Market research firm, Interact Analysis, sent this analysis of factory construction in the United States, Why has US reshoring not translated into meaningful factory construction? Written by senior analyst Matthieu Kulezak, the research notes that following a wave of investment from 2020 to 2024, the momentum is clearly fading “with leading indicators pointing to a sharp slowdown of new project activity.”
One of the clearest signals comes from the Index of Business Applications for Manufacturing Facilities. While applications fluctuated at elevated levels throughout 2024, momentum weakened significantly in early 2025. Total applications fell by 39.1% year-on-year in May 2025, for example.
As one of several indicators informing our forecast, the slowdown in new project applications points to a weaker pipeline of upcoming factory builds. We have, therefore, revised down our outlook for new factory construction in the US. The slowdown in new project applications signals reduced momentum in greenfield development, which is now feeding through into our forecast. As a result, our Q1 2026 forecast shows a much sharper decline in activity, with indexed growth falling to 76.0 in 2026, compared to 105.9 in our previous Q4 2025 view.
The first Trump administration made a concerted effort to force companies to return manufacturing to the US—or, at least, move from China. The Biden administration had similar goals using different tactics. The second Trump administration showed even more aggression in that regard through selective use of tariffs and personal “conversations” with prominent CEOs. Not to mention Harry Moser’s Reshoring Initiative that tried to use media to persuade company executives of the value of manufacturing here.
This sharp contraction in new manufacturing applications stands in contrast to the prevailing narrative around reshoring and near-shoring. While policy support and strategic intent remain strong, the data suggests that this has not translated into a sustained pipeline of new factory construction. Instead, companies appear to be delaying or scaling back new investments in response to macroeconomic uncertainty.
The rate of factory construction dipped significantly in 2025, with 2026 so far showing a similar trend. The inflation-adjusted index rose sharply from around 5,500–6,000 between 2017 and 2020 to a peak of 12,070 in December 2023, reflecting strong investment in large-scale factories. However, momentum reversed in 2024, with year-on-year declines exceeding 20% in multiple months. This weakness continued into 2025, with construction values falling 10 to 19% year on year and stabilizing at around 11,200–11,600, well below peak levels.
While factory application counts could suggest that fewer but larger facilities were still being built, the decline in total construction value shows that large-scale projects are also slowing. Because this metric reflects the total size and capital intensity of factories under construction, it confirms that overall manufacturing capacity expansion has weakened, not just small facility construction.
Rising demand for manufactured goods in the US is being met by higher utilization rates and brownfield expansion, not new factory construction.
The decline in new factory construction does not mean U.S. manufacturing is weakening. The U.S. already has a large and mature manufacturing base, so rising demand is increasingly met by expanding existing sites and increasing throughput rather than building new facilities. This is reflected in capacity utilization, which recovered from 62.5% in April 2020 to around 77–78% in 2021–2022. Although it softened in 2024, utilization stabilized and increased through 2025, rising from 74.5% in January to around 75.5% by December.
I’ve seen this following thought enacted by a few recent announcements of investment in current facilities.
An increase in production capacity or reshoring does not necessarily mean a new factory is being built. In many cases, what is described as a “new factory” is actually an expansion or repurposing of an existing site.
For example, John Deere’s announced excavator facility in Kernersville, North Carolina, is a $70M expansion of an existing campus, not a greenfield factory. The site brings production previously carried out in Japan into an existing U.S. facility and adds around 150 jobs, but it reflects capacity relocation and expansion rather than the creation of a new standalone factory. This distinction is important when analysing manufacturing growth. Output can increase through expansions, automation, or relocation without increasing the total number of factories. As a result, reshoring and investment announcements may signal higher domestic production, but they don’t always correspond to growth in the physical factory count.
I have been involved with recycling since the mid-80s. I hate waste—whether as in Lean or as in throwing stuff away. This news came to me from a company called Woodchuck—a clever play on words since they recycle wood. Also a good example of effective use of AI.
Grand Rapids, Michigan – March 24, 2026 – Woodchuck, the AI-powered climate-tech startup redefining how construction and manufacturing industries handle wood waste, today announced a joint sustainability initiative with Walbridge, one of the nation’s top industrial and automotive constructors. The program supports Ford Motor Company’s construction waste-reduction efforts at its new manufacturing facility in Marshall, Mich.
