AI takes manufacturing beyond Industry 4.0

Industry 4.0 has organized the assets in the workshop and relieved much of the dirty and dangerous physical work in the factory. Artificial intelligence (AI) goes further. These software systems are beginning to replace some of the human decision-making. The idea is that AI can make decisions based on far more data than a human brain can handle. Also, AI and making decisions without bias.s
TIBCO Software explains four ways AI is impacting manufacturing:
Supply chain optimization: AI can automate and speed up the normally manual and tedious process of aligning all orders, purchases and materials needed for production. By augmenting the process with AI, organizations can manage their supply chains more efficiently. This includes tracking supplies and finished products from the manufacturing floor to delivery.
Demand forecast: Manufacturers are beginning to use AI and machine learning to analyze customer behaviors and predict future buying patterns. The intelligence is then transmitted to the teams in charge of the manufacturing process so that they can increase or decrease production.
Quality assurance and error detection: Visual inspection tools linked to AI software detect defects on production lines. With machine vision cameras, systems can detect errors quickly and accurately. It’s better than the human eye.
Extinct factories: Using AI, the robotic equipment requires minimal human intervention. This allows companies to cut costs by eliminating things humans need that robots don’t typically need, such as lighting and other environmental controls.
We caught up Jim Chappell, Head of AI and Advanced Analytics at AVEA, to explain some of the details of AI in manufacturing.
Design News: How do AI systems support human capabilities?
Jim Chappell: AI has become a partner for humans. AI helps us become more sophisticated. It helps humans with repetitive motions and physical labor, but it also helps with decision making and insight. The software has become more powerful. It was used for computer-aided design and then computer-driven control. Now AI takes the lead role while humans oversee.
AI can plan automation and predictive maintenance. He has more tools under his control. You can start linking multiple cognizant tools together. As AI becomes more autonomous, humans will still be very much in the know. In the future, you will use AI more. AI drives data to support Industry 5.0.
DN: Explain how bias-sensitive AI and physics-based simulations can increase accuracy and transparency.
Jim Chappell: There are biases in different things. The data you have may come from an operator with poor practices. If you train a system on these bad practices, you will have a bad bias. Inefficient processes need to transition to good processes.
If you put data relating to temperatures, pressures, flow rates, amperes. If you add physics-based data, you’ll know that many amps should be used. They are not sensors, but they act as sensors and predictive models to include pure reality, and the data will reflect the good, the bad or the ugly. If any part of the operation is out of whack, the physics-based data will help bring it out. You put the sensors together and you get a smarter system. Physics-based data brings that.
DN: Explain your belief that manufacturers can achieve significant cost savings by using AI software to eliminate production losses and machine downtime.
Jim Chappell: You can combine several types of cognitive tools with predictive maintenance. More and more industries are doing it. To make it better, you layer more cognitive systems. It gives you a prescription. Let’s say we have an anomaly. We know the root cause. We produce less steam and become inefficient. AI gives us the prescriptive opportunities to improve it.
When you overlay AI and get early predictive maintenance data, you still need to know how to fix it. How much time do I have? One week? A month? The AI may indicate that you have a few months before it stops. This means it can last until the next scheduled maintenance. Some predictive systems can be too reactive. Factories will close because of predictive maintenance data that is not analytical. Still, they could wait until the downtime is scheduled. Everyone who works with AI becomes predictive asset optimization. This is quite different from the usual predictive maintenance.