Mimicking the thigmotropic behaviour of climbing plants to
design a tactile-based grasping device for the space
Study Reference Number: 12-6402
Type of activity: Standard study (25 k€)
Background and motivation
Over the last several decades the research on robotic manipulators has focused mainly
on designs that resemble the human arm. The designs are based on discrete rigid links
serially connected by joints. If we examine the manipulators available in nature, we
will see a plethora of other possibilities. Animals such as snakes, elephants, and
octopuses can produce motions from their appendages or bodies that allow the
effective manipulation of objects, even though they are quite different in structure
compared to the human arm. Moving from the animal to the vegetal world, we can
also find some example of grasping structures [Jaffe and Galston 1968].
Climbing plants are capable to grasp objects by extending themselves and then use the
objects as support by coiling around them. The goal is to achieve maximum vertical
height for rich sun exposure while avoiding the energy expenditure of developing a
supporting trunk [Isnard and Silk, 2009]. This big group of plants have continued to
fascinate biologists from Darwin’s time into the 21st century. Among all the known
climbing strategies, we can define a small group of plants possessing long, filiform,
and irritable organs called tendrils. They are specialized to grasp and climb the
surrounding environment. Unlike the examples in the animal kingdom, climbing
plants do not rely on vision to attain the support, but drive their tendrils using only the
Fig. 1 – Drawing of a caught tendril of Bryonia dioica, spirally contracted in reversed directions
There are many physiological studies of tendrils [Jaffe and Galston 1968; Putz and
Holbrook 1991] which well identify and describe their three main movements:
1. Circumnutation - an endogenous movement which increases the probability of
contact with supports.
2. Contact coiling - the stimulated tendril coils around a support.
3. Free coiling - the tendril develops helical coils along its axis to drag the stem
closer to the support providing a spring-like elastic connection.
– Schematic picture of the three main phases of a tendril
These relatively small structures can be used to support a vine weighing many times
their mass. Usually climbing plants develop many of these attaching points so that the
force is distributed between several adjacent attached tendrils, thus protecting the
plant from being torn away from their support during a stormy weather.
2 Study Objective
The ability of tendrils to allow plants to climb is anatomical surprising. Some
ecological studies prove the evolutionary success of climbers. Their strategy
constitutes a key innovation allowing them to succeed with the ability to ascend over
would-be competitors. This immediate advantage over their neighbours is at the cost
of relatively small energetic investment.
Tendrils are produced rapidly within the apical meristem of the plant and elongated
maximally before coiling. As the stem circumnutates, the tendrils touch potential
surfaces for grasping. When the tendril reaches sufficient maturity, stimulation of the
touch response sets in place a series of morphological changes so that the tendril
begins to coil around the object showing a positive thigmotropic behaviour
(thigmotropism is a directional movement in which an organism moves or grows in
response to a directional touch stimulus).
The distal portions of tendrils are highly touch-sensitive, in some cases greater than
the human’s counterpart. Some studies report that a 0.25 mg thread drawn along a
tendril can be enough to evoke a response [Simons 1992], and Darwin documented
tendril responses to stimuli in the range of 1–5 mg [Darwin 1865].
Touch stimulation leads to a rapid onset of tip coiling, and this enables a secure
attachment with the object. However, if the tendril loses the touch stimulation, it can
reverse the movement by uncoiling [Jaffe & Galston 1968] and be ready to coil again
upon another stimulation.
After the mechanical stimulus has been perceived by epidermal cells of the tendril,
plant hormones serve as mediators of the coiling response. They are released by
epidermal cells and are diffused along the structure. This leads to a growth arrest on
one side of the tendril and a promotion of growth on the opposite side, together with a
swallow and shrinkage of cells [Jaffe & Galston 1968].
Using this basic sensory-motoric loop without centralized sensing and control, plants
can blindly rely on organs that will coil around supports, providing a successful
grasping method. In addition, they are flexible to adapt to the shape of the object that
they are grasping.
How can tendrils be imitated and what are the advantages in imitating them? Robotic
grasping is a complex task and poses some very intricate problems. One of the main
difficulties is to be able to generalise to variations of object position, orientation and
shape. For this reason, most of the robotic grasping known assumes a grasped object
and all information needed for the grasping to occur. Alternatively, they rely on vision
to obtain relevant information about the objects. [Takahashi et al. 2008; Cannati and
Plants lack of a nervous system, and therefore, the exploration of the mechanical
environment and the execution of mechanical actions rely efficiently on simple reflexlike
behaviour. This reflex-like behaviour is key feature for the successful grasping of
objects of unknown shape and position driven only by the tactile information. An
interdisciplinary approach to the functional morphology, kinematics, and physiology
of tendril design will be beneficial both to biologists and engineers. The sensory and
actuation system may be less dynamic than our human senses and muscles but still
have the advantage of greater autonomy. The biomimetic of the tendril may be
extremely valuable for use in autonomous robots. Some of the potential space
applications can be: controlling the length of momentum exchange tether or
electrodynamics tethers, grasping debris to facilitate their removal, terminal docking,
space refuelling and self-assembly and, in general, all those missions where grasping
or controlling a thether length [Kruijff 2011] could offer a solution to a space
The objective of this study is to develop a tendril-like flexible mechanism that best
mimics the behaviour and the flexibility of a tendril by performing mechanical
grasping and pulling while maximizing the degrees of freedom.
