How using analytics and AI can help companies manage the semiconductor supply chain – CNBC

Yuichiro Chino | Moment | Getty Images

Businesses and consumers have been grappling with supply chain issues for months, resulting in annoying shortages of all kinds of products, including all-important semiconductor chips.

And while the CHIPS and Science Act, signed into law in August, is designed to boost semiconductor manufacturing in the U.S., there's no telling what effect the legislation will have on supply, or even when.

"The semiconductor supply chain is still constrained," said Brandon Kulik, semiconductor industry leader and principal at Deloitte Consulting. "Lead times on average have come down slightly, given softening in the consumer electronics segment [laptops and smartphones], and demand for memory has declined. But demand for higher performing data center chips, defense, and automotive chips remains historically high, with some semiconductor companies seeing growth in the area of 40% or more."

One potential nearer-term solution for companies that rely on semiconductors: advanced data analytics and artificial intelligence tools to help manage supply issues.

"The Covid-19 pandemic vividly illustrated the impact that unexpected events can have on global supply chains," said Rohit Tandon, managing director and global AI & analytics services leader at Deloitte. "However, AI can help the world avoid similar disruptions in the future."

By crunching through the massive amounts of data being generated by today's supply chains, AI can predict a range of unexpected events, such as weather conductions, transportation bottlenecks, and labor strikes, helping to anticipate problems and reroute shipments around them, Tandon said.

"AI can also enable dramatic improvements in other key supply chain areas, including demand forecasting, risk planning, supplier management, customer management, logistics, and warehousing," Tandon said.

This can lead to improved operating efficiency and working capital management, greater transparency and accountability, and more accurate delivery estimates; and fewer supply disruptions, Tandon said. "In addition, manufacturers that are using AI for visibility in their smart factory operations can better respond to potential disruptions to avoid delays and pivot if needed, enabling them to be more resilient while continuing to meet customer demands," he said.

"Organizations can leverage data analytics tools for deeper insights across the supply chain," Tandon said. "These tools are designed to improve demand prediction and support data sharing with customers and partners." In addition, organizations can use AI to predict or forecast supply chain-related events such as logistics challenges, geopolitical issues, and supply disruptions.

They can either execute actions autonomously or recommend actions stakeholders should take, "ultimately helping companies build resilience into their supply chains," Tandon said.

When deploying these tools for supply chain management, it's a good idea to start with a small and narrow scope and evolve the depth and breadth of the models and algorithms as the results show their accuracy and value, Tandon said.

High-quality data is also important. "Underlying data is key, as bad data equals bad analytics," Tandon said."Lack of transparency across the supply chain is often the result of inconsistent and incomplete data across product, supplier and customer.Standing up data governance processes that align to common definitions and [fixing] data issues provides the foundation of data quality that builds trust in the output of the analytics and AI process."

Rand Technology, an independent semiconductor distributor, is using data analytics to solve customer challenges related to supply.

"For example, if a customer has a need to alleviate inventory surplus, we use data and analytics to identify other users of these products and create an opportunity to rehome them," said Jennifer Strawn, vice president of solutions and sourcing for the Americas and EMEA at Rand. "In this way, OEMs and contract manufacturers are able to shore up their inventory mix of components."

In addition, data and analytics are especially important during a manufacturer's new product introduction phase in the bill of materials selection, Strawn said. "It is critical, during this phase, to identify where you can build flexibility into the design so that there are multiple sources for semiconductors on the approved list of materials," she said.

In this way, manufacturers are not reliant on a single semiconductor provider, which in the current environment could impact business. "We leverage advanced analytics to help determine the availability of these semiconductors and to spot trends and patterns, such as gaps, price increases or product change notices, before products are in production," Strawn said. Rand also uses the technology to drive decisions on future scenarios and to determine how much buffer stock a company might want to secure, she said.

Rand also uses advanced data analytics to identify trends and patterns that enable it to guide customers strategically through perilous market conditions. "With modeling and real-time visibility into availability, market shifts and conditions globally," Strawn said, "we are able to help reduce risks and map strategies in advance that can be employed when we note certain changes and disruptions in the industry."

See the original post:

How using analytics and AI can help companies manage the semiconductor supply chain - CNBC