AI Is Manufacturing’s Next Big Step

Artificial Intelligence can change your game, provided you lay a solid foundation.

By Jeff Winter, Hitachi Solutions

In a world constantly evolving and reinventing itself, the manufacturing industry is not one to be left behind. One can imagine the manufacturing facility of the future — machines and equipment interconnected, working in perfect harmony, driven by the unprecedented power of artificial intelligence (AI). AI’s potential to augment and automate human decision making is poised to transform every aspect of manufacturing, from the C suite to the shop floor.

The promise is enhanced efficiency, a significant boost in throughput, considerable cost reductions, and a heightened state of predictability, responsiveness, and agility.

While it’s thrilling to paint a picture of the future, it’s equally compelling to delve into where we currently stand. Recent insights from the Manufacturing Leadership Council’s, Washington, 2023 research, which is part of their “Manufacturing in 2030 Project”, throw light on AI’s burgeoning presence in the manufacturing sphere. Around 57% of manufacturers are currently exploring AI’s applications, trying to pinpoint areas where its introduction could offer maximum benefits. While a smaller subset, at 29%, has seamlessly woven AI into their strategic vision, another 28% have actively integrated it into their operations. With a whopping 96% of manufacturers keen on amplifying their AI-related investments, the future certainly seems promising.

The value of AI in manufacturing is unmistakably demonstrated when one delves into application areas. Plant floor IIoT analysis attracts a 40% focus, preventive maintenance stands at 36%, and process and quality improvements at 30% each. These findings align well with insights from IIoT analytics, which emphasizes predictive maintenance as a central area of AI application.

In addition, McKinsey’s, New York City (mckinsey.com) “2023 Global Survey” serves as a testament to the benefits of AI in manufacturing. The study found that 55% of manufacturers employing AI witnessed decreased operational costs. Meanwhile, 66% saw a rise in their revenues, with 16% reporting more than 10% increases. The standout revelation: In both cases (cost & revenue), manufacturing as a function had higher overall numbers than all eight other business functions evaluated. As a result, the adoption rate is higher and the gains are undeniable.

Looking at benefits, the World Economic Forum’s, Geneva, Switzerland (weforum.org) “AI in Manufacturing 2022” whitepaper delves deeper into potential AI applications by categorizing them into six main-use cases:

Production processes: Encompassing everything from efficient line balancing to innovative product design.

Quality assurance: Ensuring that the products are consistently of the highest caliber.

Maintenance: From routine checks to predictive strategies, ensuring minimal machine downtime.

Health and safety: Making sure the workforce remains shielded from potential risks.

Supply-chain management: Managing everything from accurate demand forecasting to optimizing costs.

Energy management: Promoting responsible and sustainable energy consumption.

Manufacturing leads all industries in AI implementation. The graph shows where the technology is being used to improve operations.

AI Success Stories

While data and projections are illuminating, the real litmus test for any technology is its practical application. For example, Microsoft’s journey to optimize its hardware-manufacturing processes in China and the U.S. saw a dramatic shift with the adoption of AI. Eager to enhance speed, accuracy, and efficiency, the team implemented a three-pronged AI strategy: connection, prediction, and cognition.

Manufacturing is the top AI application arena.

They created dashboards to ensure consistent real-time data views for all while fostering better decision making and boosting operational accuracy and efficiency. This contributed to significant financial savings. However, the process was more than merely technical. A cultural transformation was imperative for AI’s successful integration, emphasizing trust building, role modeling by executives, and constant reinforcement. The results were impressive:

• They’ve successfully managed a 95% commitment to orders within a two-day timeframe.

• Operational errors were drastically minimized, leading to a $50-million annual savings.

• Process optimization contributed to an additional $10 million in savings.

• Their ability to forecast demand improved by an impressive 15%.

As another example—according to IIoT analytics—Nissan operates an AI preventive-maintenance platform to generate remaining useful-life forecasting on 7,500 assets. Nissan claims an unplanned downtime reduction of 50% and a payback period of less than three months.

