With the rapid development of computer technology, traditional silicon-based robots have achieved substantial improvements in intelligence. However, because they still mainly rely on rigid mechanical structures, batteries, and electric drive systems, their mobility, autonomous decision-making, endurance, and natural interaction with complex environments remain inferior to those of living animals. To overcome these limitations, research has gradually shifted from bionic robots to biohybrid robots. Among them, cyborg animals represent an important branch that combines machine intelligence with biological intelligence, leveraging animals' inherent capabilities in perception, locomotion, energy supply, and environmental adaptability. This gives cyborg animal systems unique potential for tasks such as environmental exploration, emergency search, and mission execution in complex scenarios. "At the same time, the field still faces major challenges, including limited consistency of control across individuals, sensitivity of stimulation outcomes to internal states and environmental disturbances, and insufficient long-term stability, portability, and controllability of the systems." said the author Yue Ma, a researcher at Shenyang Institute of Automation, Chinese Academy of Science, "These issues make it necessary to systematically review the construction, control strategies, and applications of cyborg animals in order to support further advances in the field."
Rather than focusing on a single cyborg animal model or one specific technique, this review aims to systematically map the development of the entire field. It first adopts animal taxonomy as an organizing framework, summarizing studies on fish, reptiles, mammals, birds, and various invertebrates, thereby highlighting the differences among animal models in locomotion capability, controllability, and application potential. On this basis, the authors further examine cyborg animals from the perspective of system construction, covering key components such as brain–computer interfaces, electrical stimulation of muscles and receptors, as well as visual, chemical, thermal, and optical stimulation strategies, together with electronic backpack design, navigation-control algorithms, and representative application scenarios. Through this dual-framework approach, the review not only presents a broad landscape of cyborg animal research, but also provides a clear structure for comparing the match between different animal models and control strategies.
Cyborg animal research has evolved from an early exploratory stage dominated by insects and rats into a much broader landscape involving birds, fish, reptiles, and various invertebrates. At the same time, control strategies have expanded from simple electrical stimulation to a diverse toolbox including brain–computer interfaces, electrical stimulation of muscles and receptors, as well as visual, chemical, thermal, and optical stimulation. Together, these developments indicate that the field is moving beyond the question of whether animal motion can be controlled, toward the more advanced goal of achieving control that is more precise, stable, and application-specific. The review also highlights that progress in electronic backpack design, navigation algorithms, and closed-loop control is helping cyborg animals move from isolated laboratory demonstrations toward more sophisticated applications such as environmental exploration, swarm robotics, and human–machine interaction.In terms of technical comparison, the authors emphasize that no single control strategy is universally superior; instead, its suitability depends on the animal model, task demands, and system constraints. Broadly speaking, brain–computer interfaces and optogenetic approaches are better suited for animals with more complex neural regulation and richer movement patterns, enabling higher-level behavioral modulation. Muscle and receptor stimulation methods are more direct and often offer stronger spatiotemporal control, making them especially useful for small animals such as insects. By contrast, noninvasive approaches such as visual and electric-field-based control offer better biocompatibility but are generally more susceptible to environmental interference and lower control precision. Chemical stimulation can effectively alter behavioral states, but its delayed onset and potential side effects remain important limitations. The review therefore suggests that future progress in cyborg animals will depend not merely on adding more control methods, but on achieving a better balance among adaptability, biocompatibility, control accuracy, system complexity, and real-world deployability.
The next step for cyborg animals is not simply to add more animal models or more stimulation methods, but to push the field toward systems that are more stable, more intelligent, and more practically usable. The real challenge is that different animals vary greatly in neural organization, locomotion patterns, and behavioral traits, so control strategies need to be matched much more precisely to each animal model. At the same time, the electronic backpack, as the key interface between the organism and the electromechanical system, must better balance lightweight design, miniaturization, biocompatibility, and long-term stability, because these factors directly shape movement performance, endurance, and control quality. Beyond that, cyborg animals need to evolve from one-way stimulus-driven demonstrations into closed-loop systems that integrate sensing, positioning, navigation, feedback control, and even swarm coordination, so they can support environmental exploration, complex task execution, and higher-level human–animal–machine interaction. Ethical and animal-welfare considerations also need to be treated as a central design constraint rather than a secondary concern. "The most meaningful progress ahead will come not from making animals merely more controllable, but from integrating biological capability with engineering systems in a way that is lower-burden, more reliable, and more autonomous." said Yue Ma.
Authors of the paper include Yue Ma, Chuang Zhang, Fei Nie, Hengshen Qin, Qi Zhang, Yiwei Zhang, Lianchao Yang, and Lianqing Liu.
This work was supported by the National Natural Science Foundation of China (62373347, 62525301 and 62333021), the New Cornerstone Science Foundation through the XPLORER PRIZE, the CAS Project for Young Scientists in Basic Research (YSBR-041), the Youth Innovation Promotion Association of CAS (2023210), the Natural Science Foundation of Liaoning Province (2024JH3/10200028), the State Key Laboratory of Robotics and Intelligent Systems (2024-Z04), the Fundamental Research Project of SIA (2022JC2K01 and E4391103), and the Jiang Xinsong Young Seedlings Fund of SIA.
The paper, "Construction, Control, and Application of Cyborg Animal Composed of Biological and Electromechanical Systems" was published in the journal Cyborg and Bionic Systems on Mar 26, 2026, at https://doi.org/10.34133/cbsystems.0486.
