No brain, no problem: What robots can learn from sea stars
USC Kanso Bioinspired Motion Lab borrows a trick from nature’s toolkit that can be applied to optimize robot locomotion.
Ever feel run off your feet? Spare a thought for sea stars, creatures whose movement involves the coordination of hundreds of tiny tube feet to navigate complex environments – despite the lack of a central “brain.”
In other words, it’s as though each foot has a mind of its own. For Kanso Bioinspired Motion Lab, based within the USC Viterbi School of Engineering Department of Aerospace & Mechanical Engineering, sea stars pose an intriguing phenomenon. Kanso Lab specializes in decoding the flow physics of living systems, often applying those insights to inform developments in robotics.
Now, researchers at USC are uncovering the secret behind this decentralized locomotion. This could revolutionize how we design autonomous robots.
One thought per foot
The lab’s recent paper in PNAS, “Tube feet dynamics drive adaptation in sea star locomotion” (January 13, 2026), reveals that the movement of sea stars is directed by local feedback from individual tube feet, each dynamically adjusting their adhesion to the surface in response to varying degrees of mechanical strain.
“We began working on sea stars with McHenry Lab at UC Irvine, and later partnered with biologists at the University of Mons in Belgium,” said Eva Kanso, director of Kanso Lab and professor of aerospace and mechanical engineering, physics and astronomy. “Together with Associate Professor Sylvain Gabriele and graduate student Amandine Deridoux at the SYMBIOSE Lab, we designed a special 3D-printed “backpack” for the sea star. By loading and unloading the backpack, we could observe and measure how each tube foot responded to the added weight.”
What did the researchers discover? Each foot responded independently to changing loads. “From the outset, we hypothesized that sea stars rely on a hierarchical and distributed control strategy, in which each tube foot makes local decisions about when to attach and detach from the surface based on local mechanical cues, rather than being directed by a central controller,” said Kanso.
The experiments allowed the team to test and quantify these local responses. “At USC, we developed a mathematical model showing how simple, local control rules, coupled through the mechanics of the body, can give rise to coordinated, whole-animal locomotion.”
Sea star locomoting across a glass surface, picture from below, showing the attachment and detachment of tube feet. Image credit: McHenry Lab at UC Irvine.
No brain, no problem
This model for adaptive movement based on local feedback is highly relevant to the design of soft and multi-contact robotics. Potential application on land, under water and even on other planets, include decentralized locomotion systems for robots navigating uneven, vertical and upside-down terrain –environments that prevent consistent communication from a central “mission control” or human decision-maker. No brain? No problem.
“We also conducted experiments in which we turned the sea star upside-down – the morphology of the tube feet allows the sea star to continue to move,” said Kanso. “Just imagine if you were doing a handstand. Your nervous system would immediately let you know that you were in a position opposed to gravity. But a sea star has no such collective recognition.”
Robustness through redundancy
Instead, the sea star is equipped with the local knowledge of each tube foot experiencing the force of gravity differently. Coordinated movement is due to the fact the feet are mechanically linked to the body; when one foot pushes, the movement affects other feet. As a result, local failures do not necessarily halt the whole system – allowing for advanced robustness and resilience.
That’s a significant advantage for autonomous robots navigating extreme environments, liable to flip, lose or gain load, or be disconnected from central communication source. While fast-moving animals (from insects to gymnasts) rely on “central pattern generators” – specialized neural circuits located in the brainstem that produce rhythmic motor patters – slow-moving sea stars are primed to adapt dynamically to environmental changes.
So, it turns out there are some perks to being brainless. Whether a sea star is navigating tidal forces, currents or varying terrain roughness, they adapt and go with the flow.
Method of Research
Experimental study
Subject of Research
Animals
Article Title
Tube Feet Dynamics Drive Adaptation In Sea Star Locomotion
SMART and NUS pioneer neural blueprint for human-like intelligence in soft robots
Singapore-MIT Alliance for Research and Technology (SMART)
image:
The soft robotic arm can safely operate close to the human body without causing discomfort or injury. The AI control system is well-suited for assistive scenarios like showering, where the arm can help wipe the back — supporting people with limited mobility and easing the load on caregivers
view moreCredit: National University of Singapore (NUS)
A newly developed AI control system using neuron-inspired learning enables soft robotic arms to learn a broad set of motions once and adapt instantly to changing conditions without retraining
Inspired by the way the human brain learns, this system is one of the first to achieve three aspects needed to deploy soft robots in real-world environments — learning capabilities that can be generalised across tasks, the ability to maintain performance under diverse disturbances, and a metric that enables stability during adaptation
Validated across multiple platforms, this innovation paves the way for real-world applications across diverse industries, including healthcare, manufacturing, assistive robotics and more
Singapore, 5 February 2026 – Singapore-MIT Alliance for Research and Technology’s (SMART) Mens, Manus & Machina (M3S) interdisciplinary research group, and National University of Singapore (NUS), alongside collaborators from Massachusetts Institute of Technology (MIT) and Nanyang Technological University (NTU Singapore), have developed an AI control system that enables soft robotic arms to learn a wide repertoire of motions and tasks once, then adjust to new scenarios on the fly without needing retraining or sacrificing functionality. This breakthrough brings soft robotics closer to human-like adaptability for real-world applications, such as in assistive robotics, rehabilitation robots, and wearable or medical soft robots, by making them more intelligent, versatile and safe.