In the first three months, the program has given teams a clearer view of the materials being discarded, diversion rates, cost reductions, and operational efficiency — already achieving 40% of the project’s projected materials-related savings. This early progress offers Walbridge a powerful solution to address customers’ waste management needs and lays the groundwork for a new standard operating procedure for future large-scale construction projects.
A Legacy Builder Confronts a Modern Waste Challenge
For more than a century, Walbridge has delivered some of the most complex automotive and industrial projects in North America. As Walbridge’s customers expand their sustainability commitments, construction waste management is a growing priority — particularly on megaprojects where the volume and variability of materials can shift daily.
On the Ford project, wood waste quickly emerged as one of the most unpredictable waste elements. Crating, dunnage, international shipping pallets, and custom rigging arrived in wide-ranging sizes and material types, creating a diverse and constantly changing waste stream.
These complexities revealed opportunities for innovation. Walbridge saw the potential to elevate efficiency, reduce hauling expenses, and strengthen alignment with Ford’s sustainability goals. The need for real-time visibility into container levels and the makeup of each load became a catalyst for adopting a smarter, data-driven solution — one that made waste handling more predictable, cost-effective and sustainable.
“Our partnership with Woodchuck is built on collaboration. Transparent and real-time communication allows our team to adapt quickly to changing material waste streams on the ground. Detailed information about each load provides complete visibility not only into what is diverted from a landfill, but also into its end destination and intended use, delivering transparency and enabling measurable sustainability outcomes,” said Sander Mathijs, Walbridge Sustainability Manager. “Another key program feature is its ability to scale, allowing us to calibrate capacity and scope to meet the waste‑diversion needs of the project.”
Woodchuck’s AI Platform Delivers Immediate, Scalable Impact
Woodchuck.ai leverages its AI platform across the Ford project to track, report and validate the diversion of wood, cardboard, plastic, and metal; all with minimal onsite labor and seamless integration into Walbridge’s existing workflows.
Walbridge saw meaningful improvements within the first quarter diverting thousands of tons of wood, cardboard, plastic and metal; reducing waste, reducing landfill dependency, and reducing costs. Over the course of the project, Woodchuck will divert 8,000 tons of wood and 1,000 tons of cardboard, plastic, and metal from landfills.
Woodchuck’s detailed reporting also strengthens accountability, giving Walbridge clear data documenting recycling and reuse for both internal tracking and customer sustainability documentation.
Because the Woodchuck platform is designed for large, multi-phase construction programs, the improvements seen at the Marshall project can be replicated at scale. Whether deployed on a single megaproject or rolled out across multiple sites, contractors gain the same visibility, control, and cost efficiencies, making the solution a powerful model for nationwide waste management and sustainability performance.
“Our partnership with Woodchuck has been a game-changer,” said Ross Linton, Group Vice President, Walbridge. “In just a few short months, they’ve helped us transform our waste process to one that’s measurable, trackable, and easily managed. Our team is empowered to plan ahead, driving efficiency and sustainability. We’re excited about the future possibilities this collaboration brings.”
Creating a New Standard for Future Walbridge Projects
Based on early results, Walbridge expects the Woodchuck-enabled process to become a foundation for future large-scale builds across automotive, manufacturing, technology, and advanced industrial sectors.
“Walbridge is demonstrating what it looks like when a contractor treats waste as a strategic input rather than an afterthought,” said Todd Thomas, CEO of Woodchuck. “By embracing real-time data, AI-enabled insights, and a commitment to measurable sustainability outcomes, they’re proving that smarter waste management isn’t just good for the environment — it’s good for productivity, cost efficiency, and project certainty. Their leadership on Ford’s Marshall project shows what’s possible when innovation becomes part of the construction workflow, and they’re setting the pace for how the industry will operate going forward.”
About Woodchuck
Woodchuck is a climate impact start-up dedicated to empowering contractors, manufacturers, and biomass energy producers by streamlining wood waste diversion and processing. We are committed to leveraging advanced AI technologies to transform waste into valuable resources, reduce landfill usage, and provide a steady, sustainable supply of biomass. Based in Grand Rapids, Michigan, Woodchuck is funded by an investor syndicate led by Mason Fink, Beckett Industries, NorthStar Clean Energy and Alloy Partners. For more information, visit https://woodchuck.ai/.
