Artificial IntelligenceBy Mike Schmidt, AEM Industry Advisor Editor

The combination of emerging technologies and evolving social forces are creating new customer experiences, disrupting tried and true models for success and leaving the equipment manufacturing industry to deal with ever-changing demands from end users.

With transformative change happening everywhere – and every day – how can manufacturers expect to get a handle on a cutting-edge technology like artificial intelligence (AI), where it’s headed, and how it can be leveraged to drive efficiency and operational improvement?

“It’s critically important to look at artificial intelligence and AI processes not as AI on a standalone basis, but rather as part of an integrated whole. You really only stand to benefit if you take a holistic view,” said Quentin Samelson, senior managing consultant for IBM Global Services, who spoke at AEM’s Thinking Forward event held late last year at the Milwaukee School of Engineering (MSOE) in Milwaukee, Wisconsin.  

AI Defined

IBM defines artificial intelligence as "the general concept that machines can be 'taught' to mimic human decision-making and learning behaviors. And while AI may seem like a futuristic technology, the idea behind it has actually been around for decades. That's because, said Samelson, when it’s applied to operations, AI is nothing more than the end result of training computers to recognize and act on patterns.

“It’s really that simple, but it can be really, really powerful,” he added.

Consider IBM's supercomputer Watson, which found its way into the cultural mainstream in February of 2011 when it competed on the television quiz show Jeopardy. While stakeholders within IBM sought to have Watson go toe-to-toe and hold its own against a pair of the show’s most successful contestants of all time, they had a larger, more impactful goal in mind: develop a new piece of technology capable of finding unstructured data more effectively than ever before.

Using software called Deep QA, Watson is essentially capable of understanding and interacting with human language. And because it runs on several of what’s known as Power 750 computers – or 10 racks holding 90 servers, for a total of 2,880 processor cores powering both software and storage – it can hold approximately 1 million books of information.

IBM researchers spent years “feeding” Watson information, and in doing so, prepared the supercomputer to make an indelible mark on the world by battling human game show contestants…

According to Samelson, one key to understanding Watson or any other AI system is to recognize it’s not just one “computer,” but rather a collection of programs capable of addressing a wide range of tasks.

“Some AI programs are really good at ingesting text, and many early ones aimed to do just that,” said Samelson. “Over time, the technology has grown so it’s capable of doing things more easily and can be trained much more easily. We’ve also found we can train AI programs to (ingest) visual information and acoustic information, then decide what we want them to do – whether it’s look for differences, patterns, or even something that doesn’t belong.”

AI Applied

A second key to understanding AI is to recognize it's not a means to replace humans, but rather a means to augment humans. No matter how it's applied – plant floor processes, risk mitigation, logistics, supply chain management – AI can assist in helping manufacturers avoid costly equipment downtime and maximize uptime.

“It all starts with gathering data, looking for patterns in that data, and determining how analytics can be applied,” said Samelson. “And once you have the pattern and know what to do with it, you can determine what the program should actually accomplish.”

The benefits of investing in and leveraging artificial intelligence for manufacturers include:

  • Improved process capability
  • Enhanced performance
  • Increased opportunities for more and better analysis
  • More and better opportunities to leverage automation

For example, said Charla Stracener, electronics industry lead for IBM Watson Cognitive Computing, manufacturers now possess the ability to inspect products as they travel down the manufacturing line, identify if certain parts are good or bad, and determine why they are good or bad. In addition, manufacturers can now accurately assess their level of confidence in their insights and even train their models for improvement over time.

“Unplanned downtime, as we all know, is very, very expensive,” said Stracener. “So being able to quickly return to the last good state is very important, and manufacturers can now do that in nanoseconds.”

AI Understood

Perhaps the single most important things for manufacturers to understand about AI is that it’s not a standalone, magical fix for what ails an organization. It’s in taking a holistic view of how AI and other emerging technologies work in concert with one another to drive operational improvements that manufacturers can learn how to overcome disruption and – more importantly – better meet the evolving needs and wants of their customers.

The rise of Industry 4.0 has led to the convergence of AI with a number of technologies – blockchain, the Internet of Things (IoT), Edge, 5G and more. Organizations of all types and sizes are leveraging them and seeing incredible gains in efficiency and productivity.

“Smart factories are the keystone of Industry 4.0, and it’s represented by an aggregate of all of the technologies,” said Stracener.

All of this technology is working together to facilitate human-robot interaction, positively impact product design and development, optimize supply chain management and much more.

The increased prevalence of artificial intelligence in manufacturing is transforming activities that once seemed difficult – or even impossible – into reality. And it’s those willing to invest the time, effort and resources into leveraging AI with other cutting-edge technologies who stand to gain the most.

“AI is not magic,” said Samelson. “We’re in the fourth iteration of the Industrial Revolution, which began centuries ago. Even back then, people were nervous about change. But it didn't – and can't – prevent them from redefining themselves.”

Looking for more information about the latest trends and technologies impacting the manufacturing industry. Visit www.aem.org/think and subscribe to the AEM Industry Advisor

×