Fig. 1: Artificial fairy controlled by light. Credit: @ Jiangfeng Yang, Tampere University.

In a fascinating article discussing the first of a three-part journey to create autonomous flying robots, Dr Hao Zeng and PhD student Jianfeng Yang of the University of Tampere, Finland, outline their creation; a fairy robot inspired by the seed of the dandelion.



Profile: Hao Zeng

Dr Hao Zeng is currently a Finland Academy Research Fellow and the leader of the Light Robots Group at Tampere University.

He obtained a PhD degree in photonics from Florence University in 2015. Then he carried out postdoctoral research in Finland between 2016 and 2021.

His research includes light-responsive polymers, micro-robots and non-equilibrium soft matters. In 2022 he received an ERC Starting Grant with a project on light communicative materials.


Profile: Jianfeng Yang

Jianfeng Yang is currently a PhD researcher and a member of the Light Robots Group at Tampere University.

He obtained a Master’s degree in Chemical Engineering and Technology from Southeast University, China, in 2021.

His research includes Liquid crystal elastomers, photo actuators, flying micro-robotics and self-oscillators that resemble life-like behaviours.

How close are we to completely autonomous flying robots? Well, drones, in a very short space of time, have become an everyday occurrence. We see them in every situation from dramatic overhead footage in television shows, to their deadly use in armed conflicts. There is even talk of drones delivering packages from the air to your front garden, replacing the need for a delivery driver.

But all of these things need human input. The question is, we will ever get to autonomous flying robots and what exactly will they be able to do? Dr Hao Zeng and PhD student Jianfeng Yang from the University of Tampere, Finland, have set out to investigate just that. Here they explain how they came to create what they describe as a ‘fairy robot’…

Stimuli-responsive polymers are man-made materials that can undergo substantial shape changes in response to external stimuli, such as temperature, light, humidity, etc. Integration of these polymers in soft micro-robotic structures allow for the realisation of versatile types of robotic locomotion, such as walking, jumping, rolling, and swimming.

In this article, we report on a soft actuator assembly capable of wind-assisted dispersal and lift-off/ landing action under the control of a light beam. Zeng’s group took inspiration from the seed of dandelion, and fabricated artificial dandelion seeds that were lightweight with high porosity, and can change their shape by using light. See Fig. 1 above.


What are the challenges for soft robots?

Soft robots are machines constructed with soft actuators aiming at replacing the rigid joints and links appearing in conventional hard-bodied robotic systems. Soft robotics is expected to overcome the great challenges encountered by rigid machines that relate to the degree of freedom of actuation, safe human robot interaction and adaptive movement. Conventional soft robots use pneumatic/ hydraulic actuators controlled via input pressure, electroactive polymers deforming upon applied voltage, or hybrid actuators combining elastomer layers and rigid electronics, all relying on tube/ cable connections, thus yielding a bulky structure.

To miniaturise overall dimensions, which are required for many medical applications and robotics in small, confined spaces, responsive (smart) materials stand out as the novel design principle for soft robotics. Responsive materials are capable of large deformation upon external stimuli, allowing wireless control of robotic motion through external fields.

This nascent research field – responsive material-based robotics – has been established in the past decade, and it keeps drawing increasing amounts of attention from both material scientists and micro-roboticists as a route towards fast, accurate and reliable actuator systems for micro-robotic purposes.


What can state-of-the-art soft robots do?


Strongly motivated by the research interests in medical technologies and treatments, soft robots are developed into sophisticated swimmers, particularly micro-robots performing locomotion through microscopic channels and biofluidics-like environments. For cargo transportation and assembly, walking robots are investigated. Gripping devices are also made for micro-manipulation/ -assembling and drug delivery purposes. Two of the examples developed in the Light Robots Group can be found in Fig. 2.


Fig 2. Examples of light-driven micro-robots. Flytrap-like gripper and artificial octopus. @ Tampere University.

How can we go beyond the state of-the-art?


Thinking beyond walking, swimming and other ‘conventional’ manipulating functions, flying is a frontier yet to be conquered by responsive materials. To realise a flying robot with even smaller, sub-centimetre size, is greatly challenging because of the hurdles of miniaturisation in existing actuators and unsteady aerodynamics encountered at this small length scale. The realisation of small-scale flying robots will enable future autonomous aircraft with versatile manoeuvre function.

This is why we chose this topic. To address the above challenges, the Light Robots Group started the FAIRY project in September 2021, funded by the Academy of Finland. The goal of the project is to develop autonomous, small-scale flying robots capable of adaptive response met in biological systems.