About Walbridge
Walbridge is one of America’s largest privately held construction companies founded in Detroit in 1916. The company offers construction management, engineering, and real estate services for customers in manufacturing, hyperscale data centers, automotive, defense, higher education, health care, and government. Walbridge employs more than 1,500 professionals in North America. Visit www.walbridge.com or connect with us on LinkedIn to learn more.
Process Description
Woodchuck uses AI in two fundamentally different—but tightly connected—ways: at the job site and in the data layer. Together, they turn what was once an opaque, manual waste process into a real-time, measurable system.
1. AI at the Job Site: Shifting Sorting to the Beginning. Traditionally, construction waste sorting happens after the dumpster is full—if it happens at all.
That process is:
Manual and labor-intensive
Expensive to perform at scale
Logistically inefficient
Often skipped entirely
The result? Most mixed construction debris—especially wood—ends up in landfills, even when it could have been reused or converted into energy.
Woodchuck flips this model.
Instead of waiting until the end, Woodchuck uses AI-enabled image recognition at the point of disposal:
As materials are placed into dumpsters, cameras and sensors identify what’s being thrown away
The system distinguishes wood from other materials in real time
It guides proper usage of containers and flags contamination early
This front-end sorting approach changes everything:
Reduces contamination before it becomes a problem
Eliminates the need for costly post-collection sorting
Increases diversion rates dramatically (from <30% to >95%)
Ensures clean wood streams that can be converted into renewable biomass
In short, AI moves sorting from a reactive, end-of-process activity to a proactive, in-the-moment decision.
2. AI in the Data Layer: Turning Waste into Intelligence
Once materials are collected, Woodchuck’s platform continues to track and analyze everything that happens next. This is where the second layer of AI comes in: data aggregation, modeling, and reporting.
Through its dashboard, Woodchuck provides construction companies, developers, and asset owners with full visibility into their waste streams, including:
Material tracking
Exactly how much wood was collected, where it came from, and how it was processed
End-of-life transparency
Clear documentation showing where the material went—whether to biomass facilities or other reuse pathways
Carbon impact metrics
Precise calculations of:
CO₂e emissions avoided from landfill diversion
Carbon benefits from renewable energy generation
Energy output conversion
How much renewable energy was produced from their waste (e.g., BTUs generated, equivalent homes powered)
Operational insights
Trends across projects, contamination rates, and opportunities to improve efficiency
This transforms waste reporting from a rough estimate into a verified, auditable dataset—something increasingly critical for:
ESG reporting
Regulatory compliance
Winning sustainability-driven bids
Internal performance benchmarking
3. From Waste Management to a Measurable System
What makes Woodchuck different is not just the use of AI—it’s where and how it’s applied:
At the edge (job site): AI drives behavior change and improves material quality in real time
In the platform (dashboard): AI converts operational data into financial, environmental, and strategic insights
The result is a closed-loop system where:
Waste is captured correctly from the start
Materials are tracked through their full lifecycle
Outcomes are quantified and reported with precision
Construction companies no longer have to guess what happened to their waste—or treat it as a cost center.
They can see it, measure it, and increasingly, use it as a source of savings, energy, and competitive advantage.
Check out the sidebar ad about the Carbon Almanac. Written and edited by a hundred volunteers, this book contains many ways to help solve the carbon waste problem.
Click on the Follow button at the bottom of the page to subscribe to a weekly email update of posts. Click on the mail icon to subscribe to additional email thoughts.
Fanuc announced in recent news a significant investment to create production capacity for robot manufacturing in the US.
We experienced more announcements regarding investments in manufacturing and infrastructure in the US during the past 10 years than just about any other flurry of announcements. Most never came to pass. This thought includes the multi-billions announced for building out data center infrastructure to power AI LLMs. (Most of those will never be built as this technology levels off.)
One of my favorite analysts, Samantha Mou, Senior Analyst at market intelligence firm Interact Analysis provides comments regarding the announcement which I find relevant.
FANUC America’s $90M investment is part of a growing trend where robot manufacturers are bringing production closer to key markets, and the US is becoming a critical destination. Interact Analysis expects the industrial robot market here to see steady growth over the next five years, driven by reshoring initiatives and policies like tariffs, which are forcing robot makers to rethink their manufacturing strategies.
FANUC isn’t alone in this shift. Just last year, Yaskawa attracted attention by announcing plans for US-based production for robots and motion control components. As the largest robot supplier in the U.S. by market share, FANUC’s push toward local production aligns naturally with its market leadership and customer proximity strategy.
