The cloud’s role in data management

The connectivity of devices that use the Industrial Internet of Things (IIoT) must collect and analyze lots and lots of data.

By Rebecca Boll, General Manager, Sales, GE Automation & Controls June 15, 2017

Billions of sensors all around the world are making consumer devices smarter, safer and more efficient. Wearable products have revolutionized our understanding of wellness by providing more information about our daily health. Appliances and systems in our homes now are connected and make intelligent decisions to enhance our lives. Entire cities now are employing big data and sensors to prevent crime, manage crowds, reduce energy use, and much more.

The connectivity of and communication between these devices, as we know, is referred to as the Industrial Internet of Things (IIoT). But before these devices can start communicating with end users and each other, they must collect and analyze data—lots and lots of data.

Data for many of these devices is collected, stored, and analyzed in cloud-based environments such as the Amazons and Googles of the world. Even our Fitbit trackers have a cloud where our health analytics are stored. While there’s a lot of data gathered from consumer products, there’s a much larger amount of data found in industrial machines and an even greater need for the same type of connectivity and analytics. IIoT will increase efficiencies, reduce costs, and improve employee health and safety while generating new revenue streams for businesses.

Controls have been managing machines all around us for centuries. Just as our machines have evolved, so, too, have the analytics our machines provide. However, the data those machines collect and analyze, up to this point, has been limited to the memory of internal computers running at or inside the machine; the intelligence of those machines essentially has been limited to that memory.

New industrial control systems now are providing the same type of connectivity, communication, and analytics for our industrial assets as consumers have seen for several years. These connected controllers and systems not only run our machines like they used to run, but they also now have the capability to create an outer loop and custom applications that gather data in real time and advise the controller to make more intelligent decisions. Just as connected sensors and wireless devices transport data to consumer clouds, the vehicles to transport big data to the industrial cloud are edge devices and sensors in or near machines.

Many machines and plant systems don’t offer the type of connectivity and deep analytics today’s industry demands. These machines require third-party solutions to gather data and provide critical analytics. The more data gathered, and the closer to the machine you run data, the more accurate and intelligent the decisions the machine makes.

Some machines can’t be reached with traditional Ethernet cables because of the distance between the internet source and the machine. Not only would it be cost prohibitive to drop and run that wired connection, but the proximity would make data gathered less favorable. Connecting to these machines wirelessly makes sense.

Cloud-based, wireless solutions provide greater flexibility than wired solutions. The cloud empowers the elasticity of data storage and performance and is not subject to the constraints typical of a local computer. Connecting to the cloud also enables a fleet of machines to be managed and to communicate with each other, providing thorough analytics for the entire fleet. For example, there’s an app that leverages the power of the industrial internet, connects to a fleet of wind turbines and enables them to coordinate the speed and direction of their blades to maximize the wind farm’s overall energy output.

However, the cloud is not always a viable option due to latency, costs, regulations, and other factors. Think also about the criticality of the data being analyzed. If a self-driving car was required to go to the cloud to make important decisions, the car’s response time would be too slow. Industrial assets are the same way. Some things can be done in the cloud, but other problems need to be analyzed as close to the machine as possible. Digital industrial companies are developing products that are designed to work on the industrial edge and make collecting, storing, and analyzing data easier and more accurate.

Three years ago, we couldn’t have imagined IIoT as it is today. Similarly, we can’t speculate on where it will be three years from now. The rate at which things are changing makes it difficult to recommend specific assets or systems that should be hard-wired or should be wireless or which apps should run in the cloud versus at the industrial edge, so it’s important to think the process through strategically and go the route that makes the most sense for your plant.

Rebecca Boll is general manager, sales for GE Automation and Controls.