How Artificial Intelligence Is Used In Manufacturing

AI in manufacturing is becoming more common than you might think.

Increasingly, technology plays a major role in how products get made on the factory floor. Manufacturing plants can resemble high-tech laboratories with robotic arms handling repetitive tasks and algorithms, ensuring that products are made according to manufacturer specifications.

Image source: The Motley Fool

Industrial titans like General Electric (GE 0.91%) and Siemens have embraced artificial intelligence as the technology offers a number of advantages, including minimizing defects and errors, reducing downtime, and saving on costs.

In this look at AI in the manufacturing industry, we'll discuss what artificial intelligence is, how it plays a role in manufacturing, and review several examples of how AI is used in manufacturing.

Machine Learning

Machine learning is a technique by which artificial intelligence programs are trained.

What is artificial intelligence? What is artificial intelligence?

Artificial intelligence is a technology that allows computers and machines to do tasks that normally require human intelligence.

This includes a wide range of functions, such as machine learning, which is a form of AI that is trained data to recognize images and patterns and draw conclusions based on the information presented.

There's also computer vision, which is widely used in manufacturing; natural language processing, which includes tools like ChatGPT and other chatbots; neural networks; and robotics, which may be the most commonly used AI application in manufacturing.

Altogether, artificial intelligence capabilities allow manufacturers to redeploy human labor to jobs that machines can't yet do and to make production more efficient and cost-effective.

5 applications of AI in manufacturing 5 applications of AI in manufacturing

Manufacturing is one of many industries that artificial intelligence is changing. Keep reading to see five ways that artificial intelligence is being used in manufacturing today.

1. Robotics

When you imagine technology in manufacturing, you probably think of robotics. Now, more than ever, robots play a major role in manufacturing.

Companies like Amazon (AMZN -0.11%) use robots to move items back and forth and to pick and pack orders. Ford (F -0.16%) uses robots to operate 3D printers, and the robots save time by running them unsupervised overnight, using 3D designs that have been uploaded to the printer. While a human needs to give them the design, the process itself is autonomous.

Cobots or collaborative robots are also commonly used in warehouses and manufacturing plants to lift heavy car parts or handle assembly. Often, cobots are capable of learning tasks, avoiding physical obstacles, and working side-by-side with humans.

Expect robotics and technologies like computer vision and speech recognition to become more common in factories and in the manufacturing industry as they advance.

2. AI in quality control

Quality control is a key component of the manufacturing process, and it's essential for manufacturing. Making machines like automobiles can be deadly if there are errors.

Here, AI is also being applied in a number of different ways. AI-powered vision systems can recognize defects, pull products or fix issues before the product is shipped to customers.

Semiconductor companies, including Samsung (SSNL.F -29.7%), Google (GOOGL -0.1%) (GOOG -0.03%), and NVIDIA (NVDA -3.62%) also use machine learning for quality control, optimizing chip design, and improving manufacturing.

Semiconductor

A semiconductor is a basic element or compound substance that conducts electricity in certain situations.

Quality control is one area where AI systems consistently outperform manual testing processes done by humans. AI machines are also able to optimize production and figure out the root cause of a problem when there is an error.

3. AI in predictive maintenance

Maintenance is another key component of any manufacturing process, as production equipment needs to be maintained.

This is where predictive maintenance comes in. Predictive maintenance analyzes data from connected equipment and production equipment to determine when maintenance is needed. Using predictive maintenance technology helps businesses lower maintenance costs and avoid unexpected production downtime.

Companies like IBM (IBM -0.09%) provide predictive maintenance technology that Washington, D.C., uses to help maintain its water hydrants, for example. Similarly, C3.Ai provides predictive maintenance systems for utility companies, such as an electric grid serving more than 7 million customers. Its system relies on machine learning to prevent asset failures before they happen.

4. Automation

Automation is often the product of multiple AI applications, and manufacturers use AI for automation in a number of different ways.

Robotic process automation (RPA) is the process by which AI-powered robots handle repetitive tasks such as assembly or packaging.

One 2022 survey found that 43% of manufacturing businesses already use RPA. The benefits they've found from automation include a reduction in operational costs by up to 40%; an increase in the manufacturer's control over processes; improved employee performance; and significantly lower downtime.

Some companies that use RPA in manufacturing include Whirlpool (WHR 0.43%), which uses robotic process automation to automate its assembly line and handle materials. So-called bots are also used for quality control checks.

Supply Chain

A supply chain is a series of steps taken to obtain raw materials, turn them into a product, and deliver the product to consumers.

5. AI in the supply chain

Manufacturing goes beyond what takes place on the factory floor. It's crucial for every manufacturer to have a well-managed supply chain so they have the parts they need when they need them.

