A $60 million venture fund targeting physical AI startups has closed, backed by Silicon Valley-based Pegasus Tech Ventures and Japanese robotics manufacturer CYBERDYNE. The fund represents a deliberate bet that the next wave of artificial intelligence value creation will emerge from systems that interact with the physical world, not just digital interfaces. CYBERDYNE, known for its HAL exoskeleton and medical robotics platforms, brings manufacturing expertise and distribution channels across Asian healthcare markets to portfolio companies. Pegasus manages the fund and will lead investment decisions, while CYBERDYNE provides strategic support ranging from prototype testing facilities to introductions with hospital systems evaluating automation technologies.
Pegasus has operated as an active investor in robotics and automation for nearly a decade, with prior stakes in warehouse automation, agricultural robotics, and industrial vision systems. The firm typically writes initial checks between $500,000 and $3 million for seed and Series A rounds, then reserves capital for follow-on investments in breakout companies. This fund's structure differs from Pegasus's earlier vehicles by incorporating a strategic corporate partner from day one rather than adding strategic investors opportunistically. CYBERDYNE's involvement signals a shift in how established robotics manufacturers are approaching innovation, moving from occasional acquisitions toward continuous portfolio exposure. The Japanese firm has committed both capital and engineering resources to the fund, including access to its Tsukuba research campus where portfolio companies can test prototypes on CYBERDYNE's assembly lines.
The fund's thesis centers on what industry observers now call physical AI or embodied intelligence, systems where machine learning models control actuators, manipulators, or mobile platforms rather than merely processing data. Investment targets include humanoid robotics, assistive healthcare devices, logistics automation, and manufacturing systems that adapt to variable tasks without extensive reprogramming. Pegasus has identified gaps in commercialization support for robotics startups, particularly around hardware iteration cycles and regulatory pathways for medical or safety-critical applications. CYBERDYNE's experience navigating medical device approval in Japan, Europe, and the United States gives portfolio companies a roadmap through certification processes that often stretch two to three years. The fund also prioritizes startups building reusable robotics platforms rather than single-application products, reflecting investor preference for companies that can address multiple markets with core technology.
Healthcare automation represents a significant portion of the fund's intended allocation, driven by demographic trends and labor shortages across developed economies. Japan faces particularly acute challenges with an aging population and insufficient nursing staff, conditions that have accelerated adoption of assistive robotics in eldercare facilities. CYBERDYNE's HAL exoskeleton, which assists patients with mobility impairments, has gained traction in over 400 Japanese medical institutions and serves as a template for the type of products Pegasus hopes to back. The fund will also consider investments in surgical robotics, rehabilitation devices, and diagnostic systems that combine imaging hardware with AI interpretation. Beyond healthcare, manufacturing automation remains a priority, especially systems that can handle variable parts or operate safely alongside human workers without extensive safety cages or barriers. The recent maturation of vision-language models that allow robots to understand natural language instructions has opened new possibilities for flexible automation that Pegasus believes remains underinvested.
The fund's launch arrives as physical AI captures attention from larger technology companies and chip manufacturers racing to provide hardware optimized for robotics workloads. NVIDIA has promoted the concept of physical AI across multiple keynotes, framing it as the next frontier after large language models. Qualcomm has developed robotics-specific processors that integrate sensor fusion and motor control alongside neural network accelerators. However, venture capital deployment into robotics startups has lagged software AI investment by a considerable margin, with many investors deterred by longer development cycles and capital-intensive hardware iteration. Pegasus and CYBERDYNE are betting that this hesitancy has created attractive entry valuations for startups that can demonstrate technical progress and early commercial traction. The fund plans to deploy capital over three to four years, concentrating investments in North America, Europe, and Asia while maintaining flexibility to pursue standout companies regardless of geography. Portfolio construction will favor 15 to 20 core positions rather than a spray-and-pray approach, with meaningful reserves for follow-on rounds that allow Pegasus to maintain ownership in companies that achieve product-market fit.
What to Watch: Track whether Pegasus announces initial portfolio companies within the next 90 days, particularly any startups commercializing humanoid platforms or surgical robotics. Monitor CYBERDYNE's own product pipeline for clues about technical capabilities the firm seeks in startups. Watch for similar fund structures from other established robotics manufacturers looking to derisk innovation through venture portfolios rather than internal R&D alone. Observe whether robotics-focused funds can command premium valuations in the current fundraising environment where software-focused venture firms face pressure.
