Organoids in motion: biohybrid robotics futures

1. Introduction A driving motivation for merging organoids with robotics is to overcome the limitations of each technology alone. Organoid technology has rapidly progressed as a cutting-edge tool to recapitulate organ complexity for research and clinical applications. Yet, organoids in isolation lack the vascular perfusion, mechanical stimuli, and system-level inputs of a living body, which can limit their maturation and long-term functionality. Soft robots, on the other hand, excel at simulating physiological motions and environments—from gentle pumping to peristaltic flows—but lack the cellular metabolism, sensing, and adaptive responses of living tissue.[1] The integration of organoids with soft robots could create biohybrid systems that are greater than the sum of their parts: robots endowed with the functional intelligence of living cells and organoids supported by the dynamic environment and control of robotics.[2] Such biohybrid systems may address unmet needs in medicine—for example, more predictive disease models that respond to drugs and stimuli in real time, smarter drug delivery devices that adapt to patient physiology, and active implants that promote tissue regeneration through living interfaces. In the following sections, we discuss current progress in organoid science and soft robotics, examine strategies for interfacing organoids with robots, highlight application opportunities across biomedicine, outline key technical hurdles, and present a forward-looking vision for the next decade of organoid-integrated biohybrid robotics (Figure 1).Figure 1.: Organoids integrated with soft robots enable dynamic biomedical applications, including disease modeling, drug screening, and tissue regeneration, while addressing challenges in mechanical integration and long-term maintenance (brain–computer interface [BCI]).2. Advances in organoid technology and soft robotics Organoid technology has seen tremendous advances over the past decade, yielding ever more complex and physiologically relevant mini-tissues. From tiny cerebral organoids that model aspects of human brain development, to cardiac organoids (cardioids) that beat spontaneously, to intestinal organoids that form crypt-villus structures, researchers can now generate a spectrum of mini-organs in vitro. This includes advanced techniques, such as realizing engineering large-scale self-mineralizing bone organoids and woven bone organoid formation.[3,4] These organoids capture key features of their in vivo counterparts—for instance, brain organoids exhibit diverse neuron types and network activity and gut organoids form polarized epithelial layers with nutrient absorption and secretion functions.[5,6] Indeed, brain organoids have provided unprecedented insights into neurodevelopment and disease, and they hold promise for drug screening in neurological disorders. Similarly, patient-derived tumor organoids are used to test cancer drugs ex vivo, and liver and kidney organoids model metabolic diseases and toxicity.[7] Efforts are underway to improve organoid maturity and scalability, addressing current limitations. For example, microfluidic and bioreactor systems provide perfusion to support larger organoids with sustained viability, and bioengineering strategies (such as embedding endothelial cells or biomaterial scaffolds) aim to vascularize organoids.[8] Additionally, combining multiple organoids to form “assembloids” is an emerging approach to model interactions between organs (eg, brain-region assembloids or gut–liver axis).[9] These developments reflect an increasing demand for organoids in disease modeling, drug testing, and even as building blocks for future organ replacements. In summary, organoids have evolved into sophisticated in vitro avatars of human organs, yet they typically remain static, lacking the dynamic mechanical and electrical cues of a living organism. Concurrently, soft robotic systems have matured to become powerful platforms for biomedical interaction. Soft robots are constructed from compliant materials (elastomers, hydrogels, textiles) and employ gentle actuation mechanisms (pneumatic, hydraulic, dielectric, magnetic, etc) that are inherently suited for interfacing with delicate biological tissues.[10] Their flexibility and adaptability enable capabilities ranging from minimally invasive surgical tools to implantable assistive devices. For instance, soft robotic catheters and endoscopes can navigate the body’s tortuous pathways for targeted interventions, and implantable soft actuators have been developed to augment organ function (such as a robotic sleeve that assists heart pumping or a diaphragmatic actuator that aids breathing in respiratory failure). A recent example demonstrated an implantable soft robotic ventilator wrapped around a pig’s diaphragm to effectively increase inspiratory motion.[11] Advances in stimuli-responsive materials (eg, shape-memory polymers, liquid crystal elastomers) and additive manufacturing are expanding the design space of soft robots, allowing custom geometries and programmable motions. Importantly, soft robots can operate in close contact with living tissues without causing damage, due to their compliant nature and high biocompatibility. This makes them ideal platforms to integrate with organoids. Moreover, the field of biohybrid robotics has begun to incorporate living cells as actuators or sensors in robots—for example, skeletal muscle strips or cardiac tissues have been used to power small locomoting devices.