That's why manufacturers often use artificial intelligence systems for supply chain optimization, focusing on demand forecasting, optimizing inventory, and finding the most efficient shipping routes.

BMW (BMWYY -1.67%) for example, uses AI to predict demand and optimize inventory. In one example, the company installed an AI application to prevent the transportation of empty containers on conveyor belts. The tech also decides if a container needs to be attached to a pallet, and finds the shortest route for boxes to be disposed of.

AI systems can also take into account data from weather forecasts, as well as other disruptions to usual shipping patterns to find alternate route and make new plans that won't disrupt normal business operations.

Will artificial intelligence revolutionize the manufacturing industry? Will artificial intelligence revolutionize the manufacturing industry?

It's clear that artificial intelligence is already having an impact on the manufacturing industry. That influence will continue to grow.

Unlike some other industries, generative AI technologies like ChatGPT seem less likely to have an impact on manufacturing. But other forms of AI tech, like robotics and machine learning applications, will push the envelope, helping manufacturers produce goods more efficiently, relieve human labor, and eliminate errors that lead to product recalls or even real-world accidents.

Some forecasts estimate that the opportunity in artificial intelligence will be worth trillions of dollars. If you're looking to invest in AI manufacturers, you can consider some of the stocks above or take a look at other AI stocks, machine learning stocks, or AI ETFs.

Suzanne Frey, an executive at Alphabet, is a member of The Motley Fool's board of directors. John Mackey, former CEO of Whole Foods Market, an Amazon subsidiary, is a member of The Motley Fool's board of directors. Jeremy Bowman has positions in Amazon.Com. The Motley Fool has positions in and recommends Alphabet, Amazon.Com, and Nvidia. The Motley Fool recommends Bayerische Motoren Werke Aktiengesellschaft and International Business Machines. The Motley Fool has a disclosure policy.

Why Do AI Solutions In Manufacturing Remain At The Pilot Project Level?

CloEE Co-FounderVCSerial IT entrepreneurFormer CEO of SAP SE.

getty

While artificial intelligence (AI) has generated a great deal of interest, its practical applications have mostly found success in industries where numbers dominate, such as retail or banking, because they have a long history of appropriate data collection.

In contrast, introducing AI into manufacturing is a challenge. Integrating physical and chemical parameters, the sheer volume of data and the inherent responsibility associated with decision-making underscore the complexities. Because of these difficulties, many ambitious AI solutions often never pass the initial creation and testing stages.

Instead of jumping straight to the most "exciting" AI solutions, it's far more useful for industries like manufacturing to focus on practical, smaller AI steps to build a foundation for future innovation.

Shifting Expectations: Industrial Companies Need AI Advisors, Not Human Replacements

One of the biggest problems companies face when implementing AI solutions is that the proposed technology is too cumbersome or costly to justify the initial push past "pilot program" status. This is why many companies think that starting with more basic applications such as predictive analytics and moving toward developing digital twins is a more manageable pathway with a straightforward line toward success and good ROI.

Predictive analytics in industrial applications often refers to predictive maintenance: the ability to predict equipment failures based on current status and repair history. However, this creates a dilemma. For example, what should be done if an airplane turbine is scheduled for repair in one month according to maintenance protocols, but predictive analytics suggests that it could function for another six months? Should the maintenance schedule prescribed by the turbine manufacturer be ignored?

This problem encapsulates the central debate around predictive analytics and maintenance. Repair directors and industry experts I've worked with often express doubts, stating that while they are willing to have AI collect data from their machines to make repair decisions, they do not want to be solely responsible.

The best way to overcome this issue is to adjust expectations. It's better to think of AI in an advisory capacity, much like how it's used in healthcare settings. Doctors use AI to narrow down possible diagnoses or analyze large amounts of data to provide more targeted, helpful information for better patient outcomes. Still, the doctors ultimately make the final decisions. Similarly, AI can gather data and make maintenance recommendations, but companies will still rely on human expertise for the final decisions.

Cost Challenges Remain A Major Implementation Hurdle

In an ideal world, every machine would have a digital twin that provides complete information about the performance of each component. In practice, such data is either unavailable or limited. Each device may have unique characteristics, operate in different modes or perform various tasks, so information must be collected directly from each individual machine to get a complete picture.

Achieving full predictive analytics in this situation requires installing multiple sensors on each machine and collecting data over a long period of time. This complexity often makes the idea seem too daunting for manufacturers, especially when they must justify the initial costs of time, money, testing and labor.

