Edge Computing and the Promise of Predictive Plants

October 16, 2018 • John Crupi

In a recent survey of 250 IT and operations executives, 80 percent of the respondents affirmed that the Industrial Internet of Things (IIoT) can and will transform both their companies and their markets. Industries represented in the survey includes manufacturing, utilities, power and energy, transportation, aviation, and aerospace. And they’re not alone.

The idea of integrating existing plant information technology (IT) with increasingly connected and “intelligent” physical assets running in the environment is the foundation of the Industry 4.0 movement. IIoT may be the biggest driver of industrial productivity and growth in the foreseeable future. Management consulting firm Accenture estimates that IIoT could add $14.2 trillion to the global economy by 2030.

Increased automation to reduce operational risks and control costs is an obvious aim, but IIoT’s shifting focus to “edge” computing introduces a higher order of asset performance and production augmentation capabilities.

Broadly, edge computing involves technology that facilitates data processing at or near the source of data generation. The sources of data generation in industrial or manufacturing plant environments usually take the form of equipment and assets embedded with IIoT sensors or devices. With traditional IIoT, those “things” collect and transfer data in batches to a data center or the cloud for processing, but that model is changing. There are a couple of very good reasons why the paradigm is shifting to lean more heavily on edge computing:

  • Cost: With billions of connected things already deployed, data production vastly outpaces bandwidth availability, and much of this data on its own isn’t worth the price of transmission.
  • Time: It takes time (and network availability) to send information to the cloud and/or receive information from the cloud, and this latency is an issue in advanced automation. Things like real-time response in semi-autonomous vehicles and robots or safety-critical triggers cannot function as required at cloud speed.

Edge-based IIoT solutions address these concerns. With edge computing, “smart” equipment performs analytics on most of its own data in real-time or does so in collaboration with nearby gateways. Only the product of edge analytics (small volumes of pertinent or high-value data) is relayed to the cloud for storage or further analysis at convenient intervals, thus greatly reducing the magnitude of data being transmitted and/or stored and all associated costs.

 

Edge solutions can also automatically and near instantaneously compare local analytics results against historical conditions and performance metrics, allowing in-the-moment status reporting on the function and condition of particular assets, groups of assets, and plant facilities as a whole, depending on the deployment. This capability also accommodates automatic operational triggers based on set parameters (motion or temperature or pressure, for example), thus enabling real-time responsiveness.

But the biggest promise of edge solutions is delivered through the ability to perform predictive maintenance, offering insight into what is likely to happen to a connected asset in the future.

Predictive edge analytics forecast whether a machine is likely to function optimally, falter, or fail based on accumulated asset operation metrics, the particular item’s historical state, and incoming real-time data. That type of digital clairvoyance translates into real savings. Unplanned downtime costs industrial manufacturers an estimated $50 billion annually. Equipment failure is the cause of 42 percent of this toll. The insight gleaned from predictive analytics can ease that pain significantly. For example, a 2016 pilotimplementation of predictive capabilities for one asset class (industrial extruders) at a chemical manufacturing plant resulted in an 80 percent reduction of unplanned downtime and cost savings of around $300,000 per asset!

The types of edge analytics solutions being deployed grow more creative every day. Management consulting firm McKinsey identifies a couple of the most prized analytics capabilities in industrial settings that can be combined to deliver EBITDA margin improvements of 4 to 10 percent:

  • “Yield-energy-throughput (YET) analytics can be used to ensure that individual machines are as efficient as possible when they are operating, helping to increase their yields and throughput and reduce the amount of energy they consume.”
  • “Profit-per-hour (PPH) maximization analytics [can scrutinize] the thousands of parameters and conditions that have an impact on the total profitability of an integrated supply chain (from raw materials purchasing to final sales), providing intelligence on how best to capitalize on given conditions.”

The IIoT is the way of the future for manufacturing and plant operations, and edge computing capabilities are the way of the future for the IIoT. Real-time, cost-effective predictive prowess will dramatically reduce overhead and optimize operations in ways we can only begin to imagine.

John Crupi is the Vice President of Edge Analytics at Greenwave Systems.