[12] These early biohybrid “biobots” showcase the potential synergy: living muscle cells provide efficient, self-healing actuators, while the engineered scaffold provides structure and direction. Notably, one milestone was the development of a biohybrid fish powered by human cardiac muscle cells, which achieved self-propelled swimming by replicating the heart’s mechanoelectrical feedback loops.[12] In this system, 2 layers of cardiomyocytes were arranged antagonistically; each contraction stretched the opposite layer, triggering it to contract in turn, producing an autonomous rhythmic motion. An embedded pacemaker-like node maintained the cycle, enabling the fish to swim continuously. This remarkable example, published in Science 2022, illustrates how insights from cardiac biology can inform robotic design and result in a self-regulating biohybrid machine.[12] It also hints at broader applications—such a system provides a platform to study cardiac physiology and diseases (like arrhythmias) in a dynamic setting. Together, these advances in organoid biology and soft robotics set the stage for organoid-integrated robotics. Organoids bring authentic biological function—beating heart tissue, firing neural circuits, and secreting gut epithelium—while soft robots contribute controllable actuation, structural support, and the ability to interact with the physical world. The following sections delve into how these components can be interfaced and harnessed, and what new capabilities emerge as a result. 3. Organoid–robot interfaces: signal integration and actuation strategies Creating a functional union between organoids and robots requires carefully engineered interfaces that allow 2-way communication: robots must deliver stimuli to organoids and receive signals or actions from them. Several modes of interface are conceivable, including mechanical coupling, electrical and chemical signal integration, and even optical or genetic control pathways. The overarching goal is to achieve seamless coordination such that the organoid and robotic components operate in concert as a single system. Here, we outline key interface and actuation strategies that underpin organoid–robot integration. 3.1 Mechanical interfaces and actuation One straightforward approach is to use the organoid as a living actuator or mechanical sensor within a soft robot. Contractile organoids, such as cardiac or muscle organoids, can produce forces or motion that a robot can harness. For example, a cardiac organoid attached to an elastic robotic scaffold could generate pumping or locomotion when it beats. Conversely, a soft robot can apply controlled mechanical stimuli to an organoid—such as stretch, compression, or fluid flow—to influence its development or function. A well-known example is the biohybrid fish, in which human cardiomyocytes powered rhythmic swimming by coupling with a compliant fin structure.[12] This system demonstrated a 2-way mechanical feedback loop: the contractions of cardiomyocytes drove the fins, while the elastic tension of the fins paced the cells in return. In a similar vein, researchers have built biohybrid robots where engineered skeletal muscle strips are anchored to flexible skeletons; when the muscle cells contract (induced by electrical pulses or light in optogenetic cells), the robot bends or crawls forward. The mechanical integration of organoids may involve embedding them in soft matrices or coupling them via connectors to the robot’s movable parts. Critical to success is maintaining the organoid’s viability and adherence during actuation. The success of such systems depends heavily on preserving organoid viability and stable attachment during actuation. To this end, innovations in biomaterials—such as collagen or fibrin gels—have enabled effective coupling between soft tissues and synthetic frameworks by forming a responsive mechanotransduction interface. Moreover, mechanical interfacing must account for differences in stiffness and strain tolerance: organoids are fragile and typically require gentle forces in physiologic ranges. The use of compliant materials and strain-limiting mechanisms in the robot can ensure that the organoid experiences lifelike mechanical cues without damage. Mechanical loading can also serve as a developmental cue-constant fluid pulsation or cyclic stretch can promote maturation of organoids (for instance, applying anisotropic stretch was shown to induce formation of multichambered, vascularized cardiac organoids mimicking an embryonic heart structure). However, despite the promising prospects, organ-based actuators also face fundamental challenges. Compared with synthetic motor systems, they usually generate lower force output, have reduced durability, and offer lower control accuracy. Furthermore, biological tissues require continuous support, such as perfusion or nutrient delivery, to maintain vitality, and their functional lifespan is essentially limited by biological aging. From this perspective, the mechanical interface is not merely a driving pipeline. They also play an active role in supporting organoid health and guiding maturation through physiologically related forces. 3.2 Electrical and neural interfaces Many organoids (especially neural, cardiac, and muscle types) communicate via electrical signals. Integrating these organoids with robots often entails reading and delivering electrical impulses. High-density microelectrode arrays or flexible electronics can be incorporated to record an organoid’s electrophysiological activity (such as the spiking of neurons in a brain organoid or the field potential of a cardiac organoid) and to stimulate the organoid electrically to evoke activity. In an organoid–robot control loop, a neural organoid could serve as a biological processor: the organoid receives sensory input from the robot (translated into electrical pulses delivered via electrodes), processes this input in its neuronal circuits, and then outputs signals that are fed back to the robot’s actuators. Early demonstrations of this concept are already appearing. In the DishBrain system, a 2D neural culture (a precursor to organoids) was embodied in a simulated game environment, and it was able to learn to play Pong by responding to electrical feedback—essentially functioning as a primitive biological computer.