3 Proposed Methodology
Research groups are invited to submit one or more proposed solutions together with
an evaluation of their features and a basic demonstration of their feasibility (proof of
The research project is proposed to develop in the following steps:
-Understand the biomimetic behaviour of a tendril from an engineering point of
view in order to reproduce the natural characteristics of its hypothetic
-Assess the scope and framework of the model
-Model the behaviour of the tendril mechanism with a kinematic and dynamic
simulation in order to characterize the key model parameters, which
influence the grasping behaviour.
-Construct a 3D simulated model of the mechanism incorporating the interacting
forces and reactions.
-Evaluate the performance of the grasping and pulling of objects (symmetric
along the longitudinal axis and long enough to allow for multiple coiling) in microgravity. The objects will be of defined size, mass, material, and various
cross sectional shapes.
-Evaluate the robustness of various aspects of the mechanism (e.g. single point
A clear understanding of the tendril from a biological and engineering point of view is
essential to best reproduce the natural grasping behaviour at the base of the grasping
feature. This will be done in close collaboration with the ACT researchers who can
provide all the biological information.
The main focus of the applicant’s proposal should be a discussion on the intended
development of a tendril-like mechanism and on the creation of a simulation model.
The following points need to be understood:
4 ACT Contributions
- The two masses (the mass the tendril is attached to and the object to grasp) are
placed in the space environment therefore they are free to move.
- The grasping phase needs to be modeled and a firm attachment created with at
least two coils formed around the grasped object. A pulling phase should also
be modeled as to assess the grasp firmness.
- We suggest the mechanism to have a modular structure made of single
elements. Each element able to “sense” the touch and “actuate” a relative
expansion/contraction which results in the overall movement of the tendril in
the direction of the touch. Each element is also able to communicate with the
- In the biological tendril each segment acts as an autonomous agent that
decides its direction of growth according to the direction of touch stimuli. The
“actuate” decisions of each single segment is transmitted to the adjacent
segments and further along the tendril. Once the tendril has coiled around the
object with a certain number of coils the final stage of the grasping occurs by
the contraction of longitudinal fibers (cellulose G-fibers in nature). These
fibers run through all the tendril elements. By contracting them, it effectively
reduces the overall radius of the helix, and thus, secures the hold.
- To perform the grasping, plants work at a timescale of several hours, and
although we assume that the device should be faster than its biological
equivalent, we leave this feature open and linked to the material and actuators
chosen for the device itself.
This study is mainly addressed to research laboratories in the fields of biomechanics,
The project will be conducted in close scientific collaboration with ACT researchers.
ACT researchers will provide both knowledge concerning space related issues and
applications and plant behavioural biology as detailed in the project description
Cannata, G., Maggiali, M. (2005) An embedded tactile and force sensor for robotic
manipulation and grasping. Humanoid Robots, 5th IEEE-RAS International
5.1 Additional readings
Darwin C. (1865). On The Movements and Habits of Climbing Plants. London: John
Engelberth J., Wanner G., Groth B. and Weiler EW. (1995). Functional anatomy of
the mechanoreceptor cells in tendrils of Bryonica dioica Jacq. Planta 196 : 539 – 550
Isnard S, Silk WK. (2009) Moving with climbing plants from Charles Darwin's time
into the 21st century. American Journal of Botany 96:1205-1221.
Jaffe MJ, Galston AW. (1968). The physiology of tendrils. Annual Review of Plant
Physiology 19: 417–434.
Kruijff, M. (2011) Tethers in space: A propellantless propulsion in-orbit
demonstration Uitgeverij BOXPress.
Liss , H. , Weiler EW. (1994). Ion-translocating ATPases in tendrils of Bryonica
dioica Jacq. Planta 194 : 169 – 180 .
Putz F.E., Holbrook N.M. (1991). Biomechanical studies of vines. In Putz F. E.,
Mooney H. A. [eds.], The biology of vines, 73–97. Cambridge University Press,
Simons P. (1992). The Action Plant. Oxford, UK: Blackwell Publishers.
Takahashi T, Tsuboi T., Kishida T., Kawanami Y., Shimizu S, Iribe M., Fukushima
T., Fujita M. (2008) Adaptive Grasping by Multi Fingered Hand with Tactile Sensor
Based on Robust Force and Position Control. IEEE International Conference on
Robotics and Automation Pasadena, CA, USA, May 19-23, 2008
Bowling A.J., Vaughn K.C. 2009. Gelatinous fibers are widespread in coiling tendrils
and twining vines American Journal of Botany 96(4): 719–727.
Carrington C. M. S., Esnard J. (1989) Kinetics of thigmocurvature in two tendrilbearing
climbers Plant, Cell and Environment 12, 449-454
Engelberth J. (2003) Mechanosensing and signaltransduction in tendrils. Adv. Space
Res. Vol. 32, No. 8, pp. 1611-1619
Jaffe M. J., Galston A. W. (1966) Physiological Studies on Pea Tendrils.I. Growth
and Coiling Following Mechanical Stimulation Plant Physiol. 41, 1014-1025