Industry analysis: unleashing the potential of the IIoT

Virtually every major company within the process industries is pushing for the mass adoption of industrial Internet of Things (IIoT) connectivity.

Because of this, the ‘industrial Internet of Things’ has become one of the most recognisable buzzwords in recent years.

In essence, the IIoT is made up of any ‘thing’ that can be assigned an IP address and provided with the capability to securely transfer data over a network and into the Cloud.

It is a concept that commonly finds itself situated alongside phrases such as ‘Big Data’ and ‘Industry 4.0’ – which essentially describes the discrete manufacturing subcomponent of the IIoT.

However, engineering companies would be forgiven for not fully understanding the IIoT, or knowing why they need to be ‘IIoT-ready’.

One project applying Industry 4.0 principles is the ‘Track and Trace testbed’, which uses IIoT architecture to give added intelligence to tools and shop floor systems to simplify production process

Andrew Minturn, Bosch Rexroth business development manager 

But they do need to be ready. At least, that is the cry coming from every company offering IIoT solutions.

Many of these solutions, which are designed to drive process efficiencies, as well as give plant managers greater clarity over the operational performance of production equipment, are available today, says Bosch Rexroth business development manager Andrew Minturn.

Minturn says Bosch is already realising and testing use cases for the IIoT in more than 100 pilot projects.

“One project applying Industry 4.0 principles is the ‘Track and Trace testbed’, which uses IIoT architecture to give added intelligence to tools and shop floor systems to simplify production process,” Minturn says.

“These ‘smart tools’ can communicate to a central database, locally with operators, or to other smart tools as needed, to provide operators with situational awareness, ensuring that production efficiencies can be achieved.”

However, IIoT only works if sensors, software and services are intertwined, Minturn adds.

“The sensors [are used] to collate data, the software for intelligent interpretation, and the services to present it to the end user in a clear and meaningful way, to make it useable.”

Head in the clouds

Meanwhile, Bosch has also recently launched its own Cloud enterprise for web-based services.

This, says Minturn, will allow the company to run various applications for connected mobility, connected industries and connected building businesses.

“The Bosch IoT Cloud means we now have all the relevant technical infrastructure to deliver our Industry 4.0 needs,” he says.

Clearly, having useable data in the Cloud is one of the cornerstones of connectivity via the IIoT.

Andrew Hird, vice president and general manager, Honeywell Process Solutions Digital Transformation, says getting process data in the Cloud in a secure way is allowing businesses to leverage the entirety of their expertise.

“Once all your data is in the Cloud, you are able to benchmark the performance of your assets against every one of the asset clauses you have,” Hird says. “That gives you the capability to do some seriously clever stuff.”

One example, Hird explains, is that it allows companies to start creating applications.

Once all your data is in the Cloud, you are able to benchmark the performance of your assets against every one of the asset clauses you have. That gives you the capability to do some seriously clever stuff

Andrew Hird, VP/general manager, Honeywell Digital Transformation

He says for the best possible results, app creation needs to be moved from “on premise deployment” and into the Cloud.

“This Cloud-based approach will allow businesses to run apps across their entire enterprise.”

For Hird, apps in the Cloud play a major role in IIoT connectivity. However, he envisages this to be taken several steps further as the IIoT is understood and utilised by more companies.

“We are currently turning the industrial automation and process industry into an app-store environment. We want to turn what was a product, like a distributed control system (DCS), into an app that users can download and try for a while.”

App-lying yourself

Unlike a traditional DCS, an app equivalent wouldn’t take years to install, meaning customers could simply delete the first app and download another one from a different vendor – potentially in a matter of minutes, Hird suggests.

He says this strategy is similar to downloading apps like you would on a smart phone. “

However, we are not yet at the same level of phone apps…but we are going that way.”

Hird also concedes that traditional automation manufacturers may not deliver the game-changing blow that would ultimately bring the IIoT to the masses.

He says technological disruption could come from any player in any type of industry.

“For example, companies like IBM or Google are all potential entrants into the industry. Likewise, the disruption could come from a traditional IT player, or a small and medium-sized enterprise (SME) that comes up with a killer app that puts traditional automation firms out of business.”

But before industry can maximise the potential of a killer IIoT app, it must overcome a number of other hurdles

The skills requirement is therefore huge – from circuit design and microcontroller programming, through to machine learning, cyber security and GPS development

Andrew Minturn, Bosch Rexroth business development manager 

Take skills, for instance. Bosch’s Minturn says in the past, mechanical, electrical and software engineers could work in isolation, but that has now changed.

“The disciplines have become merged as hardware and software become ever more intertwined. The skills requirement is therefore huge – from circuit design and microcontroller programming, through to machine learning, cyber security and GPS development.”

He says this need comes at a time of chronic skills shortages across the engineering industry – which is creating an even bigger challenge for industry.

Elsewhere, there is a reluctance for some companies to fully integrate their processes with IIoT capability.

Of course, the cost of installing and upgrading a plant cannot be overlooked.

Alexander Khaytin, chief operating officer at machine learning and data analytics specialist Yandex Data Factory, says one of the main hurdles limiting IIoT connectivity is the ability for companies to get a return on investment (ROI) from the data they already have available.

Khaytin says too much emphasis on ‘Big Data’ will yield little reward for industrial companies, which provides little incentive to continue with their efforts.

“Instead of talking about the new data sources appearing with IIoT through sensors, telemetry, and various smart devices, these companies should focus on value-adding applications that make use of this vast resource.”

This is where machine learning changes things, Khaytin says.

Learning objectives

Machine learning technology can automatically determine patterns in the datasets, and therefore identify relationships between processes and inputs, delivering accurate and usable predictions and prescriptions, he says.

“These machine learning algorithms also continually teach themselves based on the new information fed. In practical terms, this opens the possibility for manufacturers to further optimise existing processes, switching from rule-based systems reliant upon statistics to highly precise recommendations and predictions.”

These machine learning algorithms also continually teach themselves based on the new information fed. In practical terms, this opens the possibility for manufacturers to further optimise existing processes

Alexander Khaytin, chief operating officer at Yandex Data Factory

However, Honeywell’s Hird says that despite machine learning being pushed as the answer to a multitude of problems, it won’t teach a business anything they don’t already know – unless it has the right tools, the customers and the right people helping with the domain knowledge.

“If you have all three, then it becomes a fantastic solution because it will allow you to come up with something new and you will be able to use it to change things.”

He also points out that the IIoT is not about finding new problems.

“The IIoT is about solving business problems that have been around for a long time. It’s all about getting 1% more production or 1% lower cost. These are problems process companies have always had – but the IIoT can help solve them.”