Humanoid Robots for Healthcare

Humanoid robots are entering healthcare at a moment of acute structural crisis. The WHO projects a global shortfall of 10 million healthcare workers by 2030, U.S. nursing turnover exceeds 20% annually, and Japan's population is now over 29% elderly. These aren't cyclical labor market fluctuations—they're demographic certainties. The humanoid form factor offers something no wheeled robot or fixed manipulator can: the ability to navigate hospitals, clinics, and homes built for human bodies, using human tools and traversing human-scale spaces, without any facility redesign. In 2026, this promise is transitioning from research demonstration to commercial deployment, though the gap between what's shipping and what's envisioned remains substantial.

The State of Deployment: Logistics First, Care Later

The most mature healthcare humanoid deployment today is Diligent Robotics' Moxi, a mobile manipulator operating in over 25 U.S. hospitals with nearly 100 units deployed. Moxi has completed more than 1.25 million autonomous deliveries—medications, lab samples, PPE, and supplies—between pharmacy, central supply, and nursing stations. Clinical data shows nurses regain approximately 30 minutes per shift previously spent on supply runs. In October 2025, Diligent unveiled Moxi 2.0, featuring 10x the compute of its predecessor via the NVIDIA IGX Thor platform, with plans to scale to 15+ units per hospital site. The first Moxi 2.0 units are deploying to hospitals in H1 2026. In January 2026, Serve Robotics announced its acquisition of Diligent Robotics for $29 million, signaling consolidation in the healthcare robotics space and Serve's push from outdoor delivery into indoor hospital logistics.

This logistics-first approach reflects a pragmatic reality: non-contact supply delivery is technically achievable today, operates outside FDA medical device regulation, and addresses a genuine pain point. Direct patient contact—lifting, bathing, rehabilitation—requires orders of magnitude more capability and regulatory clearance that no humanoid has yet achieved.

China's Elder Care Offensive

China is pursuing humanoid elder care with characteristic scale and speed. In June 2025, China's government launched a national elderly-care robot pilot program requiring companies and research institutes to conduct at least six-month trials, deploying a minimum of 200 robots to 200 families. In Chengdu's Pacific Care Home, a humanoid robot named "Yang Yang" already handles morning wake-ups, weather updates, and daily activity reminders for residents.

Fourier Robotics is the most healthcare-focused Chinese humanoid company. Its GR-3, unveiled in August 2025 and debuted in the U.S. at CES 2026, is explicitly branded as a "care-bot"—standing 165 cm tall with 55 degrees of freedom and a proprietary Full-Perception Multimodal Interaction System integrating vision, audio, and tactile feedback for emotional recognition and natural dialogue. Priced above 200,000 yuan (~$27,500), GR-3 targets B2B healthcare institutions for companionship, health monitoring, and rehabilitation support. Fourier's heritage in rehabilitation exoskeletons gives it domain credibility that pure-play humanoid companies lack.

UBTECH Robotics launched a $20,000 consumer-grade humanoid home robot in May 2025, with Walker S2 demonstrations in Chinese elder care settings. Unitree sold over 5,500 humanoid robots in 2025—surpassing the combined output of all U.S. competitors—and is targeting 20,000 units in 2026, with household and elder care as key verticals alongside its sub-$20K G1 platform.

Rehabilitation and Clinical Research

Early clinical data from robot-assisted rehabilitation programs shows patients achieving 15–20% faster recovery in mobility metrics compared to traditional therapy alone, while therapists report significantly reduced physical fatigue. This is one of the most promising near-term clinical applications for humanoid robots because rehabilitation exercises are structured, repetitive, and predictable—exactly the kind of task current AI can handle reliably.

MIT's E-BAR (Elderly Bodily Assistance Robot), tested in 2025, represents the research frontier: a mobile robot designed as robotic handlebars that follows a person from behind, physically supporting the elderly and catching them if they fall. While not yet commercially deployed, it demonstrates the trajectory toward humanoid robots providing direct physical assistance in home settings.

On the surgical side, LEM Surgical's Dynamis system—showcased at CES 2026—holds FDA 510(k) clearance and is already in clinical use for spinal and orthopedic procedures. While not a general-purpose humanoid, it represents the convergence of humanoid dexterity with surgical precision that may define the next wave of operating room robotics.

The Regulatory and Safety Landscape

As of early 2026, there is no unified global regulatory framework for humanoid robots in healthcare. Robots performing logistics (non-contact with patients) generally fall outside FDA medical device regulation. Any robot touching patients—lifting, rehabilitation, surgical assistance—requires FDA clearance as a Class II or Class III medical device. This regulatory asymmetry explains why logistics deployments lead by years: companies can ship, iterate, and scale without navigating the 510(k) or PMA process.

The collapse of Aldebaran (manufacturer of Pepper and NAO) in February 2025 raised questions about long-term vendor viability. Hospitals that deployed Pepper for patient interaction and wayfinding now face uncertain support timelines—a cautionary tale about depending on robotics startups for clinical infrastructure.

What's Next: 2026–2028

The near-term trajectory is clear: humanoid robots will first saturate hospital logistics, then expand into structured rehabilitation and elder care companionship, with direct patient handling remaining 3–5 years away from commercial viability. The integration of computer vision and large language models is enabling more natural patient interaction, while advances in sim-to-real transfer are accelerating the training of physical care skills in simulation before deployment in sensitive clinical environments. The companies that win in healthcare humanoid robotics will be those that solve not just the engineering problem, but the trust problem—earning the confidence of clinicians, regulators, and patients that a machine can safely share intimate human spaces.