
Examples of Fog use in the future could be within Smart Cities and connected vehicles, or the Internet of Vehicles (IOV). Vehicles in the future will be made safer by being able to communicate with traffic lights, other vehicles to monitor speeds, direction and braking, and even street sensors to monitor pedestrian activity.
All of these will generate a lot of data which will need to be analysed in real-time for the car to be able to react quick enough to make autonomous, potentially lifesaving, decisions. The car can’t wait for all the data to be collected in the cloud, analysed and then for a course of action sent back to it. Even a delay of a second in that data exchange could cause an accident. Instead of communicating directly with the cloud, the car could communicate with the Fog Node for that neighbourhood or block, or potentially the car itself could become a Fog Node.
Another use for Fogging could be within manufacturing; with manufacturing becoming increasingly autonomous the demand for data analytics is only going to increase. A smart manufacturing facility could have several hundred data collectors within its four walls, from sensors embedded along the assembly line to forklifts & wearable. On top of the internal data collection there could also be data streams coming from outside the facility, providing up to date information on supplies, deliveries and order requirements. All of these need instantaneous data processing to ensure the facility operates smoothly, but also to stop potential accidents. For example - an autonomous forklift truck needs to know exactly where human workers are at all times. It needs to be aware that around the corner is a human who has stopped to do up his shoelace. If there is any sort of delay in the forklift receiving that information there could potentially be a life-threatening accident. So again, all this processing can’t be left to the cloud, so the facility would install several Fog Nodes to keep as much of the processing on-site as possible.
Clearing the fog
Fog Computing essentially moves critical data processing from The Cloud to where the data is collected, at The Edge of the network.
Edge Computing initially took on some of this processing, but Fogging builds on this and brings the cloud to The Edge device. So, combined they can replicate much of the processing and analysis completed in the cloud within spilt seconds, leaving all but the largest data processing to the cloud.
Much of The Edge network is often on the edge of mobile network reach and these environments tend to also be pretty inhospitable for consumer IT equipment. You need equipment which can survive extremes: temperature, drops, shocks, exposure to water and dust. It is also important to have a device with a long battery life, to be able to go multiple shifts without access to a power source.
The Panasonic TOUGHBOOK range of rugged devices have been the European leader for the last 17 years*.
*VDC research 2017