- As artificial intelligence (AI) continues to dominate, many traditional industries such as manufacturing have been slow to trust the technology.
- AI it can automate manufacturing processes to increase efficiency and reduce errors, improve innovation through generative design, and create safer working conditions.
- 68% (PDF, p. 6) Manufacturers already have at least one use case or process powered by AI, and these small steps will demonstrate the value of AI and build trust.
Artificial intelligence, or AI, is making strides everyday life– from intelligent assistants like Siri and Alexa to personal robotics and car automation to new advances in healthcare. However, a perception problem remains as people struggle to understand the technology and fear its downsides: security concerns, job replacement, or even feelings of depersonalization.
As AI becomes more prevalent, there is also a growing reluctance to hand over tasks to technology, especially in more traditional industries such as design and manufacturing (D&M). However, the potential of artificial intelligence is hardly used.
According to projections from World Economic Forum (PDF, p. 3), this could generate up to $13 trillion in global economic activity and increase global GDP by 2%. For businesses, the choice to use AI-based tools can raise concerns, especially when it comes to data sharing and security. But as businesses see real benefits in using AI without risk to their data or specific expertise, confidence in AI will grow.
State of AI in D&M
AI may seem like a recent phenomenon, but it is deeply rooted in manufacturing. “I started my AI career 40 years ago in robotic automation systems with 3D vision for a General Motors manufacturing plant,” he says Dr. Jay Lee, a pioneer in industrial artificial intelligence, Clark Distinguished Professor and director of the Center for Industrial Artificial Intelligence in the Department of Mechanical Engineering at the University of Maryland College Park.
“If people tell you that artificial intelligence is just starting, no, it worked for us 40 years ago. The robots assembled the cars using intelligent vision to automatically identify and adjust the trajectory with compensation,” adds Lee, who is also member from the World Economic Forum’s Global Future Council on Advanced Manufacturing and Production.
Companies have long been turning to Dr. Lee to help them improve their operations. When a compressed air system failed at a Toyota plant in Georgetown, Ky., the unplanned shutdown cost money and slowed production at a plant that typically saw a new car roll off the production line every day. 25 seconds.
Lee integrated AI into the production line using sensors and AI to detect anomalies and prevent crashes. Maintenance costs have fallen 50%and this problem has not caused any downtime since the solution was implemented in 2006.
Since these early use cases, AI has become more robust and has moved beyond basic operational functions. It can now help businesses innovate through generative design that allows iterating and simulating different scenarios to achieve the best possible results.
Sixty-six percent of business leaders believe they will need AI in the next two to three years. But a recent Boston Consulting Group study found just that 16% manufacturing companies have achieved their AI goals. Despite early progress, the manufacturing sector has been slow to adopt AI.
You need the right data to trust the process
The manufacturing industry produces approx 1,812 petabytes of data every year, and turning that data into insights and actions can be a driver of innovation if manufacturers allow it.
But according to Deloitte67% of executives are not comfortable providing their data to other organizations.
“If you’re not creating data for a specific thing, it probably can’t be used for that new purpose without being reworked,” says Alec Shuldiner, director of data acquisition and strategy at Autodesk.
“Data mining is the work required to reuse data so that it can be used to drive a new process, such as for analytics or machine learning applications.”
AI is only as good as the data it receives. This will yield the desired results only if this data is reliable, accurate and relevant. “If you give me junk data, I can’t help you,” Lee says. “You must provide me with useful and actionable data.” You need to have the right context to connect the data to the goal you want to achieve.
For example, I want to predict a machine failure. Well, you need to give me some information about the machine’s condition. If you have fish, it is useful, but if it comes from polluted water, it is not edible.
To bridge the gap between the continued reluctance to adopt AI and harnessing its full power, manufacturers must learn to trust what they can’t see. They are comfortable letting AI handle predictive maintenance, but generative AI remains largely unknown.
But it’s a risk worth taking. As manufacturers better understand how AI enables end-to-end visibility, it will create more opportunities for their organizations.
Building trust and unlocking the value of AI
Dr. Lee defines the benefits of AI as the “three Ws”: less work, less waste, and less worry. “There’s a lot we don’t know,” he said.
“For example, some people walk around the factory, they want to check everything. Why? They worry even if the machine never breaks down. AI alleviates these concerns by enabling greater visibility. “If everyone in the community has a security camera, don’t worry. You can have apps to view your house. Oh, who’s there? Oh, Amazon delivery.
As AI proves itself and people become more aware of how it works, they are starting to integrate it more into their operations.
As cloud-connected factories become the norm, AI can be supercharged, gathering all this data in real-time and generating insights quickly. But in the meantime, manufacturers are stuck in a decision-making process.
“In today’s design, we are often forced to make compromises that we would rather not have to make,” says Dr. Shuldiner. “You can design something quickly, or you can design it to be easy to manufacture, or you can design it to achieve a sustainability goal like recyclability. But often you can’t do all these things at once.
So if you want to add recyclability to the design, then you have to spend a lot more time on the design and risk making it more expensive to manufacture.
AI will take us to a point where many of these trade-offs disappear. You will be able to design quickly and efficiently while achieving more complex design goals.
Dr. Lee points to outliers that used advanced technology early on, such as Toyota and General Motors, as companies that continue to innovate, using cloud computing and AI to make better, more vehicles, lighter and more efficient.
But often handing over more of their operations to AI is a gradual process for manufacturers. “Our traditional industry will need continuous improvement,” says Lee. “It’s not an overnight success. Do one little thing first to make it happen. Wow. Got it. GOOD. Let’s move on to the next one.
Sixty eight percent (PDF, p. 6) manufacturers have at least one use case or process powered by AI, and these small steps will demonstrate the value of AI and build trust.
“The priority is to be aware of the benefits of AI,” says Lee. “People are afraid of threats or negative things from AI. But you shouldn’t stop moving forward because you’re worrying too much.