When changes in wafer fabrication added more links to Texas Instruments' supply chain, the semiconductor maker responded by streamlining its operations.
James Cooke is a principal analyst with Nucleus Research in Boston, covering supply chain planning software. He was previously the editor of CSCMP?s Supply Chain Quarterly and a staff writer for DC Velocity.
Over the years, its supply chain has expanded to include additional steps and players. Yet Texas Instruments hasn't allowed that to slow down the process. James Cooke tells the story in this article, which first appeared in the Quarter 1/2008 edition of CSCMP's Supply Chain Quarterly.
This story represents the type of analytical articles featured in the Quarterly. Readers can obtain a subscription by joining the Council of Supply Chain Management Professionals (whose annual membership includes a free subscription). Subscriptions are also available to non-members for $89 a year. For more information, visit www.SupplyChainQuarterly.com.
You might think that a product made from silicon would flow through a supply chain like sand through an hourglass. But that's hardly the case for Texas Instruments Inc. (TI), one of the world's largest makers of silicon chips. Though its products may not present handling challenges, the Dallas-based company still faces the same difficulties that confront any business trying to coordinate the movement of materials through a far-flung global supply chain. And for TI, those challenges have only been magnified as the company's manufacturing process has become increasingly complex.
Those changes in Texas Instruments' manufacturing process over the years are largely a reflection of marketplace developments. Since TI introduced the first integrated circuit to the world in the middle of the last century, the market has exploded into a multibillion dollar global business. The company, which once concentrated on making chips for use in computers, now produces 55,000 different semiconductor products that are incorporated into a wide variety of electronic devices, including mobile phones, modems, and DVD players. Today, the computer industry accounts for only about 30 percent of TI's semiconductor sales, while customers in the communications industry consume at least half of the 10 billion-plus chips TI makes each year.
Not surprisingly, these changes have also had repercussions for TI's supply chain. Compared to a few decades ago, there are now more steps in the manufacturing process and, because TI has contracted with third parties for some of its manufacturing and distribution needs, more players in the chain. The conventional wisdom holds that the more links in the chain, the greater the potential for delays and other disruptions to the flow of materials and information among the various parties. Yet that hasn't been the case at TI. In the face of added challenges, the semiconductor maker has actually responded by streamlining its operations.
Complicating factors
Semiconductors—also called integrated circuits or chips— are big business for TI, which makes both analog and digital versions. In 2006, the most recent year for which financial figures are available, $13.7 billion of the company's $14.2 billion in revenue (96.5 percent) came from its semiconductor business. Asia and Europe each accounted for 25 percent of the company's semiconductor sales; the United States accounted for another 20 percent.
The chips themselves are made from silicon wafers in highly specialized factories known as fabrication plants, or "fabs." From the fabs, they move to assembly and test (A/T) facilities, where the wafers are cut into individual chips, which are then assembled into packages and tested.
Not so long ago, back when most of TI's chips were sold for use in personal computers (PCs), chip manufacturing was a relatively straightforward business. "The world used to be fairly simple," says Jan De Meulder, TI's director of supply chain logistics. "You were building chips for one PC application in one fab, sending them to one assembly and test site for completion, and your end customer was IBM, who was doing everything else. That was back in the '70s."
These days, it's no longer so simple. "Today, we have some processes [that] start in a first fab and then go to a second fab to complete the wafer's fabrication," explains De Meulder. "Then, they could be packaged in an assembly plant, followed by burn-in (tests in which the devices are subjected to high temperatures to assure quality) and final test processes in yet another site. Quite often, we then have to send those to an electronic manufacturing corporation that builds subassembly boards." From there, the units might be shipped directly to the customer's factory, where end products such as cell phones are assembled.
At the same time, more players have been introduced into the process. Three decades ago, Texas Instruments owned most of its chip-making facilities. But over the years, the company has outsourced part of its production to thirdparty foundries. At present, Texas Instruments uses more than 30 fabs and foundries around the globe. (Fabs make product strictly for TI's customers; foundries take orders from multiple customers, including TI's competitors.)
Delivering the goods
TI's process for moving semiconductors out to its customers, which now number more than 3,000 worldwide, is no less complicated. About 75 percent of the company's semiconductor products move through its distribution network, which includes four big regional distribution centers and more than 70 smaller hubs (see Figure 1). The company uses a regional DC in Singapore to serve customers in Asia; one in Dallas to serve North America; one in Utrecht, the Netherlands, to serve Europe; and one in Tsukuba, Japan, to serve markets in that country. But Texas Instruments doesn't manage these facilities itself; it uses third-party logistics service providers (3PLs) to handle the day-to-day operations of those warehouses.
Another 10 percent of TI's chips move directly from its factories to customers. The other 15 percent are sent to warehouses near key customers' factories. TI maintains inventories for those customers at the sites on a consignment basis.