Unlike regular robots that move using rigid motors and joints, soft robots are made from flexible materials such as soft rubber and move using special actuators – components that act like artificial muscles to produce physical motion. While their flexibility makes them ideal for delicate or adaptive tasks, controlling soft robots has always been a challenge because their shape changes in unpredictable ways. Real-world environments are often complicated and full of unexpected disturbances, and even small changes in conditions – like a shift in weight, a gust of wind or a minor hardware fault – can throw off their movements.
Despite substantial progress in soft robotics, existing approaches often can only achieve one or two of the three capabilities needed for soft robots to operate intelligently in real-world environments: using what they’ve learned from one task to perform a different task, adapting quickly when the situation changes, and guaranteeing that the robot will stay stable and safe while adapting its movements. This lack of adaptability and reliability has been a major barrier to deploying soft robots in real-world applications until now.
In a study titled ‘A general soft robotic controller inspired by neuronal structural and plastic synapses that adapts to diverse arms, tasks, and perturbations’, recently published in Science Advances, the researchers describe how they developed a new AI control system that allows soft robots to adapt across diverse tasks and disturbances. The study takes inspiration from the way the human brain learns and adapts and was built on extensive research in learning-based robotic control, embodied intelligence, soft robotics and meta-learning.
The system uses two complementary sets of “synapses” – connections that adjust how the robot moves – working in tandem. The first set, known as “structural synapses”, is trained offline on a variety of foundational movements, such as bending or extending a soft arm smoothly. These form the robot’s built‑in skills and provide a strong, stable foundation. The second set, called “plastic synapses”, continually updates online as the robot operates, fine-tuning the arm’s behaviour to respond to what is happening in the moment. A built-in stability measure acts like a safeguard, so even as the robot adjusts during online adaptation, its behaviour remains smooth and controlled.
“This new AI control system is one of the first general soft-robot controllers that can achieve all three key aspects needed for soft robots to be used in society and various industries. It can apply what it learned offline across different tasks, adapt instantly to new conditions and remain stable throughout — all within one control framework,” said Associate Professor Zhiqiang Tang, who was a Postdoctoral Associate at M3S and at NUS when he carried out the research, is the first and co-corresponding author of the paper, and is now Associate Professor at Southeast University (SEU China).
“Soft robots hold immense potential to take on tasks that conventional machines simply cannot, but true adoption requires control systems that are both highly capable and reliably safe. By combining structural learning with real-time adaptiveness, we’ve created a system that can handle the complexity of soft materials in unpredictable environments. It’s a step closer to a future where versatile soft robots can operate safely and intelligently alongside people — in clinics, factories, or everyday lives,” said Professor Daniela Rus, Co-lead Principal Investigator at M3S, Director - Computer Science and Artificial Intelligence Laboratory (CSAIL) at MIT, and co-corresponding author of the paper.
The system supports multiple task types, enabling soft robotic arms to execute trajectory tracking, object placement and whole-body shape regulation within one unified approach. The method also generalises across different soft-arm platforms, demonstrating cross-platform applicability.
The system was tested and validated on two physical platforms – a cable-driven soft arm and a shape-memory-alloy–actuated soft arm – and delivered impressive results. It achieved a 44–55% reduction in tracking error under heavy disturbances, over 92% shape accuracy under payload changes, airflow disturbances and actuator failures, and stable performance even when up to half of the actuators failed.
“This work redefines what’s possible in soft robotics. We’ve shifted the paradigm from task-specific tuning and capabilities toward a truly generalisable framework with human-like intelligence. It is a breakthrough that opens the door to scalable, intelligent soft machines capable of operating in real-world environments,” said Professor Cecilia Laschi, Principal Investigator at M3S, Provost’s Chair Professor, Department of Mechanical Engineering at the College of Design and Engineering and Director of the Advanced Robotics Centre at NUS, and co-corresponding author of the paper.
