Life—and Data Analytics—on the Network’s Edge
The Internet of Things (IoT) is growing fast, and that growth will only accelerate in the coming years. Some estimates have as many as 30 billion connected devices worldwide by 2020, according to IDC, with IoT spending to hit $1.29 trillion by that year. And all those connections — among devices, systems, sensors, humans, services and business processes — will help generate zettabytes of traffic that will flow into data centers within the next three years.
All this is driving the need to bring more compute capabilities — more processing power, more storage, more analytics — closer to the devices and systems that are creating all this data. Capacity at the edge of the network is going to become even more crucial as the demand for more storage, processing power, networking and analytics grows along with the amount of data being generated. At Dell EMC, we’ve been supporting our customers’ edge computing initiatives for more than two years, and other vendors are now following suit.
The IoT promises businesses a level of insight and knowledge that has not been seen before. Data is the key. The data being generated from the various distributed, connected and increasingly intelligent devices and sensors can lead to better and faster business decisions, improved efficiencies and reduced costs. To save network bandwidth, speed up the analytics and decision-making process, and make the data more secure, these capabilities are going to have to move closer to the network edge, where the systems and applications themselves reside.
The power of real-time data analytics
Rather than having to route the data from the edge endpoints back to central data centers for processing, storage and analyzing, the data can be managed and analyzed in real time at the edge. Think of an oil rig in the middle of the North Atlantic, generating massive amounts of data. Transmitting the data from the oil rig to a central data center would be costly and pose a high security risk. But analyzing the data in real-time not only means faster, better business decisions, but also the ability to determine what data needs to be sent back to the data center and which data is relevant and interesting, and which is not. Network bandwidth is saved by reducing the amount of data sent to the data center or public cloud, and the amount of data that needs to be stored is also reduced.
Growing numbers of enterprises and SMBs already are beginning to use intelligence devices and sensors to fuel better business decisions and reduce costs. Dell EMC partner Action Point has a customer in manufacturing that uses boxes in its production processes that include sensors that generate 19 points of data, from temperature to the amount and impact of shaking. In all, the sensors generate 3,000 data points a minute.
The Weir Group, a huge manufacturer of mining equipment that is often use in such remote areas as the Rockies and mountainous regions of Peru, is using sensors, IoT gateways and analytics at the network edge to enhance the predictive maintenance and asset management of its machines. Where once technicians had to travel to these areas to inspect equipment and find and fix problems — a costly and time-consuming effort — the company can now collect and analyze the data at the edge and address issues before they become problems. In cases where people do have to travel to the remote sites, they can have the right equipment already in hand, thanks to the insights gained through edge analytics. All of this means the company can keep its 400,000 assets up and running more easily, while reducing maintenance and management costs.
Ramping up artificial intelligence at the network edge
The drive for more edge computing will only ramp this year. Connected systems are becoming more intelligent and more are coming online, all the while generating more data. Companies will increasingly leverage artificial intelligence (AI) to drive greater automation in the managing and analyzing of data at the edge and to bring more intelligence to applications. More sensors and more computing power at the edge will also drive the need for more AI capabilities.
The growing number of communications protocols — everything from ZigBee to 6LoWPAN to the upcoming 5G — will drive lower latency and faster speeds as they give businesses more and better options for bringing more capacity and analytics capabilities to the network edge.
In addition, there is a push to develop more technologies to address the growing demand to store more data at the edge. IoT systems are creating massive amounts of data — and a growing amount of that data is video — but not all the data needs to be stored or analyzed. Some of it isn’t useful or relevant. So what is needed is technology that analyzes the data and sorts out what is important and what isn’t. For example, V5 Systems, a Dell EMC partner, offers a self-supported video surveillance system that includes algorithms that can sort through all the video taken, keeping what’s relevant and deleting what isn’t.
Much of the time, nothing happens in the data that is recorded by surveillance equipment. V5’s technology can capture data, analyze it and, in real time, determine when something out of the ordinary is happening — a child is alone on the street or a man with a gun appears — enabling users to respond immediately. By understanding what is important and what isn’t, such technology can save companies from having to save and store useless data out on the edge, thereby sparing the expense of sending and storing it in a cloud.
The need for edge computing capabilities will only grow along with the IoT. The Internet of Things requires a technology platform that can stretch from the data center and the cloud out to the edge, and users must find a vendor that can offer such a broad portfolio of products.
Andy Rhodes is the vice president and general manager for IoT at Dell EMC.