NTT Research & Harvard Scientists Revolutionize Biohybrid Ray Design with Machine Learning

In a groundbreaking collaboration, NTT Research and Harvard scientists have optimized the development of biohybrid rays using machine-learning directed optimization (ML-DO), leading to significant improvements in swimming efficiency and adaptability. This innovative approach, detailed in a new paper published in Science Robotics, offers promising prospects for the future of biohybrid robotics.

The conventional biomimetic approach to biohybrid design involves recreating existing biological structures, which has limitations in terms of adaptability and efficiency. For biohybrid lifeforms resembling batoid fishes, there is a wide range of natural aspect ratios and fin morphologies, making it challenging to select the most suitable fin geometries for a given working environment. Additionally, biomimetics may neglect the natural biomechanical and hydrodynamic forces that govern swimming speed and efficiency, leading to inefficient muscle mass and limited swimming speeds (Zimmerman et al., 2025).

To overcome these limitations, the research team led by Harvard SEAS Postdoctoral Fellow Dr. John Zimmerman, NTT Research Medical and Health Informatics Scientist Ryoma Ishii, Harvard SEAS Tarr Family Professor of Bioengineering and Applied Physics Kevin Kit Parker, and members of the Harvard SEAS Disease Biophysics Group, employed ML-DO to efficiently search for high-performance design configurations in the context of biohybrid robots.

The team developed an algorithm for expressing a multitude of different fin geometries and described a generalized ML-DO approach for searching within a large discontinuous configuration space. By using this methodology, they identified biohybrid fin geometries for high-performance swimming with smooth and orderly flow. The ML-DO-driven results included a quantitative exploration of fin structure-function relationships and reconstruction of general trends in open-sea batoid morphology, as well as a winning design: Fins with large aspect ratios and fine tapered tips, which preserved their utility across multiple length-scales of swimming.

On the basis of these findings, the team built biohybrid mini-rays out of engineered cardiac muscle tissue, which were capable of self-propelled swimming at the millimeter length scale. These biohybrid rays demonstrated improved swimming efficiencies approximately two times greater than observed in previous biomimetic designs, marking a significant advancement in the field of biohybrid robotics.

Looking ahead, researchers note that additional work is needed to completely match natural scaling laws. While the devices presented in this study demonstrated greater efficiency than other recent biomimetic designs, they were still slightly less efficient on average than naturally occurring marine lifeforms. However, the potential applications of this ML-DO-informed approach are vast, including remote sensors, probes for dangerous working environments, and therapeutic delivery vehicles. Furthermore, this research advances scientific understanding of 3D organ biofabrication, such as a biohybrid heart, by demonstrating the potential of ML-DO in optimizing biohybrid designs.

In conclusion, the collaboration between NTT Research and Harvard scientists has resulted in a significant breakthrough in biohybrid ray design, leveraging machine-learning directed optimization to enhance swimming efficiency and adaptability. This innovative approach holds great promise for the future of biohybrid robotics and related fields, paving the way for more advanced and efficient biohybrid robots.

Word count: 598