To move its products, Texas Instruments relies heavily on air carriers, although the company has begun experimenting with ocean shipments for products with long lead times. Given its global reach, TI must keep track of rates on more than 1,000 different shipping routes.
All in all, the process is a lengthy one. On average, it takes eight months to source material, build the various semiconductor components, and deliver them into the end customers' hands. Texas Instruments keeps about 90 days' worth of inventory on hand to meet demand.
All together now!
As for how TI has managed to streamline its supply chain despite the additional steps, the explanation lies partly in its effective use of electronic communications. For starters, TI has standardized its information technology platform, deploying a common global IT system to handle all orders and shipments. The company now uses a single version of SAP's enterprise resource planning (ERP) system across all business units around the globe. The SAP system allows Texas Instruments to take an order from any customer anywhere on the planet and check whether it can meet the order. "People can look at orders and inventory anywhere around the world," says De Meulder. "It allows us to track changes in real time."
At the same time, the company has set up a global connectivity platform to exchange information with its external partners. It uses a variety of methods, ranging from electronic data interchange (EDI) to XML-messaging on the Internet, to connect with those external players. "When you have so many touch points with suppliers, you have to be connected by IT to your partners," De Meulder says.
TI's efforts to standardize its information base have extended to other areas of the operation as well. For example, the company uses a single version of i2's production planning software and has adopted a sales and operations planning (S&OP) process that has pushed everyone in the organization to use a standard set of data. "Everybody is aligned on what the company will aim for, what revenues we're looking for, and what the factories will build to," De Meulder says. "And everybody will execute the same plan."
As part of that process, Texas Instruments conducts a monthly review to spot potential gaps between supply and demand, so that it can act swiftly to resolve disparities. That monthly plan drives the weekly production plan for the factories, which must be kept running in order to be cost effective. "We can't afford to have the factory idle," De Meulder explains.
Fast-moving parts
Along with its IT and planning initiatives, Texas Instruments has also introduced several innovative transportation programs. To coordinate inbound parts shipments from offshore suppliers to its North American production plants, for example, the company has set up an Inbound Routing Center in Dallas. As it did with its regional distribution centers, TI has contracted with a 3PL to run the Inbound Routing Center. Among other functions, the 3PL is responsible for determining the most cost-efficient manner of shipping.
To help ensure a steady flow of parts and components needed for production, the Dallas chipmaker has also contracted with suppliers to set up vendor-managed inventory hubs near its fabs and A/T factories in Asia and Mexico. "We are asking suppliers to put key components right in front of our factories," says De Meulder, who notes that these programs are similar to the ones TI operates for its own customers.
Initiatives aimed at streamlining the company's outbound transportation include what TI calls the "shippacked-to-sales-order" program. Under this program, which is designed to expedite order shipping, a factory in, say, Malaysia will consolidate individual orders for Japanese customers into one container for shipment to TI's Tsukuba distribution center. Before it leaves the factory, each package is labeled with the final customer's name and other delivery details. When the shipment arrives at Tsukuba, workers at the DC break down the load into individual orders for final delivery.
Texas Instruments has also begun deploying transportation management software (TMS) to help it move shipments to customers swiftly at the lowest possible cost as well as to obtain better visibility of shipments in transit. Although it now uses the software only at its distribution centers, the company has plans to deploy the same TMS applications at all of its assembly and test facilities, which will allow those sites to take over tasks like filling out customs and shipping documents. "We want the factory to be able to do the same thing the DC can do," says De Meulder. The company has also developed contingency plans to keep the network running in case of supply chain disruptions. It has re-examined its shipping patterns and routes to ensure that it has an alternative carrier to move products and supplies on every key inbound and outbound lane. "If something happens, we have a backup plan with another strong provider," De Meulder reports.
A perfect delivery network
For all of its successful improvement initiatives, Texas Instruments still considers its supply chain to be a work in progress. It recently engaged an outside research firm to model its distribution network to determine the optimum locations for its distribution centers. "We will use that model to project the future," De Meulder says. He adds that he expects TI to relocate more DC operations to Asia in an effort to get closer to Asian customers, increase its presence in that part of world, and reduce its operating costs.
"We are busy working on developing a truly optimized network that can handle any shipment to any customer anywhere," says De Meulder. "Customers are not ... forgivng today; they want a perfect delivery network."
Congestion on U.S. highways is costing the trucking industry big, according to research from the American Transportation Research Institute (ATRI), released today.
The group found that traffic congestion on U.S. highways added $108.8 billion in costs to the trucking industry in 2022, a record high. The information comes from ATRI’s Cost of Congestion study, which is part of the organization’s ongoing highway performance measurement research.
