Bio-syncretic robots composed of biological and electromechanical systems

Chuang Zhang, Jia Yang, Wenxue Wang, Lianqing LiuNational 2022 · DOI

TL;DR

This paper proposes a roadmap for developing "bio-syncretic robots" by integrating biological systems with electromechanical components. The authors argue that combining the strengths of both domains is crucial for creating robots with advanced sensing, intelligence, and actuation capabilities that surpass current limitations.

What Problem Does This Paper Address?

Traditional electromechanical robots, while excelling in precision and strength, often fall short in areas where biological systems naturally shine: sensing, intelligence, and actuation. Biological systems, honed by eons of evolution, possess remarkable capabilities like high-temperature sensitivity in snake pit organs, efficient learning in brains, and powerful leaps in insects. These biological feats are difficult to replicate with current artificial systems. Conversely, biological systems can be fragile and lack the repeatability and power of electromechanical counterparts. This paper addresses the challenge of bridging this gap by proposing "bio-syncretic robots" that leverage the complementary advantages of both biological and electromechanical systems at molecular, cellular, and tissue levels.

Methods

The paper outlines a conceptual framework for bio-syncretic robots, focusing on three key functional areas: sensing, intelligence, and actuation.

  • Sensing: The authors suggest moving beyond single-spectrum, single-pixel sensing to multi-spectrum and array-based sensing. This involves engineering biological materials responsive to a wider range of light (visible and infrared) and utilizing microelectrode arrays (MEAs) to capture signals from ordered photosensitive cells. Nanomaterials like graphene are proposed to enhance cellular signals, and image reconstruction algorithms are needed for efficient multispectral imaging.
  • Intelligence: Current bio-syncretic intelligence relies on 2D neural networks, limiting complexity. The paper advocates for 3D neural networks derived from stem cells or brain tissue, cultured in vitro. These networks would be integrated with advanced MEA chips for complex information processing and accurate training. The authors also propose exploring multi-source stimulation (electrical, optical, chemical) for selective training and using motion feedback for intelligent learning.
  • Actuation: The focus shifts from miniature cell clusters to macroscopic muscle actuation tissues fabricated using bioprinting technology. The concept of "bionic integration" inspired by bone and muscle is suggested for coupling nonliving structures with biological actuators. The goal is to create robots with multiple cooperative biological actuators capable of complex, flexible behaviors, controlled via artificial or biological interfaces.

The paper also emphasizes the importance of a closed-loop system where sensing, intelligence, and actuation are integrated and interact with each other and the external environment.

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • Synergy is Key: The core idea is that combining biological and electromechanical systems offers a path to overcome the limitations of each individual domain, leading to robots with unprecedented capabilities.
  • Beyond Monomial Functions: Current bio-syncretic robots are limited to simple, single functions. The future lies in integrating sensing, intelligence, and actuation into a cohesive, cooperative system.
  • 3D is the Future for Intelligence: Moving from 2D to 3D neural networks is crucial for increasing the complexity and processing power of bio-syncretic intelligence.
  • Bioprinting for Scale: Fabricating macroscopic biological actuators through bioprinting is essential for developing bio-syncretic robots that can perform complex, large-scale tasks.
  • Closed-Loop System for True Intelligence: A closed-loop system that integrates sensing, intelligence, and actuation is necessary for robots to achieve intelligent, cooperative behavior and effectively interact with humans and their environment.

Key Results

The paper is primarily a conceptual proposal and roadmap, not a report on experimental results. It "preliminarily verifies that it is feasible to use biological materials as robotic function cores" but notes that "the majority have only demonstrated simple functions." The key "results" are the proposed pathways and strategies for advancing the field.

Limitations and Future Work

  • Limited Functionality: Current bio-syncretic robots are limited to simple functions and struggle to interact effectively with humans and the environment.
  • Complexity of Integration: The paper acknowledges that significant work is needed to fully realize the potential of biological systems in sensing, intelligence, and actuation for practical applications.
  • Future Directions: The authors highlight the need for:
  • Engineering biological materials responsive to multiple spectra for advanced sensing.
  • Developing 3D neural networks and matching 3D MEA chips for enhanced intelligence.
  • Utilizing bioprinting to create macroscopic muscle tissues for complex actuation.
  • Developing sophisticated control methods for bio-syncretic robots.
  • Advancing interdisciplinary research in areas like neuromuscular control, autonomous robotics, and in vitro biological maintenance technologies.
  • Achieving a "coexisting-cooperative-cognitive working mode with humans and the environment."

Summary updated

Abstract

Potential approaches to developing bio-syncretic robots are proposed in different dimensions, including monomial functions and closed loop systems of living cell-based sensing, intelligence, and actuation.
BibTeX
@inproceedings{zhang2022bio,
  author = {Chuang Zhang and Jia Yang and Wenxue Wang and Lianqing Liu},
  title = {{Bio-syncretic robots composed of biological and electromechanical systems}},
  booktitle = {National},
  year = {2022},
  doi = {10.1093/nsr/nwac274},
}

Loading related content...

Figure