If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Developed soap bubbles exhibit various biological and physicochemical properties
•
The soap bubbles allow effective flower pollination
•
A flying robot equipped with a bubble maker can be used for autonomous pollination
Summary
Natural and artificial flower pollination are critical processes in the life cycle of flowering plants. Declines in the number of global pollinator insects, the heavy labor of conducting artificial pollination manually, and the rising cost of pollen grains are considered to be significant worldwide problems. Here we show that chemically functionalized soap bubbles exhibit effective and convenient delivery of pollen grains to the targeted flowers thanks to their stickiness, softness, high flexibility, and enhancement of pollen activity. By exploring the physicochemical properties of functional soap bubbles, we could prepare mechanically stabilized soap bubbles capable of withstanding the windmills produced by robotic pollination. An unmanned aerial vehicle equipped with a soap bubble maker was autonomously controlled to pollinate flowers. Such technology of automatic intelligent robotic pollination with functional soft materials would lead to innovative agricultural systems that can tackle the global issues of pollination.
). Hand pollination with a cotton swab or a small brush is an effective method that has been used since ancient times as it allowed operators to apply pollen grains directly to flowers (
); however, this method required heavy labor to manually apply the pollen grains to all flowers within a farm in a timely manner. Machine pollination methods such as pollen blowers, dusters, and spray dispensers have been alternatively used recently to reduce the human labor and the reliance on insect pollination (
); however, the expenses incurred from these conventional machine pollination methods have largely increased owing to the cost of pollen grains. In fact, these approaches produce a large number of inefficient pollen grains, especially those scattered from machines, which are not directly targeted toward the flowers. Alternatively, robotic crop pollination has attracted significant attention because of its potential advantageous performance in terms of individual flower detection, autonomous operation ability, and utilization of biomimetic strategies (
Elamvazhuthi, K., and Berman, S. (2015). Optimal control of stochastic coverage strategies for robotic swarms. in: 2015 IEEE Int. Conf. Robot. Autom., IEEE 1822–1829.
). As such, the notion of robotic pollination might address the problem of reduction of natural insect pollinators as an alternative way. Moreover, developments in the field of robotic pollination seem to be directed at paying attention to reducing the pollination workload of farmers engaged in the agricultural business. Very recently, there appeared a study on a materially engineered artificial pollinator equipped with sticky ionic liquid gel coated by vertically aligned animal horse hair, which can work as a biomimetic honeybee to transport pollen grains among flowers (
). However, our previous pollination using a small toy drone lacked autonomous controlling system and the sticky animal hairs on a drone were needed to physically scrabble onto the flowers for delivering pollen grains. Operability of the technology itself was unfortunately impractical, and flowers were seriously damaged. Bearing fruit was thus not achieved by our previous technique due to these critical problems. Nonetheless, there remain huge demands for the development of material-engineering-based intelligent robotic pollination for effectively and efficiently pollinating crops. In addition, developing and designing of intelligent functional robots is attractive as an emerging technology and promising for autonomous precision robotic pollination (
Strader, J., Nguyen, J., Tatsch, C., Du, Y., Lassak, K., Buzzo, B., Watson, R., Cerbone, H., Ohi, N., Yang, C., and Gu, Y. (2019). Flower interaction subsystem for a precision pollination robot. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2019, Macau, China.
For centuries, scientific studies on soap bubbles have fascinated people of all ages mainly because of their beautiful rainbow colors and thin-film-based geometric structures based on simple scientific principles (Behroozi et al., 2008;
). However, despite the immense use of soap bubbles for entertainment, scientific, and educational purposes, establishing methodologies and applications using soap bubbles as a functional material remains substantially unaddressed. We believe that the steady liquid membrane and the large surface area of soap bubbles are suitable media for delivering microscopic lightweight pollen grains for pollination and that the easy degradation and the low cost of the eco-friendly ingredients of soap bubbles are attractive and unique features. In addition, flowers do not sustain substantial physical damage from directly shooting soap bubbles because soap bubbles are lightweight, soft, and highly flexible.