Total hours of congestion fell slightly compared to 2021 due to softening freight market conditions, but the cost of operating a truck increased at a much higher rate, according to the research. As a result, the overall cost of congestion increased by 15% year-over-year—a level equivalent to more than 430,000 commercial truck drivers sitting idle for one work year and an average cost of $7,588 for every registered combination truck.
The analysis also identified metropolitan delays and related impacts, showing that the top 10 most-congested states each experienced added costs of more than $8 billion. That list was led by Texas, at $9.17 billion in added costs; California, at $8.77 billion; and Florida, $8.44 billion. Rounding out the top 10 list were New York, Georgia, New Jersey, Illinois, Pennsylvania, Louisiana, and Tennessee. Combined, the top 10 states account for more than half of the trucking industry’s congestion costs nationwide—52%, according to the research.
The metro areas with the highest congestion costs include New York City, $6.68 billion; Miami, $3.2 billion; and Chicago, $3.14 billion.
ATRI’s analysis also found that the trucking industry wasted more than 6.4 billion gallons of diesel fuel in 2022 due to congestion, resulting in additional fuel costs of $32.1 billion.
ATRI used a combination of data sources, including its truck GPS database and Operational Costs study benchmarks, to calculate the impacts of trucking delays on major U.S. roadways.
There’s a photo from 1971 that John Kent, professor of supply chain management at the University of Arkansas, likes to show. It’s of a shaggy-haired 18-year-old named Glenn Cowan grinning at three-time world table tennis champion Zhuang Zedong, while holding a silk tapestry Zhuang had just given him. Cowan was a member of the U.S. table tennis team who participated in the 1971 World Table Tennis Championships in Nagoya, Japan. Story has it that one morning, he overslept and missed his bus to the tournament and had to hitch a ride with the Chinese national team and met and connected with Zhuang.
Cowan and Zhuang’s interaction led to an invitation for the U.S. team to visit China. At the time, the two countries were just beginning to emerge from a 20-year period of decidedly frosty relations, strict travel bans, and trade restrictions. The highly publicized trip signaled a willingness on both sides to renew relations and launched the term “pingpong diplomacy.”
Kent, who is a senior fellow at the George H. W. Bush Foundation for U.S.-China Relations, believes the photograph is a good reminder that some 50-odd years ago, the economies of the United States and China were not as tightly interwoven as they are today. At the time, the Nixon administration was looking to form closer political and economic ties between the two countries in hopes of reducing chances of future conflict (and to weaken alliances among Communist countries).
The signals coming out of Washington and Beijing are now, of course, much different than they were in the early 1970s. Instead of advocating for better relations, political rhetoric focuses on the need for the U.S. to “decouple” from China. Both Republicans and Democrats have warned that the U.S. economy is too dependent on goods manufactured in China. They see this dependency as a threat to economic strength, American jobs, supply chain resiliency, and national security.
Supply chain professionals, however, know that extricating ourselves from our reliance on Chinese manufacturing is easier said than done. Many pundits push for a “China + 1” strategy, where companies diversify their manufacturing and sourcing options beyond China. But in reality, that “plus one” is often a Chinese company operating in a different country or a non-Chinese manufacturer that is still heavily dependent on material or subcomponents made in China.
This is the problem when supply chain decisions are made on a global scale without input from supply chain professionals. In an article in the Arkansas Democrat-Gazette, Kent argues that, “The discussions on supply chains mainly take place between government officials who typically bring many other competing issues and agendas to the table. Corporate entities—the individuals and companies directly impacted by supply chains—tend to be under-represented in the conversation.”
Kent is a proponent of what he calls “supply chain diplomacy,” where experts from academia and industry from the U.S. and China work collaboratively to create better, more efficient global supply chains. Take, for example, the “Peace Beans” project that Kent is involved with. This project, jointly formed by Zhejiang University and the Bush China Foundation, proposes balancing supply chains by exporting soybeans from Arkansas to tofu producers in China’s Yunnan province, and, in return, importing coffee beans grown in Yunnan to coffee roasters in Arkansas. Kent believes the operation could even use the same transportation equipment.
The benefits of working collaboratively—instead of continuing to build friction in the supply chain through tariffs and adversarial relationships—are numerous, according to Kent and his colleagues. They believe it would be much better if the two major world economies worked together on issues like global inflation, climate change, and artificial intelligence.
And such relations could play a significant role in strengthening world peace, particularly in light of ongoing tensions over Taiwan. Because, as Kent writes, “The 19th-century idea that ‘When goods don’t cross borders, soldiers will’ is as true today as ever. Perhaps more so.”
Hyster-Yale Materials Handling today announced its plans to fulfill the domestic manufacturing requirements of the Build America, Buy America (BABA) Act for certain portions of its lineup of forklift trucks and container handling equipment.