[13] Extending this to 3D brain organoids, one can envision living “brains” controlling robots. Indeed, researchers have reported that brain organoid cultures interfaced with microchips can drive mobile robots, adapting their activity based on feedback from the robot’s sensors. A recent open-source platform called MetaBOC has been developed to connect brain organoids to external devices, showing that human neurons grown on chips can receive inputs from and send outputs to robotic systems. These findings highlight that organoid networks are capable of embodied learning and closed-loop control, given appropriate interface hardware and training feedback. Electrical coupling is equally powerful for actuation: for instance, a soft robotic gripper lined with muscle organoids could have embedded electrodes that trigger the organoids to contract on demand, causing the gripper to bend and hold an object. Alternatively, light-based stimulation can be used if the organoid cells are genetically modified with optogenetic channels—light can penetrate soft tissues and avoid direct electrode contact, offering a wireless mode of control. In sum, electrical interfaces enable organoids to act as controllers and actuators in biohybrid robots, effectively merging biological neural networks with robotic systems in real time. 3.3 Chemical and microfluidic integration Another interface modality involves the exchange of chemical signals (eg, metabolites, hormones, and neurotransmitters) between organoid and robot. Organoids, by nature, secrete a variety of biochemical factors and respond to their culture milieu. A robot equipped with microfluidic channels can perfuse an organoid with specific solutes (drugs, nutrients, and signaling molecules) under precisely regulated timing and conversely sample the organoid’s secreted factors for analysis or feedback control. This type of interface is especially pertinent for metabolic organoids (liver, gut, pancreatic islets) and for drug testing applications. For example, a gastrointestinal organoid integrated in a robotic device could be exposed to varying concentrations of a drug via automated micropumps, while sensors measure the organoid’s secreted enzymes or electrical responses, creating a closed-loop drug screening system. There are already organs-on-chips that incorporate human organoids or tissue constructs with flowing media to emulate blood flow and nutrient delivery, some even linking multiple organoid types in sequence to represent a physiological system. Robotic automation has been used to assemble and cultivate organoids on-chip, ensuring reproducible placement and connection of tiny tissue spheroids. In advanced setups, one could imagine a “robotic symbiote” where the robot provides life support to the organoid—continuous perfusion, waste removal, and even immune isolation—similar to a heart–lung machine but on the microscale. The organoid, in return, provides real-time biochemical sensing or production. A concrete illustration is a “robot-on-a-chip” human motor unit developed in 2024, which integrated a brain organoid with a network of motor neuron spheroids and a muscle tissue strip, all on a microfluidic chip.[14] In this device, microfluidic channels supplied nutrients and drugs (like levodopa [L-DOPA] for a Parkinson disease to the organoids, while the neurons signals from the brain organoid to the effectively coupling chemical with mechanical actuation in one This how and chemical interfaces can connect organoid and robotic into a functional signal integration effective organoid–robot interfaces be and chemical the of a cerebral organoid to a soft for one the organoid with electrode arrays it in a scaffold that forces and perfuse it with a fluid The signals flowing across these interfaces can be integrated via electronics and signal and intelligence can complex organoid outputs (eg, neural spiking or metabolic and them into for the robot’s actuators. Conversely, can the stimulation to responses from the organoid it over time, to adaptive seamless integration also requires addressing the differences between biological responses and robotic control instance, neurons in over and tissue over A interface this by on multiple feedback for control and chemical for long-term A is that the organoid–robot interface not be biological components can serve as For example, in the human motor endothelial cells endothelial cells were to promote between the brain organoid and essentially forming a vascularized neural The endothelial network not the neurons but also the integration of neural circuits, signal from the organoid to the Such use of cellular components cells, cells, as living interfaces the and integration of organoid and robot parts. This of is to biohybrid interface is not a single of contact but an engineered where living and In summary, organoid–robot interfaces require a of bioengineering and flexible and control all into it to organoids as or of a robotic system, enabling the robot to and act The following concrete application where such organoid-integrated robots could be game the in in The of organoids with soft robotics a of across In this we present disease modeling, dynamic tissue regeneration, and These highlight how organoid-integrated robots could function in each the of living