Clone's robot hand has artificial muscles
PLUS: Samsung's national AI gambit and a sweet potato harvesting bot
Welcome back to your Robot Briefing
Clone Robotics unveiled a robotic hand powered by artificial muscles that mimics human grip strength and speed, using a neural network trained on real hand movements to control 27 degrees of freedom with remarkable precision.
With carbon-fiber bones and water-powered synthetic muscles that survived 650,000 test cycles, this could be the breakthrough that finally gives humanoid robots the dexterity they need to handle everyday objects—or just another impressive prototype that struggles outside the lab.
In today's Robot update:
The Uncanny Hand
Snapshot: Polish robotics firm Clone Robotics demoed a hyper-realistic robotic hand that uses artificial muscles and a neural network controller to achieve human-like speed and grip strength, tackling one of the biggest challenges in humanoid dexterity.
Breakdown:
Takeaway: This hand represents a meaningful step toward humanoid robots that interact with objects as naturally as people do. Clone plans to integrate this technology into only 279 units of their Alpha humanoid, suggesting they're focusing on refinement over mass production.
Korea's Robot Gambit
Snapshot: South Korea launched a government-led manufacturing alliance that signed a partnership with Seoul National University to develop foundational AI models for humanoid robots, autonomous vehicles, and smart factories, aiming to generate over $67 billion in economic value by 2030.
Breakdown:
Takeaway: This represents one of the most ambitious national-level efforts to coordinate industry, academia, and government resources around humanoid robotics and autonomous systems. South Korea's centralized approach could accelerate development timelines that typically take individual companies years to achieve independently.
The EV-to-Robot Pipeline
Snapshot: Chinese electric vehicle makers are aggressively entering the humanoid robotics market, with Xpeng's Iron robot leading the charge toward planned mass production in late 2026. This pivot leverages their manufacturing expertise and AI development to compete in what they see as the next frontier of artificial intelligence.
Breakdown:
Takeaway: The shift from EVs to robotics shows how Chinese automakers view physical AI as a natural extension of their autonomous driving and manufacturing capabilities. With over two million robots already operating in Chinese factories, these companies are positioning themselves to dominate both the industrial and consumer robotics markets globally.
The Sweet Potato Picker
Snapshot: Researchers at Hainan University developed SPECNet, an AI vision model that enables agricultural robots to identify and harvest sweet potatoes in complex field environments with a 78% success rate. The system processes images at 35 frames per second, making it fast enough for real-world deployment.
Breakdown:
Takeaway: This vision system addresses a critical gap in agricultural automation for crops grown in challenging terrain where traditional machinery can't reach. The model's ability to maintain consistent performance across varying light and soil conditions makes it particularly valuable for real-world farming operations that can't control environmental factors.
Other Top Robot Stories
MarketsandMarkets projects the agricultural robot market will surge from $17.73 billion to $56.26 billion by 2030—a 217% expansion driven by labor shortages, precision farming demand, and automation adoption across major segments including robotic prostatectomy systems, partial nephrectomy platforms, and farm produce handling.
Distalmotion raised $150 million in Series G funding to accelerate U. S. commercialization of its DEXTER surgical robot, targeting the rapidly growing ambulatory surgery center market with a mobile platform that fits any operating room without modifications and keeps surgeons at the patient's bedside.
MindOn trained the Unitree G1 humanoid to perform household chores like watering plants, closing curtains, and tidying up with natural fluid movements—notably without any teleoperation, marking a shift toward fully autonomous home robots that can handle tasks requiring sensitivity and dexterity.
🤖 Your robotics thought for today:
What's a collaboration between humans and robots that doesn't exist yet—but would make you genuinely excited to pioneer it?
Tell me – what do you think?
Until tomorrow,
Uli
