Unplanned downtime is a major burden for manufacturers, costing an estimated $50 billion a year and resulting in anywhere from a 5 to 20% drop in productivity.

In the past, predicting maintenance has been a game of chance. However, as smart technology becomes more accessible, manufacturing is entering the Fourth Industrial Revolution, AKA Industry 4.0.

“Now we’re connecting equipment and processes and getting feedback from it,” said Michael Ger (pictured) managing director of manufacturing and automotive at enterprise data cloud company Cloudera Inc. “That’s really what we’re talking about in the Fourth Industrial Revolution, and it is intrinsically connected to data and a data life cycle.”

Ger spoke with Dave Vellante, host of theCUBE, SiliconANGLE Media’s livestreaming studio, during the Cloudera “Transform Innovative Ideas Into Data-Driven Insights” event. In the second of two interviews focused on leveraging data insights in the manufacturing industry, Ger and Vellante discussed the importance of improving manufacturing uptime and quality, as well as how Cloudera is working with manufacturers to bring facilities and operations into the data-driven era. (* Disclosure below.)

Cloudera specializes in data life cycle management for manufacturers

Manufacturing Edge2AI is the name Cloudera gives to the connected manufacturing data life cycle, and it’s an area that Cloudera has focused on with its Data Platform.

“We’re not doing it just narrowly to implement technology. We’re doing it to move these drivers, improving quality, reducing downtime,” Ger stated.

The impact of implementing connected technology in the manufacturing pipeline is shown in internal research by Deloitte, which found a 10-20% increase equipment uptime, 20-50% reduced maintenance planning time, and 5-10% reduction in overall maintenance costs. That’s a lot of potential savings, and with sensor prices dropping dramatically and interconnectivity increasing, transitioning to smart manufacturing and collecting and analyzing data from manufacturing processes is now accessible to most companies.

The spectrum of use-case scenarios ranges from simple to complex

The implementation of technology for data collection and analysis can be done incrementally, according to Ger.

“There’s a spectrum of use cases, ranging from simple to complex, but you can get value even in the simple phases,” he said.

The easiest use cases to start with involve monitoring equipment and processes to streamline them and reduce expenses due to inefficiencies.

“No machine learning; we’re just talking about simple monitoring,” Ger stated.

Then companies can move onto a process called “quality event forensic analysis,” which involves tracking back to find the cause of recurring issues. As an example, Ger described a company that is experiencing a sudden increase in warranty claims.

“This is about connecting the dots. I’ve got warranty issues. What were the manufacturing conditions of the day that caused it?” he said.

Once this has been established, the company can identify exactly which other products were impacted by the same conditions. This allows a proactive recall of only those products rather than a blanket recall of an entire product line.

At the more complex end of the spectrum comes machine learning. But with that complexity comes an equal range of benefits. Describing “a whole slew of machine learning use cases,” Ger explained how models can be created to train visual sensors to identify quality issues in products as they are produced or track equipment maintenance and predict failure before it happens.

“Start with monitoring; move to machine learning. But at the end of the day, you’re improving quality and improving equipment uptime,” he stated.

The data life cycle explained

Each use case is unique, but the data life cycle follows a set pattern, which Ger described in detail.

“You’re going to ingest that data. You’re going to store it. You’re going to enrich it with enterprise data sources,” he said. “Now … you’re getting really nice data sets. They’re becoming very compatible with machine learning, so you bring these data sets together. You process that; you align your time-series data from your sensors to your timestamp data from your enterprise systems and your maintenance management system. Once you’ve done that, we can put a query layer on top. So, now we can start to do advanced analytics, query across all these different types of data sets.”

Once a long history of data is stored, it can be used to build out machine learning models that can be deployed out to the edge to work in inference mode. This is where the true value comes into play, as the models continuously monitor incoming data in real time and can predict failures or other potential problems and act proactively to avoid them. The most common example would be ordering parts and scheduling maintenance to avoid lengthy downtime during peak production.

The benefits of connected technology in the manufacturing process are undisputed, and the barrier to entry due to sensor costs has dropped. Yet, only about 10% of companies have made the leap into Industry 4.0. Over half of companies surveyed in a plant engineering maintenance study used in-house spreadsheets for maintenance, and 39% were still in the Dark Ages of paper recordkeeping.

The key challenge, according to Ger, is collecting both the information technology and operational technology data sets together in a single location to enable advanced analysis.

“The single hardest thing in this type of environment, connected manufacturing, is that operational technology has kind of run in its own world and … the silos abound,” he stated.

This is why Cloudera is stepping in to help manufacturing companies overcome this barrier and harness their valuable operational data to decrease downtime, improve productivity and reduce costs. Cloudera’s data platform was designed for digesting IT data, and thanks to the company’s robust partner ecosystem, operational technology data can now also be easily imported as well.

“Suddenly, you’ve got all the data you need to implement those types of Industry 4.0 analytics use cases,” Ger said. “It boils down to, can I get to that? Can I break down that IT/OT barrier that we’ve always had and bring together those data sets that we can move the needle in terms of improving manufacturing performance?”

Watch the complete video interview here (registration required), and be sure to check out more of SiliconANGLE’s and theCUBE’s coverage of the Cloudera “Transform Innovative Ideas Into Data-Driven Insights” event. (* Disclosure: TheCUBE is a paid media partner for the Cloudera “Transform Innovative Ideas Into Data-Driven Insights” event. Neither Cloudera Inc., the sponsor for theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: Michael Ger