Although cloud computing is pervasive and popular in consumer technology, the cloud isn’t always ideal in industrial applications because factory automation needs real-time connectivity without latency or intermittent-connectivity issues, as well as industrial-grade
equipment that can handle the factory environment.
The next generation of computing--fog and mist--and the applications for factory automation and embedded devices are just now being realized. On the hardware side, the latest processers are small-form-factor systems that enable industrial applications.
By bringing connected hardware, data processing, data analytics, and data storage closer to the edge with nearer physical proximity, fog and mist computing address many cloud-computing issues. Fog computing brings that proximity to the factory floor (close to the machine) while mist computing brings it right onto or into the machine. IDC predicts that by 2019, 45% of IoT-created data will be stored, processed, analyzed, and acted upon close to/at the edge of the network.
Fog computing
Fog computing locally addresses the needs of factory connectivity by offering distributed data-and-control resources while increasing efficiency and reliability. Making use of new software-designed automation elements, such as software-PLC controllers and digitization of equipment/processes, is necessary to address the fragmented state of communication and control at the lowest hardware levels.
Mist computing
Mist computing takes the idea of fog computing one step further by integrating data storage, processing, and analytics at the hardware level. There may also be security benefits to mist computing, since the farther the data travels, the more susceptible it is to an interception. While this idea has yet to be proven in real-world testing, the more local you can secure data, the safer it is going to be, which makes mist computing an attractive option for increasing data security.
Hardware requirements
The hardware necessary for fog and mist computing at factories must be industrial-grade, since it is not being maintained in a controlled environment. With wide operating temperatures or a wet environment, looking for a Microsoft Azure or similar certification provides confidence that the hardware will be able to perform in rough conditions.
Additionally, space is an issue, with factory floors requiring smaller devices that can embed into or nearby machines. With cloud computing, the hardware elements are designed around server farms and large control centers intended to store and analyze massive amounts of data for management, control, and monitoring of the enterprise, down to the factory floor.
Fog and mist computing require much smaller embedded form factors. Beyond providing the equipment control and monitoring function, this hardware must also support fog/mist computing sharing of resources.
Especially with mist computing, the physical integration onto and into machinery is not a simple technology change for hardware vendors. In addition to the size and ruggedness, the product lifetime of mist-computing hardware must support the overall quality and product lifetime of the machinery it is controlling.
With fog and mist computing, adding this hierarchy of elements requires greater interoperability. The OpenFog consortium was founded by industrial giants like Dell, Intel, and Microsoft to address these challenges. The goal is to create a framework for efficient and reliable networks and intelligent endpoints, combined with identifiable, secure, and privacy-friendly information flows among clouds, endpoints and services based on open-standard technologies.
We're still learning how fog and mist computing can best be realized in industrial environments. What we do know is that the speed of localization of data analytics can create far-reaching benefits in factory automation and other industries where it is vital to have quick decision-making. Watch for new, small-form-factor embedded CPUs that will bring full-featured compact CPUs and industrial-embedded PC products to enable fog and mist computing for factory automation.
