Published on 07/29/2016 | Technology
This article originally appeared in Industrial IoT/Industrie 4.0 Viewpoints .
Industrial network gateways have rarely been a primary purchase consideration for OEMs or end users scoping out automation projects. Relied upon to bridge disparate network environments, or to bring legacy equipment into automation or enterprise architectures, gateways are often an afterthought in the quest to integrate operations in pursuit of business benefits.
But today’s gateways, whether Ethernet, wireless, or cellular-based, are harbingers of the changes coming to the industrial IoT network edge in the age of the industrial internet.
Originally relegated largely to network protocol conversion, today’s gateway products target sensor-to-cloud integration in furtherance of industrial internet-based strategies designed to improve business performance. This is leading to a broadening functional spectrum in the gateway market, one that ranges from its origins in protocol conversion for interfacing disparate networks to sensor-to-cloud integration and edge computing. It is also ushering in a broadening of the supply base with numerous new entrants and a host of new product introductions.
Why? Because the network edge is where the assets and the associated data to be analyzed reside, and the industrial landscape is rife with them. The combination of a large installed base of disconnected assets and long hardware replacement cycles is behind the need to incorporate legacy equipment to feed real-time data to monitoring, optimization, analytics and other applications. Gateways are a primary means for incorporating these legacy assets into the automation or enterprise architecture.
Edge devices are, and will be, increasingly called upon to serve data up to enterprise clouds and their associated business applications, such as analytics. This need to “feed the beasts” is driving the need for edge-to-cloud integration and migration of enterprise platform agents to the edge.
Edge devices like network gateways are assuming more IT-associated functionality due to the need to filter and massage data so that low capacity remote networks are not overloaded and higher-level platforms aren’t flooded with the tidal wave of data generated by industrial devices. Edge computing, fog computing, and similar strategies will be paramount to both prevent this data deluge and deliver feedback locally and as close to real-time as possible. Ability to execute applications on edge devices will therefore be increasingly required.
Many gateway suppliers have offered cloud integration for many years, often to their own private remote access and monitoring cloud, but even these suppliers must evaluate how best to meet the demand for making data available to enterprise applications. Reliance on standard microprocessors and operating systems that support tool-based application integration strategies, including support of REST APIs and MQTT, is a place to start. The Linux environment, which is supported in the Intel IoT platform solution, readily supports these tools and provides fast time-to-market. Windows 10 is also an option, one usually pursued by Microsoft partners and those aligned with their Azure cloud platform.
Beyond simple integration via a tool-based strategy, edge device suppliers must also assess their cloud strategy from the perspective of whether or not to support any of the cloud platform field agents migrating to the edge. Several suppliers, such as Dell and HPE, have already declared their support for Predix Machine in their IIoT gateway devices, while Cisco has allied with IBM and their Watson IoT Platform and attracted industrial partners such as Mazak as a result of its recent unbundling of the IOx platform. Further information on this trend is available in the separate ARC blog The Battle for the Industrial Network Edge.