Competing in the Age of AI-centric Supply Chains
Artificial intelligence (AI) is the ultimate evolution of business intelligence and analytics as it uses data to generate smart insights. At the same time, it gets smarter by learning from patterns, behaviors and feedback from various digital and human interfaces.
* By Mary Lou Jay *
Artificial intelligence (AI) is the ultimate evolution of business intelligence and analytics, according to the 2019 MHI Annual Industry Report, “Elevating Supply Chain Digital Consciousness.”
AI gathers digital information from a variety of sources, including the IoT, data from the cloud, automated equipment (robots, autonomous vehicles and wearable technology) and advanced analytics. It uses all of this data to generate smart insights. At the same time, it gets smarter itself by learning from patterns, behaviors and feedback from various digital and human interfaces.
Supply chain executives are aware of AI’s potential: 88% of those surveyed for the report believe that AI will allow an organization to manage risk and improve predictability, while 79% said that AI will become a core competency within three years.
Increased efficiency
Today’s supply chain has to satisfy consumers who want lower costs for goods, faster response times and free delivery.
“One of the greatest challenges that we saw in this report is trying to grapple with the changing expectations of consumers,” said Randy Bradley, assistant professor of supply chain management, University of Tennessee. “We’re trying to figure out how to adapt and to adjust. Things have changed more quickly than we have been able to change our organizational processes and activities.”
Through AI and machine learning, an organization can look for patterns that predict consumer behavior so they can improve their processes for faster response times.
In the non-physical world, AI will help companies do better forecasting, figuring out how much product to get, where to put that product and how to allocate it so it’s closer to the customer. “AI algorithms add layers of real-time modeling that can look at data such as history and geographies in a way that it’s never been done before, “said Jim Liefer, CEO of MHI member Kindred AI.
“In the physical world, I believe that where you’ll see the greatest impact from the use of AI is in the control of robots by using inputs from the sensors like cameras and barcode scanners,” he said. A robotic arm that incorporates AI can take that sensor information to learn to navigate in an unstructured environment, just as humans do.
“I want the robotic arm to get from point A to point B, but I’m not going to tell it anything about how to get there,” said Liefer. The arm would start off by “babbling”—moving around randomly—but with every movement it makes it understands when it gets closer to the destination point. “Within a few hours it has tried every possible permutation and then it can easily navigate directly to that end point.”
If the endpoint changes, the robot knows how to get to that new endpoint faster because it already learned how to get to the initial endpoint.