For decades, researchers have marveled at the remarkable capabilities of biological structures like elephant trunks, octopus arms, and climbing plants. These natural continuum structures exhibit an enviable versatility and adaptability despite possessing no central control unit or “brain.” Scientists have long sought to replicate such traits in robotic systems for applications like search and rescue, surgery, infrastructure inspection, and space exploration. However, emulating the complexity of biological structures with traditional rigid robots has remained an enduring challenge.
Recent breakthroughs in soft robotics and smart materials have brought the goal of building highly capable continuum robots closer to reality. Yet most research has focused on advancing actuation and control systems rather than matching the structural optimizations that enable such extraordinary functionality in organisms shaped by eons of evolution. As a result, even state-of-the-art soft continuum arms require extensive arrays of sensors, actuators, and control hardware to accomplish the fluid movements and environmental adaptations that plants and invertebrates perform intrinsically.
Now in a paper published in Advanced Intelligent Systems ("A Soft Continuum Robotic Arm with a Climbing Plant-Inspired Adaptive Behavior for Minimal Sensing, Actuation, and Control Effort"), scientists at the Italian Institute of Technology’s Bioinspired Soft Robotics Lab demonstrate a pioneering approach to replicating biological structures all the way down to the tissue level. Their plant-inspired soft continuum arm, named Mandy, matches the gradient distribution of flexural rigidity observed along the searching stems of a climbing plant species. This breakthrough biomimetic design enables Mandy to detect supports and twine itself around them using only a single sensor and actuator.
The researchers performed biomechanical tests on Mandevilla plant samples to analyze how their flexural rigidity decreased from the stem base to the growing tip. This gradient enables sufficient apex flexibility for the searching circumnutation movements needed to locate supports while maintaining the basal stiffness necessary to bear the structure’s weight.
Upon finding a suitable support, the Mandevilla’s searcher stem alters its growth pattern and establishes contact through thigmomorphogenesis (the process by which plants change their growth and development in response to mechanical stimulation, such as touch or wind). Like many climbing plants, the Mandevilla cf. splendens attaches itself by twining its stem around supports. The authors set out to replicate in a robot both the gradient distribution of flexural rigidity that enables searching behaviors and the automated twining response triggered by physical contact.
Rather than concentrating sensors, actuators, and control systems in a small section of the arm, the researchers distributed intelligent physical attributes throughout the arm’s structure. The result is an elegantly simple robotic system with emergent capabilities far beyond its hardware specifications.
The key innovation enabling Mandy’s advanced functionality with such compact onboard resources was matching the flexural stiffness distribution along the Mandevilla stem in each section of the arm. This was accomplished by 3D printing the arm from rigid base segments transitioning gradually to extremely flexible rings toward the tip.
The researchers also mimicked the plant’s thigmomorphogenesis by integrating small air chambers as distributed contact sensors. Wall thicknesses descending from base to tip ensure equal air pressure changes. This enables detecting suitable supports anywhere along the arm using a single air pressure sensor rather than an array. When triggered, a lone tendon actuator initiates Mandy’s twining response.
Simulations and experiments verified Mandy’s ability to replicate the Mandevilla’s autonomous behaviors using fraction of the hardware required by other continuum arms. Simply by incorporating biological design principles, the researchers achieved cutting-edge functionality on a drastically reduced budget.
This breakthrough study opens exciting new directions for soft robotics research. Matching biological structures as closely as possible unlocks capabilities unimaginable in traditional rigid systems. And minimizing mechanical complexity through bioinspired design enables accomplishing more with less.
Mandy’s adaptable body and distributed sensorimotor control closely replicate the Mandevilla’s elegant autonomous behaviors. Yet this early prototype accomplishes such feats with only simple 3D printed segments, a single sensor, and a lone actuator.
By progressing down the development path illuminated by nature herself, soft continuum arms could soon deliver on longstanding promises to transform applications from minimally invasive surgery to disaster relief. And advanced functionality may finally become accessible even with extremely compact, low-cost hardware specifications.
As the researchers emphasize, investing more effort in quantitative biomechanical analysis and biomimetic structural design looks poised to pay immense dividends moving forward. Work like Mandy highlights the enormous gains still waiting to be unlocked simply by studying and implementing time-tested biological solutions.