Avoid Common Missteps

This revolutionary technology promises a future in which machines can learn, adapt, and optimize operations in ways previously seen only in the realm of science fiction. Yet, like all tools, AI’s effectiveness hinges on its proper use. As many industries are finding out, the road to AI adoption is riddled with pitfalls. According to the same Manufacturing Leadership Council study, only 22% of manufacturers currently employ metrics to assess the effectiveness of AI. An even more substantial number, 61%, don’t have specific metrics to measure the effectiveness/impact of AI deployments.

Here are some common missteps to avoid:

• Putting the cart before the horse: Many organizations get so enamored with the allure of AI that they lose sight of their objectives. They invest heavily in technology, only to realize they need to clearly define its purpose or objective. AI, even with its myriad applications, can easily lead companies to tangential initiatives or capabilities, much like a spiderweb sprawling in all directions.

• Overestimating AI capabilities: It’s essential to temper expectations. AI isn’t an all-powerful magic wand. While it can process vast amounts of data at lightning speed, it stumbles in areas demanding human intuition. Complex decision-making scenarios, unpredictable situations, and nuanced analyses are often better left to human expertise.

• Ignoring ethics and privacy concerns: The AI landscape is fraught with ethical landmines. From biases in algorithms that can perpetrate discrimination to issues of transparency and accountability, AI poses a new set of challenges. Add to this the dilemmas of creativity versus ownership, the potential for misinformation, and a slew of data concerns such as privacy, security, and surveillance. Companies must tread carefully because, as technology improves and social sentiment changes, new ethical considerations constantly emerge.

• Neglecting a data strategy: AI’s lifeblood is data. According to McKinsey Global Institute, manufacturing collects and stores more data than any other industry, almost double the next highest industry. However, merely having data is not enough. The quality, diversity, and relevance of this data is crucial. Without a robust data strategy that ensures you have the correct data at the right time, probably contextualized, normalized, and properly indexed, even the most advanced AI models will be unreliable and underperform.

• Treating AI as a project, not a strategy: This is a pivotal distinction. Viewing AI as a short-term project restricts its potential. Instead, seeing it as a strategic ally can unlock its transformative capabilities, guiding long-term growth and innovation.

For industries, the allure of artificial intelligence is undeniable. It promises optimization, efficiency, and innovation, yet, as with all tools, its success hinges on judicious use. By sidestepping these common pitfalls, companies can better harness AI’s true potential, unlocking new possibilities.

Research indicates that 55% of manufacturers employing AI witnessed decreased operational costs and 66% saw a rise in their revenues.

The Path Forward

Understanding where and how to begin is crucial for businesses eager to embark on this transformative journey. The initial step is often introspective — understanding the company’s current capabilities, its potential challenges, and identifying the areas where AI could bring about the most impactful change. This requires a blend of top-down leadership commitment and bottom-up skills development.

Starting small is essential, focusing on pilot projects that can demonstrate quick wins and gradually build internal confidence. As organizations gain more proficiency, they can expand AI applications across broader domains, eventually moving from mere problem solving to strategic, future-oriented solutions.

As AI’s role in manufacturing expands, it’s more than just a passing trend or a mere tool. It signifies a larger shift toward an efficient, innovative, and sustainable future world where machines and humans collaborate seamlessly, driven by the shared goal of continuous evolution and growth. EP

Jeff Winter is the Senior Director of Industry Strategy for manufacturing with Hitachi Solutions, Santa Clara, CA (global.hitachi-solutions.com). He is active in Industry 4.0, with leadership roles on the Executive Board and Smart Manufacturing & IIoT Divisions of ISA (International Society of Automation, Pittsburgh, isa.org), the International Board of Directors for MESA (Manufacturing Enterprise Solutions Association), and the Smart Manufacturing Advisory Board for Purdue Univ.