However, because maintenance plays such a critical role in industrial plants, the long-term value of predictive analytics should be taken more seriously. Imagine a scenario in which a company runs 10 coffee shops, each with one coffee machine. Knowing in advance when the coffee machine will need maintenance allows the company to avoid disrupting daytime operations by scheduling it at night. This simple example underscores the importance of maintenance and the potential value of predictive analytics to improve scheduling, productivity and profitability.

Since predictive analytics have the potential to take the industrial sector to the next level, the answer to the cost and expertise problems is to start smaller.

The Solution Is To Step Back

Instead of aiming for full predictive analytics right away, enterprises should focus on data related to the underlying production processes and how each production stage is affected by each specific machine.

There is always room for greater production efficiency, so companies can start small. For instance, taking stock of the current manufacturing processes can show managers the most significant bottlenecks and give the company a more narrow focus. Implementing AI solutions for one or two machines—or even just specific parts of machines—that will yield major gains is an excellent way to lay the foundation for future technologies that can further improve operations.

Understanding that the entire production floor does not have to be pulled into full AI data collection right away can ease some of the resource strain and show managers the value of continuing to invest in enhancing manufacturing processes.

As industries continue to push the boundaries of innovation, finding the right balance between the potential of AI and human expertise remains crucial. We can only unlock the true benefits of AI in these areas through a combination of technological advances, research, and collaborative efforts.

As we look toward the future, it's important to remember just how many iterations of AI there are and how many more there could be in the future. Basic AI is already in use, and the next generation to come already has a solid foundation on which to build. Neural networks and other newer types of machine learning are able to analyze and predict problems, which is something everyone is currently trying to solve. Overall, as AI and ML continue to advance and become more widespread, we will see more AI solutions move past the pilot stage and into implementation, allowing the manufacturing sector to become more efficient, sustainable and connected.

Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?

Time To Redesign Your Career For The Age Of Artificial Intelligence

More AI, more people skills needed

getty

It's clear that technology professionals will see their roles transformed with artificial intelligence, requiring skills refreshes and learning new approaches. Not so clear is the impact of AI on career prospects for business managers and professionals, with confusing messages about job replacement and usurping of decision-making authority.

The rise of AI, particularly generative AI, is likely to have a strong impact on managerial and professional jobs, an analysis by Rakesh Kochhar of Pew Research suggests. "Jobs with a high level of exposure to AI tend to be in higher-paying fields where a college education and analytical skills can be a plus. Workers with a bachelor's degree or more (27%) are more than twice as likely as those with a high school diploma only (12%) to see the most exposure."

Yet, the Pew study also suggests, professionals in more exposed industries do not feel their jobs are at risk – "they are more likely to say AI will help more than hurt them personally. For instance, 32% of workers in information and technology say AI will help more than hurt them personally, compared with 11% who say it will hurt more than it helps."

Vittorio Cretella, CIO of Procter & Gamble, doesn't see AI as replacing human talent, but rather, amplifying those talents. "The continued rise of AI will change the type of work we do, and how we do it, but augmenting rather than replacing human capabilities," he argues. "We will still need the skills of digitally-savvy, creative human employees who can work effectively with machines."

AI "will have a profound impact on employees across the whole organization, not only on expert roles such as data science or machine learning engineering," Cretella points out. "Nearly all employees, regardless of function, will need to get familiar with working with machines, exploring insights and leveraging recommendations which may often be different from what their previous experience would suggest."

For corporate leaders, the priorities of the AI age must shift to greater investments in "talent and upskilling employees while in-sourcing strategic capabilities such as data science and machine learning engineering," Cretella advocates. This calls for a balancing act among managers and executives, who "must facilitate the combination of human and machine strengths, creating the organizational focus and culture that encourages continuous learning and the application of AI to improve business outcomes."

AI, implemented successfully, "will amplify human skills, not simply substitute, or replace them," says Cretella. "Key human-centric skills include curiosity, creativity, critical thinking, compassion and collaboration."

Where humans will make the greatest difference is in problem definition, he continues. This consists of "decomposing a problem through key questions and identifying patterns before attempting to define an algorithmic solution. We need leaders and teams to focus on that phase to develop inquisitive skills and dedicate enough time before skipping to solutions."

P&G's approach is to "always start from the job to be done, whether is about maximizing media reach, improving manufacturing quality or defining the shelf layout for the best consumer shopping experience," Cretella illustrates. "What matters is the capability to define the hypothesis and problems, the curiosity to explore data and the power of AI to find the answers."

It still takes curiosity, and a sense of what people need, to succeed in today's hyper-competitive economy. "Technology alone does not change things – people do," he emphasizes. "The future of management is with an AI-savvy generation of business leaders who are curious, have little or no cognitive bias, and understand which organization design, processes, and resources are needed to unleash the power of data and machine learning."

---