That means the Greenville, North Carolina-based company now plans to expand its existing American manufacturing with a targeted set of high-capacity models, including electric options, that align with the needs of infrastructure projects subject to BABA requirements. The company’s plans include determining the optimal production location in the United States, strategically expanding sourcing agreements to meet local material requirements, and further developing electric power options for high-capacity equipment.
As a part of the 2021 Infrastructure Investment and Jobs Act, the BABA Act aims to increase the use of American-made materials in federally funded infrastructure projects across the U.S., Hyster-Yale says. It was enacted as part of a broader effort to boost domestic manufacturing and economic growth, and mandates that federal dollars allocated to infrastructure – such as roads, bridges, ports and public transit systems – must prioritize materials produced in the USA, including critical items like steel, iron and various construction materials.
Hyster-Yale’s footprint in the U.S. is spread across 10 locations, including three manufacturing facilities.
“Our leadership is fully invested in meeting the needs of businesses that require BABA-compliant material handling solutions,” Tony Salgado, Hyster-Yale’s chief operating officer, said in a release. “We are working to partner with our key domestic suppliers, as well as identifying how best to leverage our own American manufacturing footprint to deliver a competitive solution for our customers and stakeholders. But beyond mere compliance, and in line with the many areas of our business where we are evolving to better support our customers, our commitment remains steadfast. We are dedicated to delivering industry-leading standards in design, durability and performance — qualities that have become synonymous with our brands worldwide and that our customers have come to rely on and expect.”
In a separate move, the U.S. Environmental Protection Agency (EPA) also gave its approval for the state to advance its Heavy-Duty Omnibus Rule, which is crafted to significantly reduce smog-forming nitrogen oxide (NOx) emissions from new heavy-duty, diesel-powered trucks.
Both rules are intended to deliver health benefits to California citizens affected by vehicle pollution, according to the environmental group Earthjustice. If the state gets federal approval for the final steps to become law, the rules mean that cars on the road in California will largely be zero-emissions a generation from now in the 2050s, accounting for the average vehicle lifespan of vehicles with internal combustion engine (ICE) power sold before that 2035 date.
“This might read like checking a bureaucratic box, but EPA’s approval is a critical step forward in protecting our lungs from pollution and our wallets from the expenses of combustion fuels,” Paul Cort, director of Earthjustice’s Right To Zero campaign, said in a release. “The gradual shift in car sales to zero-emissions models will cut smog and household costs while growing California’s clean energy workforce. Cutting truck pollution will help clear our skies of smog. EPA should now approve the remaining authorization requests from California to allow the state to clean its air and protect its residents.”
However, the truck drivers' industry group Owner-Operator Independent Drivers Association (OOIDA) pushed back against the federal decision allowing the Omnibus Low-NOx rule to advance. "The Omnibus Low-NOx waiver for California calls into question the policymaking process under the Biden administration's EPA. Purposefully injecting uncertainty into a $588 billion American industry is bad for our economy and makes no meaningful progress towards purported environmental goals," (OOIDA) President Todd Spencer said in a release. "EPA's credibility outside of radical environmental circles would have been better served by working with regulated industries rather than ramming through last-minute special interest favors. We look forward to working with the Trump administration's EPA in good faith towards achievable environmental outcomes.”
Editor's note:This article was revised on December 18 to add reaction from OOIDA.
A Canadian startup that provides AI-powered logistics solutions has gained $5.5 million in seed funding to support its concept of creating a digital platform for global trade, according to Toronto-based Starboard.
The round was led by Eclipse, with participation from previous backers Garuda Ventures and Everywhere Ventures. The firm says it will use its new backing to expand its engineering team in Toronto and accelerate its AI-driven product development to simplify supply chain complexities.
According to Starboard, the logistics industry is under immense pressure to adapt to the growing complexity of global trade, which has hit recent hurdles such as the strike at U.S. east and gulf coast ports. That situation calls for innovative solutions to streamline operations and reduce costs for operators.
As a potential solution, Starboard offers its flagship product, which it defines as an AI-based transportation management system (TMS) and rate management system that helps mid-sized freight forwarders operate more efficiently and win more business. More broadly, Starboard says it is building the virtual infrastructure for global trade, allowing freight companies to leverage AI and machine learning to optimize operations such as processing shipments in real time, reconciling invoices, and following up on payments.
"This investment is a pivotal step in our mission to unlock the power of AI for our customers," said Sumeet Trehan, Co-Founder and CEO of Starboard. "Global trade has long been plagued by inefficiencies that drive up costs and reduce competitiveness. Our platform is designed to empower SMB freight forwarders—the backbone of more than $20 trillion in global trade and $1 trillion in logistics spend—with the tools they need to thrive in this complex ecosystem."