[New York = E-Daily Seong Joowon Correspondent] As artificial intelligence (AI) expands beyond the digital realm into the real world as “physical AI,” forecasts suggest that the commercialization of the robotics industry will gain full momentum within the next two to three years. However, some have pointed out that robots currently operating in laboratories must demonstrate high reliability and cost-effectiveness before they can be deployed in actual industrial settings.

Marc Weinstein, Chief Operating Officer (COO) of RoboStrategy—a Nasdaq-listed investment fund specializing in robotics and physical AI—met with Edaily on the 13th (local time) at the Knickerbocker Hotel near Times Square in New York and “Robots specialized for specific tasks could see widespread commercialization within two to three years,” he predicted. He estimated that it would take about five years for general-purpose robots capable of performing a variety of tasks to become commercially available.
COO Marc Weinstein is a financial expert who worked in investment banking at Morgan Stanley and Jefferies, served at the digital asset investment firm Wave Financial, and later served as a General Partner (GP) at Mechanism Capital. He majored in economics at the University of Pennsylvania’s Wharton School and, as a co-founder of RoboStrategy, oversees investment strategy and capital markets operations.
He cited advancements in AI models, falling hardware prices, and demographic shifts as the driving forces behind the rapid growth of the robotics industry. “Robots are approaching full autonomy, capable of performing various tasks even in unfamiliar environments,” explained COO Weinstein. “As hardware costs continue to decline, the cost of robotic labor will ultimately fall below that of human labor.”
In particular, he predicted that concerns over labor shortages due to aging populations and retirements in South Korea, the U.S., and Europe would drive demand for robots.
Robostrategy, a closed-end investment fund that went public on the Nasdaq last May, invests in private robotics companies such as Figure AI, Dyna Robotics, and Standard Bots. The firm hosted a robotics event in New York that day, attended by executives from its major portfolio companies.
Weinstein, the COO, emphasized the ability to generate actual revenue as the most important criterion for robotics investments. “It’s not enough to stop at technology demonstrations or letters of intent; what matters is whether actual revenue is being generated from customers using robots in industrial settings,” he said.
He also issued a warning about investment risks. “While shares in publicly traded funds are tradable, venture investments take much longer,” he explained, adding, “Stock prices can fluctuate significantly regardless of the net asset value of the underlying assets or the growth of the investee companies.” This means investors should focus on a 5- to 10-year investment horizon rather than short-term gains.
In fact, as of the end of September, StandardBots, an industrial robotics company, accounted for 31.3% of RoboStrategy’s net assets. COO Weinstein highly rated the growth potential of the industrial robotic arm market amid the trend of U.S. manufacturing returning to the country. He expected that concentration risk would be mitigated as the number of investment targets increases in the future.

Tao Chen, CEO of Dexmate—another U.S. robotics manufacturer in which RoboStrategy has invested—identified reliability as the biggest challenge in robot commercialization. In an interview that day, he said, “In university labs, a success rate of 60–70% may be considered a meaningful achievement, but to supply products to actual customers, a success rate of 99.99% is required.”
CEO Chen is a researcher-turned-founder who earned a Ph.D. in robot learning from the Computer Science and Artificial Intelligence Laboratory (CSAIL) at the Massachusetts Institute of Technology (MIT). He holds a master’s degree in robotics from Carnegie Mellon University and has conducted research on robot object manipulation, mobility, and navigation technologies.
He explained that it would be difficult to scale the business if significant engineering resources had to be allocated every time a robot was deployed in a new customer’s work environment. He emphasized that AI models must evolve to be applicable to a wide range of tasks and be capable of learning new tasks with only a small amount of additional data.
According to CEO Chen, DexMate has currently supplied hundreds of robots worldwide. Customers—primarily in manufacturing and logistics warehouses—are utilizing robots or testing their implementation for tasks such as parts sorting, task-specific bundling, box handling, and item retrieval.
Collaboration with Korean companies is also drawing attention. #LG CNS made a strategic investment in U.S.-based DexMate last March and is currently conducting technical validation for industrial applications. In an interview that day, CEO Chen stated, “LG Corp. is using our robots and is also assisting with distribution in Korea.”
CEO Chen predicted that over the next one to two years, the application of AI-based robots to specific tasks in the manufacturing and logistics sectors will advance, and that over the next three years, operational reliability and the ability to adapt to new environments will improve significantly. He emphasized, “Physical AI is not just a matter of AI models; it is a system-level challenge that requires solving both hardware and software together.”