Physical artificial intelligence is progressing at an exponential rate. Autonomous or semi-autonomous robots are already used in warehousing, logistics, mining, construction and farming. Several manufacturers are planning to launch commercially available humanoids in the near term, including for consumers. There are even real-world examples of so-called “dark factories,” manufacturing facilities with minimal human oversight in which robots navigate using machine vision technologies. These sites produce smartphones, electric vehicles, air conditioners, and appliances, among other things.
The robotics boom will be especially profound in the industrials sector. Industrials companies are grappling with shortages in skilled labor that are only projected to increase in the coming years. For example, from January to March, 2026, there were more than 440,000 to 510,000 vacant manufacturing positions in the U.S. That gap is expected to increase to between 1.5 million and 2 million vacant positions by the early 2030s.
To manage labor shortages and stay competitive, labor-intensive industrials businesses will become more capital-intensive as autonomous robots augment, replace or change traditional labor roles. The extent of that shift on the labor market is unclear. Robots might replace some roles, but will also fill existing labor vacancies or hours where humans aren’t available, such as the midnight shift. There may be offsetting new skilled roles created as well, such as for robot technicians, AI training and oversight roles, and data and system security.
This trend towards reindustrialization will fundamentally change the balance sheet of industrials companies and, with it, possibly trigger a credit and financing boom for the industrials sector. As industrials companies automate, they will be shifting operating expense into capital expenditures, which should result in more physical assets (robots) and thus more collateral in the business. This creates a feedback loop: The increased collateral value should make these businesses more attractive to lenders and other financing sources, which will in turn provide these businesses more financing to reindustrialize further and thus become more creditworthy.
The underlying issue, though, is how to prepare for this transformation. Questions remain about what a robotic future will look like, including what robots will be available, how they’ll be owned and operated, what their relative costs will be, what maintenance they’ll need, and what efficiency gains they’ll produce. In addition, a robot’s value as an asset may depend on its safety record, software, cybersecurity, data rights, maintenance arrangements and ability to operate reliably within a company’s compliance and safety programs.
Following are a few key variables that will be particularly relevant for industrials companies, their balance sheets, financing capability and governance and risk management over the next five to 10 years, as they build out their robotics capabilities.
Ownership models: Own or lease? The ownership question will largely be driven by how expensive owning a robot is compared to leasing one over its useful life span. If robots become affordable enough, then industrials companies will likely purchase them outright. All things being equal, owning the robots will strengthen the balance sheet and add to the available collateral for a lender, which may provide a business with better access to credit.
In contrast, some manufacturers or intermediaries may operate and lease out fleets of robots under a “robot-as-a-service” (RaaS) or leasing model, similar to how vehicles are often acquired. That might be attractive to an industrials business that doesn’t have the cash or financing capability to purchase a large fleet of robots, or simply wants to avoid a larger upfront capex spend. An industrials business that cuts its costs significantly by leasing a fleet of robots might also have a compelling story to tell as a “capital-light” business, which could be attractive to lenders lending against cash flow or recurring revenue.
Obsolescence: What is a robot’s lifespan? In the first few years of the robotic roll-out, there might not be good enough data to predict the useful life span of a robot. Will, for example, a specific robot become obsolete within three, five, seven or more years?
If the effective lifespan of a robot is short (three years, for example), there will likely not be much secondary resale value or collateral value to a lender. Even if a business is investing into capex and purchasing robots, the replacement costs could be so high as to counteract much of the benefit of a stronger balance sheet, in which case the business reality might be closer to an RaaS model.
Obsolescence risk also affects contracting. Buyers may seek upgrade commitments, maintenance obligations, parts availability, cybersecurity patching, software support and clear end-of-life terms to manage obsolescence risk.
Specialization: Can the robots be repurposed? The more general purpose a robot is, the more collateral and resale value it should have. Intuitively, an industrial-scale robotic arm that’s tuned to a specific factory setting might not command much of a secondary market price if it can’t be repurposed for other factory settings or industries, or because there’s a smaller pool of buyers for the technology.
If the robotics industry moves toward general-purpose autonomous robots, then businesses and lenders would have a broad, fungible market to resell goods. A humanoid working in a warehouse might be re-deployed across any industry or even take care of a household (folding clothes or doing laundry, for example). That’s unlike most physical capital used as collateral today, such as industry-specific machines or assembly lines, and would again lower the relative risk profile of a robot compared to other forms of collateral.
Operationalizing data: Is the underlying data valuable? Most of the attention for robots or physical AI applications goes to the “physical” portion of the equation. These machines also have an AI component and are a product of intensive training over large and often proprietary datasets.
Robots will need to begin with a baseline set of capabilities, but each industry and company will have its own processes, production methods and workflows for manufacturing products. These processes could constitute trade secrets which are protected under relevant law, but it’s difficult to measure their value, since much of it has historically been tied to the specific people who know how to implement them. By contrast, robots can take that “secret” human knowledge and transform it into operational data that can be applied to similar businesses, though adequate safeguards would be needed to avoid disclosing the data in a way that compromises the trade secret protection.
Its value will depend largely on who owns the deployment data — is it the robot’s manufacturer, owner, or someone else? Companies should also understand whether deployment data might be shared with vendors, used to improve generalized models, or commingled with data from other buyers.
With these unknowns, how do industrials businesses and lenders prepare for robotic reindustrialization?
On the industrials side, businesses must be able to clearly demonstrate the financial and technological value of their use of autonomous robots to stakeholders and lenders. That requires a full-fledged robotics strategy, including evaluating where and how to deploy robots, their relative costs, ownership models, financing options, risk management, and data ownership and use. These questions cut across finance, technology, operations, cybersecurity, procurement, safety, insurance and legal risk. Robotics and other subject-matter experts will need a seat in discussions alongside CFOs and CTOs.
Lenders, for their part, will need to develop, recruit or in-source the technical expertise to understand the relative risk of this wave of reindustrialized, robotics-driven businesses, and build a robotics-focused diligence framework. They’ll also need to evaluate not only cash flow and collateral value, but also the useful life of robotic assets, resale market depth, maintenance arrangements, data value and control, and the borrower’s ability to deploy the technology safely. Lenders who can make this shift and appraise and price the risk competitively but responsibly will be in the best position to succeed. Those who are too aggressive could take on bad deals, while those who are too conservative might miss out.
Most importantly, both sides will need to adapt. This is perhaps one of the largest credit and financing opportunities for industrials companies in modern history, but threading the needle will require developing a disciplined robotics strategy and expertise.
Edmund Mokhtarian is a partner with the law firm of DLA Piper.