With this aside, a large number of pollen grains are easily scattered in orchards through conventional artificial pollination with a brush and a machine pressure sprayer (
). Unfortunately, the scattered pollen grains non-specifically attach to superfluous flowers, which often disturb the bearing and harvest of fruits. In fact, the inefficient fruits produced from such non-specific attachments are eventually thinned out, accompanied by heavy human labor resources for maintaining fruit development and tree vigor (
). High controllability of the directional flying of soap bubbles by bubble-making devices is useful for simply shooting soap bubbles directly onto the target flowers so as to systematically reduce the workload of thinning out superfluous fruits. Moreover, scattering of pollen grains can be physically restrained as they are tightly confined to the liquid membrane of the soap bubbles. Thus, we are confident that soap bubbles are an ideal material for delivering pollen grains through robotic artificial pollination.
In this study, we demonstrate that (1) chemically functionalized soap bubbles exhibit unique properties, such as delivering pollen grains to the targeted flowers in a simple manner, reducing the usage of pollen grains, effectively attaching soap bubbles on the pistils of the targeted flowers using the high stickiness of the soap bubble membrane, preventing severe damage to delicate flowers using the softness and high flexibility of soap bubbles, and enhancing the pollen activity by promoting germination ratio and length of pollen tube; (2) chemically functionalized soap-bubble-mediated pollination can be used for practical Pyrus pyrifolia var. culta pollination at orchards aside from its contribution toward the healthy expression of fertility for various pollen grains; (3) mechanically stabilized soap bubbles capable of withstanding windmills due to robotic pollination can be successfully prepared; and (4) an autonomous controllable unmanned aerial vehicle (UAV) equipped with a mechanically stabilized soap bubble maker can fully automatically transport pollen grains to Lilium japonicum flowers, thus successfully aiding in plant pollination. The findings and concepts demonstrated in this study will undoubtedly influence the design of next-generation soft-materials-based robotic technologies (
), including applications in agriculture, material chemistry, biomimetic science, and aviation engineering.
Results
Preparation of Soap Bubble Solutions
Figures 1A and S1 show a schematic illustration of how soap bubbles containing pollen grains are prepared using a bubble gun. At the interface between a soap film and air, the surfactant transforms a molecular bilayer in which the heads are directed toward the aqueous phase and the tails are directed toward the air (
). We thus hypothesize that pollen microparticles are pushed out with the solution and then physically absorbed onto the bilayer of a bubble membrane by mechanical blowing from the fun in the bubble gun. We accidently found that natural pollen grains can be easily incorporated into a soap film and flown in the air using various bubble devices. One interesting device is the battery-and-motor-driven bubble gun, which can produce a lot of soap bubbles (Figures 1A and S2, and Video S1). We used pear pollen grains from Pyrus bretschneideri as an experimental model for the hereinafter described practical pollination at orchards and measured the germination ratio and growth of the pear pollen tubes using in vitro pollen activity assays (measurements of germination ratio and length of pollen tube) according to instructions from previous works (
(A–C) (A) A schematic illustration of soap-bubble-mediated pollination using a bubble gun. Influence of the type of surfactant on the (B) pollen germination ratio and (C) pollen tube growth.
(D) Pollen activity assay on an agar dish after soap-bubble-mediated pollination, at an A-20AB concentration of 0.5%.
(E) Impacts of the addition of different concentrations of A-20AB (0%–1%) and pollen grains (1–10 mg/mL) on the number of bubbles. Soap bubble quantity is increased by increasing the surfactant concentration and reducing pollen concentration. Data are represented as mean ± SEM in (B, C, and E).