This breakthrough opens doors for more robust soft robotic systems to develop manufacturing, logistics, inspection and medical robotics without the need for constant reprogramming – reducing downtime and costs. In healthcare, assistive and rehabilitation devices can automatically tailor their movements to a patient’s changing strength or posture, while wearable or medical soft robots can respond more sensitively to individual needs, improving safety and patient outcomes.
The researchers plan to extend this technology to robotic systems or components that can operate at higher speeds and more complex environments, with potential applications in assistive robotics, medical devices and industrial soft manipulators, as well as integration into real-world autonomous systems.
The research conducted at SMART was supported by the National Research Foundation Singapore under its Campus for Research Excellence and Technological Enterprise (CREATE) programme.
###
About Mens, Manus and Machina (M3S)
M3S is an interdisciplinary research group (IRG) launched in July 2023 by SMART, MIT’s research enterprise in Singapore. Inspired by MIT’s motto of mens et manus (“mind and hand”), the programme aims to promote AI and machine use for practical applications through an intersectional approach. The research at M3S addresses critical questions concerning the design of technology, the development of human skills, and the adaptation of institutions and social structures to effectively navigate the transformative impact of AI, automation and robotics. By exploring the intricate interplay between human capabilities, emerging technologies and societal structures, M3S seeks to drive scientific, societal and commercial impact that will pave the way for the design of inclusive, resilient and innovative solutions that empower individuals, institutions and cities in Singapore and beyond.
For more information, visit https://m3s.mit.edu/.
About Singapore-MIT Alliance for Research and Technology (SMART) [新加坡-麻省理工学院科研中心]
Singapore-MIT Alliance for Research and Technology (SMART) is MIT’s Research Enterprise in Singapore, established by the Massachusetts Institute of Technology (MIT) in partnership with the National Research Foundation of Singapore (NRF) since 2007. SMART is the first entity in the Campus for Research Excellence and Technological Enterprise (CREATE) developed by NRF. SMART serves as an intellectual and innovation hub for research interactions between MIT and Singapore. Cutting-edge research projects in areas of interest to both Singapore and MIT are undertaken at SMART. SMART currently comprises an Innovation Centre and six Interdisciplinary Research Groups (IRGs): Antimicrobial Resistance (AMR), Critical Analytics for Manufacturing Personalized-Medicine (CAMP), Disruptive & Sustainable Technologies for Agricultural Precision (DiSTAP), Mens, Manus and Machina (M3S), Wafer-scale Integrated Sensing Devices based on Optoelectronic Metasurfaces (WISDOM) and Wearable Imaging for Transforming Elderly Care (WITEC).
SMART research is funded by the National Research Foundation Singapore under the CREATE programme.
For more information, please visit http://smart.mit.edu
About National University of Singapore (新加坡国立大学)
The National University of Singapore (NUS) is Singapore’s flagship university, which offers a global approach to education, research and entrepreneurship, with a focus on Asian perspectives and expertise. We have 15 colleges, faculties and schools across three campuses in Singapore, with more than 40,000 students from 100 countries enriching our vibrant and diverse campus community. We have also established more than 20 NUS Overseas Colleges entrepreneurial hubs around the world.
Our multidisciplinary and real-world approach to education, research and entrepreneurship enables us to work closely with industry, governments and academia to address crucial and complex issues relevant to Asia and the world. Researchers in our faculties, research centres of excellence, corporate labs and more than 30 university-level research institutes focus on themes that include energy; environmental and urban sustainability; treatment and prevention of diseases; active ageing; advanced materials; risk management and resilience of financial systems; Asian studies; and Smart Nation capabilities such as artificial intelligence, data science, operations research and cybersecurity.
For more information on NUS, please visit nus.edu.sg.
The 160g soft robotic arm with a 37.2 g soft gripper performs a pick-and-place task with a 56.4 g object, bending smoothly to grasp, lift and reposition the object in a controlled motion. The combined payload is 58.5% of the arm’s mass, demonstrating stable manipulation under a relatively high payload while maintaining compliant, precise operation
The AI control system enables the soft robotic arm to learn and achieve precise movements, such as bending into a curved ‘C’ shape, similarly to how a human arm bends. In an anti‑disturbance test, the arm was challenged under fixed and continuously changing fan speeds, and still achieved the target shape with 93.8% accuracy under the most challenging scenario
Credit
SMART M3S
Journal
Science Advances
Article Title
A general soft robotic controller inspired by neuronal structural and plastic synapses that adapts to diverse arms, tasks, and perturbations
Article Publication Date
7-Jan-2026