That said, questions remain about the depth of localization. It is possible that the new facility will primarily support assembly instead of full-scale manufacturing. Given that FANUC produces its core motion control components in Japan, and with limited domestic supply of key parts such as precision gearboxes in the US, it is likely that critical components will continue to be imported, with final robot assembly conducted locally.
I’m always happy to see news of investment in manufacturing. But experience has made me skeptical about the real impact. We can hope.
Click on the Follow button at the bottom of the page to subscribe to a weekly email update of posts. Click on the mail icon to subscribe to additional email thoughts.
We met in a conference room at an office in Barrington, IL. A place where sometime later a couple guys thought they’d screw me in a business deal. I came out ahead in the end, but the place has mixed memories.
This meeting involved thinking about the future of asset data and systems interoperability. We had a system diagram. The idea was to solve a huge problem for owner/operators of process manufacturing enterprises—flowing engineering data into other software systems for operations, maintenance, and enterprise. The incumbent system was a morass of paper (or pdf documents which was much the same thing).
We did trademark searches and domain name searches and eventually settled on the Open Industrial Interoperability Ecosystem—OIIE.
I plot this history for context for the conference I attended recently—the 2nd ADIF Workshop at Texas A&M University dubbed Driving Asset Data and Systems Interoperability Toward an Open and Neutral Data Ecosystem.
This workshop brought together owner/operators, EPCs, System Integrators, university researchers, standards organizations, and software vendors. Each group conducted a panel discussion of its needs and successes. I was there for a short presentation and to moderate the standards panel.
Professor David Jeong from Texas A&M and the session leader previewed the discussions. One of his colleagues later presented research his team has performed to provide a method for taking P&ID documentation into a standard format usable by other software systems.
The message that came to me from the panel of owner/operators (grossly summarized, as will be all the discussions) included two key words—collaborate and operationalize. They are impatient about solving this data interoperability problem. One panelist quipped, “We know the project is finished when the large van backs into the loading dock and disgorges mountains of paper.”
What blows my mind is that I was moved to a position called Data Manager in 1977 to tackle the (much smaller) mountain of paper our product engineering department provided to operations, accounting, and inventory management. I led a digitalization effort in 1978 to tackle the problem. The problem not only remains, but it is immensely more complicated and critical.
The EPCs basically said that their hands were tied by the owner/operators mandating which design and engineering software to use and the inflexibility of the vendors of said design and engineering software. When owner/operators had requested digital documentation, they had responded with pdfs. Hardly interoperable data.
Our standards panel included the leader of DEXPI, whose organization has developed a method of changing P&ID data into an xlsx (Excel) format. That, of course, is a good start.
An organization called CFIHOS (see-foss) presented their take on standards. I’m afraid I got a bit lost in the slides (note: more research needed). What I gathered was that they were attempting one overriding standard—and that that work was years away. Interesting that I listened to Benedict Evans’ podcast this morning. He is a long-time tech industry analyst. He remarked in another context, “It seems that where there are 10 standards and someone comes along with a standard to encompass them all, you wind up with 11 standards.”
The ISA-95 was presented. This messaging (and more) standard is incorporated with the OIIE, which was presented next. Dr. Markus Stumptner of the University of South Australia presented his research work on proof of concept of the OIIE.
If we can get enough momentum focusing on this area and find some SIs willing to take the OIIE to an owner/operator, perhaps we can finally prove the business case of asset data and systems interoperability.
Click on the Follow button at the bottom of the page to subscribe to a weekly email update of posts. Click on the mail icon to subscribe to additional email thoughts.
I’ve become a bit cynical about all the investment announcements by companies currying political favor (Hi, Tim Cook) wondering if these are merely publicity events. We have seen a number of big announcements in the past that have not paid off in actual manufacturing.
Hitachi Energy forwarded a release announcing a $106 million investment to expand transformer component manufacturing capacity at a facility in Alamo, Tennessee. The spokesperson assured me that they were actually breaking ground today and that real production would commence when completed.
I hope so.
The news in brief.