Video S1. Preparation of Soap Bubbles Containing Natural Pollen Grains Using a Bubble Gun, Related to Figure 1
We initially tested the influence of five different surfactants—lauramidopropyl betaine (AMPHITOL 20AB [A-20AB]), sodium polyoxyethylene lauryl ether sulfate (EMAL E-27C [E-27C]), laurylhydoxysulfo betaine (AMPHITOL 20HD [A-20HD]), sodium polyoxyethylene alkyl ether sulfate (EMAL D3D [E-D3D]), and [N-cocoyl-(2-aminoethyl)-N-(2-hydroxyethyl)-N-sodiumcarboxymethyl] ethylenediamine (AMPHITOL 20YB [A-20YB])—on the pollen activity and bubble formation within the range of 0.2%–2.0% (Figures 1B, 1C, and S3). These surfactants were randomly chosen from among many commercially available products for their foaming ability to produce many soap bubbles by triggering a bubble gun once. The pollen activity assays indicated that the five surfactants showed a dose-dependent inhibition effect on pollen germination and tube growth (Figures 1B and 1C). Pollen germination ratio (G) was calculated as G = N/Nt × 100 (%), where N and Nt denote number of observations for pollen tubes by optical microscopy and total number of observations (100), respectively. Besides, the length of pollen tubes was measured by the results of direct observation and ImageJ software. The neutralized surfactant A-20AB demonstrated the highest performance in terms of pollen germination and tube growth compared with the other surfactants. In fact, the generated pollen tubes that were treated with a low A-20AB concentration were healthy after being subjected to soap-bubble-mediated pollination in an agar dish (Figures 1D and S4). In particular, A-20AB possessed the highest soap bubble formation ability among the utilized surfactants (Table S1). As such, we subsequently applied it as a foaming component in pollination solutions to create soap-bubble-carried pollen grains because of its powerful bubble-formation ability and superior capability of pollen germination and tube growth. A-20AB also showed dose- and incubation-time-dependent inhibition effects on pollen germination and tube elongation within the range of 0.0%–1.0% at a smaller interval (Figure S5). Nevertheless, the concentrations of A-20AB and pollen grains had a direct influence on the formation of soap bubbles (Figure 1E). We investigated the effects of different concentrations of surfactant and pollen grains on bubble formation by recording the number of bubbles produced per one trigger of the gun. Generally, higher surfactant concentration can help to make a lot of soap bubbles. Besides, larger number of pollen grains, which often have water-insoluble agglomeration, tends to disturb the formation of bubble membrane in the nozzle of the gun so that it decreases the number of soap bubbles. For instance, no bubbles could be produced at 0.0% or 0.2% A-20AB with pollen grains in a wide range of 1–10 mg/mL, whereas at least more than one soap bubble could be produced in the case of 0.4%–0.8% A-20AB and pollen grains of no more than 4 mg/mL. Moreover, with 1.0% A-20AB, 4–11 bubbles were formed within a pollen grain concentration of 1–10 mg/mL. In any case, pollen grains could be certainly carried by the formed soap bubbles, and the number of pollen grains loaded on the soap bubble increased with more pollen grains (Figure S6). Considering the above-mentioned inhibition property of surfactants against pollen grains and the available number of soap bubbles formed, we chose 0.4% A-20AB and a pollen grain concentration of 4 mg/mL for subsequent soap-bubble-mediated pollination. In this case, we could load a maximum of approximately 2,000 pollen grains on every soap bubble after triggering the bubble gun once (Figure S6).
To ensure successful pollination, we further optimized the components of the soap bubble solution under various physiological conditions (Figure S7), considering the pH value as an essential influencing factor on pollen germination and tube growth (
). The germination ratio reached its highest value (ca. 30.7%) at pH 7.0 (Figure S7A), and the pollen tube length exhibited a similar trend and reached its maximum value (ca. 1,010 μm) at pH 7.0.
Moreover, moderate addition of boron, calcium, magnesium, and potassium promotes pollen germination and tube length through direct or indirect mechanisms (
). Among them, calcium plays an important role in the germination and growth of pollen. It can improve the germination and growth of pollen due to the binding of calcium to pectate carboxyl groups along the pollen wall. Other elements such as boron, magnesium, and potassium can enhance the calcium effect. Here, we added individual concentrations of H3BO3, MgSO4·7H2O, CaCl2, and KCl to the solution to improve the pollen activity (Figures S7B–S7E). Although H3BO3 (0–60 ppm) showed a negligible effect on promoting pollen germination, we observed an obvious increase in the pollen tube growth at 5 ppm (p < 0.001) (Figure S7B). The pollen tube length actually reached 1,187 μm, which is 1.3 times higher compared with the control without H3BO3 (Figure S7B). Similarly, MgSO4·7H2O did not significantly promote pollen germination, but it obviously stimulated tube elongation at a low concentration of 0.1 mM (p < 0.01), within which the tube length reached a maximum value of 1,127 μm, which is 1.3 times higher compared with the control without MgSO4·7H2O (Figure S7C). With this aside, we found that CaCl2 concentrations within the range 0.1–2.0 mM dramatically improve the pollen germination ratio and tube growth (Figure S7D). More specifically, the germination ratio and tube length reached a maximum of 45.7% and 1,152 μm, respectively, at 1.0 mM CaCl2 (Figure S7D), which are approximately 1.6 and 1.3 times higher compared with the control without CaCl2. Moreover, although there was no sign that KCl promoted pollen germination, it otherwise improved the tube elongation significantly (p < 0.05) (Figure S7E). Eventually, at a KCl concentration of 1 mM, the pollen tube length reached a maximum value of 1,032 μm. Furthermore, under optimal conditions, the germination ratio and tube length of the pollen grains reached 45.0% and 1,232 μm, respectively, which were approximately 1.8 and 1.4 times higher compared with the control without chemical additives (Figure S7F).