organoids to challenges. and drug screening Organoid–robot are disease by physiologically relevant in systems. A example is the brain organoids, motor and muscle tissue on a soft robotic to the in Parkinson The system in real time, by muscle contraction following with high soft robotic platforms enable of or gastrointestinal diseases by mimicking breathing and how physical forces and The of mechanical actuation to robotic screening systems and drug and drug delivery Integrating organoids as within robotic systems a new of platforms that and in real time. One application is a biohybrid where a pancreatic organoid embedded in a soft in to In such a system, robotics through controllable actuation or programmable microfluidic channels can the and of This delivery by with physiological demand, and Similarly, tumor robotic catheters could responses appropriate drugs tumor organoids and in real based on feedback. A soft robotic embedded may be used to test of in real time, offering a of Advances in synthetic biology allow engineered organoids to secrete on demand, while soft robotic components and Such instance, in which neural organoids electrophysiological activity to triggering soft robotic components to deliver targeted the of living devices for closed-loop tissue regeneration and soft robots offer an active in to devices that the during tissue In muscle regeneration, robotic embedded with muscle organoids apply controlled stretch and and maturation into In brain organoids with interfaces have been shown to integrate and with electrical For cardiac soft robotic with the and deliver electrical functional integration and enabling These systems on developmental where and force organoids to form and constructs offer unprecedented and in by combining the biological of organoids with the of robotic In liver organoid robots to real-time metabolic sensors provide early by or In vivo, soft robots immune or neural organoids could early or triggering automated For robotic platforms equipped with arrays (eg, and serve as living for chemical or offering a sophisticated to sensors. These systems not but physiological enabling closed-loop and from to From these it is that organoid-integrated offer They the between biochemical sensing and physiological than a they measure the to that with robotics, this For instance, of a in a cardiac organoid sensor trigger a to or sensing a in could an This the for and In all application regeneration, and act as limited to living metabolism, regeneration, and with soft robotics, they to a new of biohybrid that could by more lifelike in function and more in and organoid-integrated soft robotics a of complex technical and biological challenges that mechanical integration, systems, biological and long-term sustained viability of organoids within robotic platforms strategies that integrate endothelial networks and microfluidic to in vivo perfusion and mechanical between fragile organoids and soft robotic requires compliant hydrogels, and such as to structural and ensure functional For real-time control, the development of minimally invasive sensing from optical and chemical sensors to flexible for organoid and enabling adaptive robotic These systems must navigate biological signal and which across organoid types (eg, neural control in organoid and function automated and in organoids in function due to metabolic or strategies such as or control must be to ensure and as systems or sensory and and device become For instance, the integration of neural organoids into responsive robotic systems the of or current models remain from future adaptive or learning require Similarly, the use of patient-derived tissues in biohybrid devices for frameworks ensuring and long-term These while are active in research and the and biohybrid robotics. The integration of organoids with biohybrid robotics is to over the next to from in vitro demonstrations to relevant as closed-loop systems where neural organoids control muscle actuation under metabolic input from liver to the functional integration of multiple organoid types within robotic Concurrently, and including platforms and support of organoids with actuation or sensing The in vivo applications, such as implantable pancreatic organoid robots or vascularized neural are to emerge through advances in immune and robotic the control in and while organoid by neural tissues capable of learning offer a biological These systems new models of embodied developmental and in engineered biological and frameworks in for neural organoids tissue robotics, synthetic and the future of this supported by emerging and In the integration of organoids with soft robotics a and that could new for and combining the capabilities of living tissues with the and of we to creating devices that are not but Such devices one as an of the human body, capable of sensing, and in that or function. However, with this vision biological and Organoids require metabolism, have a limited and are by functional long-term requires a powerful support system, such as vascularized real-time and to maintain long-term viability and synthetic living tissues can adapt to or respond to of these must be into account in design and control. the next decade, we organoid-integrated biohybrid robots to from demonstrations to early applications, the stage for a new of living The require challenges and but the potential disease implants that respond to the body, and a on biomedical engineering and human The of organoids and robotics is more than an integration of 2 it is the of a new of that the of living and as as they are This was supported by the Science of and of Science and and the by for Science and of The that they have of and the and and the and and the