Can supply chain decision-making be automated?

LAKE BUENA VISTA, Fla. — Decision intelligence a relatively new term in the supply chain. Many companies manage by exception, but that is only one part of decision intelligence, explains Henry Hwong, vice president of product marketing for Aera Technology.

Hwong told FreightWaves that decision intelligence is still an evolving field that includes simple decisions such as shifting a load to another carrier because the lane price has exceeded a preset threshold, to more complex decisions such as moving production when a factory shuts down.  

“We are transforming from an era of people making decisions guided by machines to an era of machines making decisions guided by people,” explained Fred Laluyaux, president and CEO of Aera Technology, during a presentation of the company’s product on Monday at the Gartner Supply Chain Symposium/XPO 2022 conference at Walt Disney World’s Dolphin Resort.

Hwong detailed the company’s new product, which is designed to automate decision-making across organizations, although he noted that humans are and will always be part of the process.

Gartner has predicted that by 2025, 95% of decisions that use data will be at least partially automated. That makes AI- and machine learning-driven software a must for supply chain businesses. Aera’s latest iteration of its Decision Cloud platform updates AI/ML operational capabilities to help the system better understand the relationship and effectiveness of decisions, the company said.

Decision Cloud aggregates data from both internal and external systems and sources to build decision-making models that can accommodate a wide range of business decisions.

“Through Aera Decision Cloud, we are re-inventing decision-making and accelerating adoption of decision intelligence,” said Shariq Mansoor, founder and chief technology officer of Aera Technology. “Today’s comprehensive release for both business end user and data science teams underscores how we are continually advancing and evolving our platform to enable digital decisions at scale in this increasingly complex business environment.”

The new product release builds on the platform’s Aera Cognitive Skills, a prebuilt set of capabilities for demand forecasting, planning, inventory, logistics, procurement, finance and revenue.

Specific updates include:

• Enhancements to Aera Cortex, the data science engine inside the Aera Cognitive Operating System, to improve the user experience and including a new Aera notebook, new AutoML and new model performance and versioning.

• Advanced intelligence capabilities to provide a deeper understanding of relationships and effectiveness of decisions, including a new graph explorer, new confidence score and new action item node.

Hwong said the technology is designed to learn what decisions are frequently made and automate those. He said a confidence score gives users the chance to see how often a certain decision has been made. If it is common enough, it can be automated.

“How you react to a change in these plans [is important],” he said. “There is always a person that has to make a decision.”

Hwong noted that “everybody is trying to automate the supply chain, but it is very difficult to do because the supply chain is dynamic.” Aera is not trying to replace the human element but rather supplement it.

“It’s all built around how people make decisions,” he said. “There is never enough planners out there, so we are working to increase the capacity of decision-making.”

Aera Decision Cloud collects data from siloed systems by working on top of existing systems such as warehouse management, transportation and finance. With that data, individual departments can make better decisions, on their own or through an automated process.

As an example, Hwong pointed to a company’s sales department that might plan a promotion in Germany only to find out too late that there isn’t enough inventory to support the promotion. The Aera Decision Cloud could have identified that concern early in the process.

“The first step is always getting visibility into the data,” Hwong said. “We attack things on a case-by-case basis. Working with us is usually a journey.”

Hwong added that one company did an analysis of Aera’s technology and found that it had made 12,000 recommendations in a single month, with 71% of them able to be automated.

“It would have taken 45 man years to do that manually,” he said.

What really distinguishes Aera’s technology, though, is its ability to go back and “rewrite” data. As decisions are made through the system, either automatically or through augmented means with human involvement, the technology goes back to the underlying data sources and inserts rules that ensure future decisions are made in the same manner. Hwong noted that humans are involved so that the system doesn’t automate decision-making every time simply because it was automated once, and only a human can make the decision to automate a process.

It typically takes between six and eight weeks for Aera to get the system installed, collect the data, and start helping businesses make the right decisions.

Click for more articles by Brian Straight.

You may also like:

Drones are flying into weather data deserts. Can they be stopped?

Navigating COVID-19 shipping chaos: Finding capacity and servicing the customer

Need a warehouse? You may have to wait 9 months

Source: freightwaves - Can supply chain decision-making be automated?
Editor: Brian Straight