Four edge computing use cases in industry

Considering the relationship between edge computing and iot, there will be a sub-category of iot - industrial iot. Today, the edge computing use cases of Iiot are maturing.

The industrial sector, which is usually a broad term for companies such as manufacturing and energy (such as heavy machinery manufacturing plants or power plants), is actually one step ahead on a marginal concept: industrial SCADA systems. In short, these are independent local control systems responsible for a variety of key industrial and other processes in the locality. These can therefore be seen as the forerunners of modern edge architectures.

"Industrial SCADA is a fringe form that has been around in one way or another for over 30 years," said Andrew Nelson, chief architect at Insight. Today, most facilities are equipped with some sort of isolation control system. In fact, they often have multiple such systems and processes, and edge computing deployments are increasingly likely to augment or even replace them."

Industrial environments themselves are essentially marginal locations, meaning they are often far from centralized data centers or cloud platforms. They are therefore suitable for increasing marginal adoption. Oil and gas RIGS in the ocean seem to fit anyone's definition of "marginal."

At this point, the industrial sector deals with inherently difficult sites: edge computing use cases overlap with other environments, such as warehousing or logistics, but often in harsher environments.

All of this makes the industrial sector a good marginal use case. So how do CIOs and other IT and business leaders in the industrial sector think about and implement edge infrastructure and applications?

First, Red Hat technology evangelist Gordon Haff provides some macro background: There are basically two mainstreams of industrial edge computing.

"On the one hand, sensor data is often filtered and aggregated from the operational/shop floor edge layer to the core; On the other hand, code, configuration, master data, and machine learning models flow from the core (where development and testing takes place) to the factory."

This has big implications for the fringe strategies of various industries. Edge to core is the need for IT leaders to decide what needs to actually live at the edge and what can or should be kept in a centralized cloud platform or data center.

"The idea is that you generally want to centralize management if you can, but keep it decentralized when you need to. For example, sensitive production data may not be allowed to leave the site, or running industrial processes may need to be protected from any disruption related to network issues outside the plant." The latter is an important part of SCADA connectivity because unexpected outages are not allowed in many industrial environments.

The core-to-edge process is primarily about operational soundness and efficiency. As with edge architecture in general, you can't expect IT professionals to fix IT every time you need to update a configuration or patch a system at an edge location. "In industry, there may be hundreds of factories running thousands of processes: automation and consistency are key," says Haff.

Red Hat technology evangelist Ishu Verma adds that core-to-edge is how enterprises extend the same practices and techniques they use to deploy applications on cloud platforms or in-house to their edge nodes, even in the harshest of industrial Settings.

"This approach allows enterprises to extend the best practices of emerging technologies to the edge -- microservices, GitOps, security, etc. This allows edge systems to be managed and operated using the same processes, tools, and resources as a centralized site or cloud platform."

Edge computing in manufacturing and energy
Within these two-way flows, here are four scenarios where edge computing is used in industry.

(1) Real-time simplified operation
Those traditional SCADA and other control systems are just as important as monolithic or legacy applications in many other areas, but are not particularly easy or flexible to use in modern environments.

"Traditional SCADA and control systems infrastructure tends to be closed and vendor-specific. Iot/edge deployment can help with real-time operations in a single control platform, rather than jumping between systems."

Monitoring and predictive maintenance is a good example in this category: Sensors and meters in a plant can be used for real-time operations and help industrial operators better plan when critical maintenance and other work is needed. This has been difficult in the past due to data silos, a common challenge for CIOs at many companies.

"Many industrial facilities will have multiple control systems that may or may not be integrated," Nelson said. Iot/Edge use cases can extract data across systems, correlate events and predict failures."

(2) Run AI/ML workloads in industrial sites
Reducing or eliminating latency is one of the main drivers of edge computing strategies. This is especially true for artificial intelligence and machine learning applications, as well as other automations that require data and large amounts of data to be effective.

There is huge AI/ML and automation potential in Iiot, but there are also huge data and latency implications.

"It takes a lot of data to make smart machines work seamlessly at the edges," said Brian Sathianathan, CTO of Iterate. Good AI requires data. Great AI requires a lot of data, and it needs to be available immediately."

This can be a problem in the first data flow scenario described above by Haff from Red Hat: sensor data flows from the edges to the core.

"I've seen it in manufacturing facilities where robots in the factory transmit 'too much' data through the local network and then all the way to the cloud and back," Sathianathan said. It's not ideal because, as manufacturing CIOs know, decisions have to be made immediately to be effective."

If delay is a problem, actual downtime is a killer -- especially in industrial Settings (where, for example, data outages or network problems can shut down gas pipelines) and related areas such as manufacturing.

While some downtime is usually acceptable in a standard IT environment, this is not the case in manufacturing. The cost of stopping a production line due to unstable edge applications can run into hundreds of thousands of dollars per minute, so there is simply no room for error.

Keeping the necessary data at the edge will be the driving factor to align edge computing with AI/ML use cases and minimize what Sathianathan describes as the "data overload" scenario.

(3) Improving energy management
Having edge applications that can automatically monitor and optimize energy consumption in industrial locations will not only improve productivity, but may also boost business revenue.

"There's a big push to monitor energy use and control loads in manufacturing and industrial applications," says Insight's Nelson. In industry, significant cost savings can be achieved simply by turning off or metering power loads at peak times."

In fact, rising energy consumption and costs in industrial organizations are such an issue that it is the subject of presentations and papers at the 2021 conference: Edge Computing Energy Management Systems for Industrial Facilities.

Cios and other IT leaders can certainly understand the revenue: designing an edge application that automatically adjusts and optimizes energy consumption in response to price fluctuations could be a real drive for use case.

"Reducing electricity costs has become an urgent issue, while remote monitoring of connected devices and intelligence pushed to the edge of monitoring devices have become critical in the Industrial Internet of Things," the report's authors wrote.

(4) Strengthen staff safety and on-site safety
A pattern will be seen here: Industrial Edge/iot use cases rely on a large number of sensors and other machines in these environments. But it's not just about machines, it's also about people. Industrial advantages also have important employee safety and on-site safety possibilities, Nelson said.

"Tracking employees and contractors and alerting them when they're not where they're supposed to be is a big deal for safety and security," Nelson said.

Like many edge applications, this is a category that typically involves or integrates with other technologies, such as AI/ML. It's also the kind of seemingly low-tech device (such as ubiquitous employee ID badges) that can be given a modern makeover.

"Computer vision, RFID, and BLE can all be used in this use case, and building a security badge reader and security camera is a useful integration," Nelson said.

Or try another universally recognized security product that predates edge, cloud and digital computing as it's known: the hard hat.

"They built hard hats with built-in sensors that can track these use cases through WiFi access points," Nelson said.

No comments

Related recommendation

No related articles!


Four edge computing use cases in industry