The first question is how do we make them? The Light Robots Group took its inspiration from Nature, and the design research is to be done in three parts:

Firstly, we made our flying robots through passive means, whereby a structure can be carried by a gust of wind or relative airflow, resembling the dispersal of the seeds of the dandelion plant; secondly, we want to make the flying motion much faster by developing a system that can glide just as many birds do in Nature; and finally the ultimate goal of the project is to develop active flight, in which wings flap to create enough thrust to counteract gravity, analogous to birds’ ability to both take-off and hover.

These three natural processes are shown in Fig. 3., with the light-controlled dandelion in this article the result of the first part of the trilogy.


Image of Dispersal: Credit: © Tagooty (CC BY-SA 4.0) -sa/4.0/legalcode. 


Image of Gliding: Credit: © Ron Knight (CC-BY 2.0) 2.0/legalcode. Image of Hovering: Credit: © Becky Matsubara (CC-BY 2.0) 2.0/legalcode.


Learning from Nature – achieving ‘dispersal’


Dandelions are one of the most widely spread herbs in the world. This is because of its remarkable skills of dispersal assisted by winds. The secret behind the dispersal capability is the parachute-shaped pappus attached to each seed. The pappus consists of a bundle of bristle filaments with a diameter of about 15 microns. This unique structure leads to high porosity and the formation of a separated vortex ring above the dandelion seeds. This specific vortex ring generates low-pressure air bubbles, which suck the dandelion upwards and thus prolongate the floating time in air. Fig. 4.


Fig. 4: Separated vortex patterns created by natural dandelion seed (left) and a synthetic dandelion seed (right). Credit: @ Jiangfeng Yang, Tampere University.


What have we achieved?


We fabricated an artificial structure by glueing 54 bristle filaments on a 2D rectangular light-controlled soft actuator, and arranged them in a centro-symmetrical fashion. The actuator is made of a liquid crystal elastomer membrane – it can bend/ unbend upon light being switched on or off. The artificial dandelion has several biomimetic features, such as high porosity (0.95) and lightweight (1.2mg), and separated vortex ring generation. Superior to its natural counterparts, this artificial structure can reversibly open/ close upon visible light excitation. Hence, a fairy-like robot is created.


How to test them?


We set the artificial dandelion inside a wind tunnel with controllable flow speed, and used a home-used humidity generator to seed the air, so the airflow could be visualised by camera upon a laser illumination. We observed similar separated vortex ring structures above the structure, and this vortex pattern is able to be changed by the opening or folding of the structure when we used another LED light beam shining from the top. Fig.5.


Fig. 5: Photographs of the vortex ring pattern of an artificial structure @ Jiangfeng Yang, Tampere University.

Ready for take-off


Finally, the researchers placed the robot inside a wind tunnel with constant flow velocity. By shining on the light, the structure quickly unfolds and takes off inside the wind flow. Fig. 6 below.



The loss of pollinators, such as bees, is a huge challenge for global biodiversity and affects humanity by causing problems in food production. Could this artificial fairy provide some technological possibility in artificial pollination?


We believe many practical questions remain to be answered. But we are basic scientists, and not good at selling our story through science-fiction or providing a long-term vision for our work. Our biggest curiosity in this research today is people nowadays think they have synthesised smart materials as good as animal muscles, performing the same level of actuation force, frequency, and so on. But why it is so difficult to make them fly?

Looking at the small insects, butterflies and hovering birds, they all have well-defined trajectories for their wing movement. So, flying in air by wing flapping should be a very simple skill to learn. If we consider the simplest step in natural flying species – the flapping of wings to take off – can we reproduce them in inanimate material?

In the first instance, material sensitivity needs to be improved, in order to allow efficient operation under sunlight. Secondly, we need to explore other control strategies to steer the motion of trajectory in the air and significantly increase the precision of the landing spot. Thirdly, we also need to take the environmental issues into concern. To minimise the pollution and unexpected side effects, the devices should be reusable, and the material should be biodegradable after some time.

To achieve all of this will require huge progress in the research of micro-robots and smart materials, which will therefore require close collaboration with materials scientists and people working on micro-robotics. It will be the next step of our research.

Fig 6: Series of photographs showing the take-off action upon light irradiation @ Jiangfeng Yang, Tampere University.

Jianfeng Yang,

PhD student researcher


Dr Hao Zeng,

Group leader


Faculty of Engineering and Natural Sciences,

Tampere University.