We start with a basic fog question: When and where should we use fog computing in our network?
The basic premise of fog computing is decentralization of data processing as some processing and storage functions are better performed locally instead of sending data all the way from the sensor to the cloud and back again to a control mechanism. This reduces latency and can improve response times for critical system functions, saving money and time. Fog computing also strives to enable local resource pooling to make the most of what’s available at a given location.
I believe the opportunity for this kind of distributed intelligence and the associated intelligent gateways needed for fog computing are the strongest when these two conditions are met:
1. The focus of data analytics is at the aggregation level so the closer the better; and
2. There is a complex degree of protocol complexity where doing it locally actually makes more sense.
Markets that have these needs include manufacturing, extraction industries (energy, as an example), and healthcare. Applications such as smart metering can benefit from real-time analytics of aggregated data that can optimize the usage of resources such as electricity, gas, and water. Local level analytics is suited for those applications that require the data to be stored and analyzed locally due to either regulatory reasons or because the cost of transportation of the data upstream and the associated wait-time for analysis is prohibitive, such as airline maintenance data.
One major network bandwidth issue for IoT in the coming years is subsidiarity, making sure that the data analysis is done at the appropriate level to the speed and efficiency demanded for the application. In most cases, there will be a blend of approaches and the functionality to manage local as well as central application management will be increasingly critical to data analysis speed and functionality.
Use cases for fog computing and IoT
Good use cases for fog computing will be ones that require intelligence near the edge where ultra-low latency is critical. Some good case examples of fog computing usage in energy can be found in both home energy management (HEM) and microgrid-level energy management. HEM can use IoT to transform an existing home to a smart home by integrating various functionalities such as: temperature control, efficient lighting, and management of smart devices. A microgrid is a smart distribution device that can connect and disconnect from the grid to enable it to operate in both grid-connected or standalone mode.
My own personal interest is in connected building and smarter rooms in office buildings. Here there is a demonstrated need for edge intelligence and localized processing. A commercial building may contain thousands of sensors to measure various building operating parameters: temperature, keycard readers and parking space occupancy. Data from these sensors must be analyzed to see if actions are needed, such as triggering a fire alarm if smoke is sensed. Fog computing allows for autonomous local operations for optimized control function. This is useful for building automation, smarter cities, smarter hotels and more automated offices.
One good example of an architecture that has taken this into account can be found here with the Flextronics’ Smart Automation project. Another good example can be seen here in the Raiffeisenbank Romania headquarters, with redundant control systems for maximum reliability.
To conclude, there is a whole sub-layer of functionality where fog computing can quickly and autonomically assess control and develop edge intelligence within the enterprise. Our Industry 4.0 research continues to examine edge intelligence activities where central computing resources can still retain a viable role in the enterprise.