Project will create approximately 100 new jobs in Crockett County; leverages support from the State of Tennessee
Investment supports fast-growing demand for transformer components such as HVDC and dry-type bushings from Hitachi Energy
An additional 60,000 sqft will be added to the manufacturing facility to create room for more production lines to boost capacity and support customer demand
The Alamo expansion will significantly boost production of transformer bushings, including dry bushings for high-voltage direct current (HVDC) and alternating current (AC) transmission systems up to 800 (kilovolts) kV. Once complete, the site will become the largest bushings manufacturing facility in North America, and one of the largest in the world. This investment recognizes the Alamo facility’s strategic importance in the U.S. domestic supply chain for power transmission infrastructure. The components manufactured in Alamo will support the growing need for long-distance power transmission to enable the rapid expansion of data centers, AI infrastructure, and other electrifying industries.
The project adds over 60,000 square feet to the facility, including over 35,000 square feet for manufacturing, 20,000 square feet of warehouse space, and 5,000 square feet of new office space. The expanded site is expected to be operational by mid-2027 and will create over 100 new jobs in West Tennessee, offering long-term employment opportunities in engineering, skilled trades, and operational support roles.
The project includes the implementation of an integrated logistics center and vertically integrated machine shop, enabling just-in-time delivery, reducing external warehousing needs, and enhancing operational efficiency. The added office space supports recruiting and workforce development in line with long-term growth plans.
By increasing the business’s total HVDC bushing capacity and enhancing global redundancy, the expansion de-risks single-site operations while supporting technology shifts in the market for dry bushings supporting IEEE standards.
This project builds on Hitachi Energy’s broader $6 billion USD global investment strategy announced earlier this year, which includes $1.5 billion USD dedicated to scaling transformer manufacturing worldwide. As the world’s largest transformer manufacturer, Hitachi Energy is advancing grid modernization with industry-leading solutions and a commitment to achieving carbon-neutral operations by 2030.
Time to first production. This mantra has stayed with me since the time a company I worked for built automated assembly machines. This story tells of how some new engineering tools from Festo and additional assembly services enabled a company called CODI Manufacturing to meet a customer’s immediate need.
CODI Manufacturing introduced today a compact, flexible 12-cycle-per-minute case packer for small to mid-sized food and beverage companies that achieves an industry leading price/performance ratio.
The company utilized the Festo Handling Guide Online (HGO) engineering tool to design the two-axis handling system, the core of the new machine’s package handling. By having Festo assemble and ship a bolt-in-ready handling system, the company was able to build and commission the new machine in just four weeks.
“This machine is incredibly flexible,” said Jared Jones, Chief Operating Officer, CODI Manufacturing. “All the end user has to do is swap out the gripper head to change the number, type, and size of bottles, cans, jars, or pouches being packed into shipping cases, select a new recipe on the HMI, and start the machine.”
The CODI design features:
Festo’s new multiprotocol CMMT-AS servo drives and matching EMMT-ST one-cable motors
New CPX-AP-A configurable remote I/O terminal
New VTUX valve terminal
A simplified motion series (SMS) low-cost electric-linear actuator
The CDPX IP67 machine mounted HMI
Festo robust linear axes
The Festo CPX-CEC controller with EtherCAT communications for advanced motion control.
Collaboratively, Festo Engineering and CODI developed a pneumatic gripping head that provides feedback on successful gripping of each bottle. “Customers are amazed at the advanced motion automation technology on this machine for the price point,” Jones said. “It’s striking how much state-of-the-art automation is packed into such a small footprint machine.”
CODI Manufacturing had wanted to build a compact case packer for small to mid-sized companies for several years. Recently, CODI was four weeks into an eight-week project when Jones and the customer discussed adding a case packer to the order. “When I mentioned a price, the customer said he’d buy the machine if we could deliver it in four weeks with the rest of his order,” recalled Jones. “I said we’d try.”
The core of the machine would be a two-axis handling system used to grip and lift six bottles off one conveyor and place them into a box on an adjacent conveyor. The CODI team used the Festo HGO engineering tool to specify the system where the axes, servos, motors, mounting plates, and accessories were all properly sized and interoperable. The design session took minutes, saving CODI several days of engineering time.
“One of my lead controls engineers came into my office and said that Festo would supply a bolt-in-ready handling system at basically the same cost as buying and assembling the pieces and parts,” Jones said. “There was simply so little price difference between assembled and unassembled that we went with assembled. We used CAD files from the HGO session to build the frame while we waited for the assembly. Three weeks later, the system arrived. We used eight bolts to attach it to the machine frame. For a multiple-axis system, this was the easiest and fastest assembly and commissioning we’d ever had and the lowest total cost.”