Gelatine is a water-soluble protein that consists of large amounts of glycine, proline, and hydroxyproline, which might play an essential role in pollen germination and tube elongation (
). We added 0.2%–2.0% gelatine to the aforementioned pollination solution including optimized chemicals. Apparently, after treating the solution with 0.8% gelatine, the germination ratio and tube length reached a maximum value of approximately 50% and 1,363 μm, respectively (Figure S8A). We also added 0.0%–0.5% hydroxypropyl methylcellulose (HPMC) (
) to the solution and investigated its effect on pollen germination and tube elongation. HPMC was chosen to mechanically stabilize the films of the soap bubbles for robotic pollination (as mentioned earlier) and to moisturize the pollen particles in the solution to improve the pollen activity. Interestingly, the presence of HPMC (at a concentration of 0.2%; p < 0.01) significantly promoted the elongation of pollen tubes to a maximum value of 1,408 μm, whereas it slightly promoted pollen germination (Figure S8B). As we expected, HPMC also helped make the bubbles more stable, which was helpful in retaining the pollen grains on the thin film of the soap bubbles and transporting them to the targeted flowers.
The membrane thickness of a soap bubble can be determined using the following formula (
where τ, M, ρ, and R represent the membrane thickness, weight (ca. 7.7 mg), density (ca. 0.99 g/cm), and radius (ca. 1.6 cm) of the soap bubble. If we assume the value of π to be 3.14, then τ is 2.4 μm, which is a reasonable value for a conventional soap bubble within the thickness range of 1–10 μm as calculated using spectroscopic methods (
For artificial pollination at an orchard, we selected pears as our model mainly because of their potential large market size, as well as their easy handling during experiments. We initially examined the activity of pear pollen grains in an optimized soap bubble solution during the pollination process for 3 h, for comparison with other methods such as powder pollination and solution pollination, in addition to non-optimized soap bubble pollination. At the beginning of the optimized soap-bubble-mediated pollination, the pollen germination ratio and tube length were 49% (Figure 2A) and 1,221 μm (Figure 2B), respectively, which are about 1.9 and 1.5 times higher compared with a non-optimized soap bubble solution. After 3 h of pollination, the pollen germination ratio and tube length decreased to 28% and 990 μm, respectively, which were about 5.9 and 1.9 times higher compared with a non-optimized soap bubble solution. This result indicated that optimization remarkably improved the pollen activity in the bubble pollination process. More particularly, the pollen activity of the optimized soap-bubble-mediated pollination did not decrease, when compared with that of other pollination methods (powder pollination and solution pollination), after 3 h of pollination. In summary, soap-bubble-mediated pollination exhibited higher pollen activity compared with other conventional pollination methods for at least 3 h after the pollination process.
Figure 2Field Work of Soap-Bubble-Mediated Pollination
(A) Pear pollen germination ratio of various pollination methods during the pollination process for 3 h.
(B) Length of pear pollen tubes after soap-bubble-mediated pollination for 3 h incubation time in solution when compared with different pollination methods.
(C) Photographs of soap-bubble-mediated pollination for pear flowers using a bubble gun.
(D) Fluorescence microscopy images of pollen tube formation after soap-bubble-mediated pollination with two bubbles. Scale bar, 100 μm. The white arrows indicate the pollen tubes.
(E) Formation of young pear fruits by soap-bubble-mediated pollination. The inset image shows young pear fruits formed by 5–10 soap bubbles shot onto a flower after 53 days. Scale bar, 1 cm. ∗∗p < 0.01, ∗∗∗p < 0.001. Data are represented as mean ± SEM in (A, B, and E).
To demonstrate the feasibility of soap-bubble-mediated pollination, we shot different numbers (0, 1, 2, 5, 10, 20, and 50) of soap bubbles on natural pear flowers (Figure 2C). After overnight incubation at 25°C, the pistils of the flowers were cut and stained with aniline blue. Fluorescence microscopy showed that the pollen grains successfully landed on the pistils and that the growth of pollen tubes became clearly visible after pollination (Figures 2D and S9). In the control group, in which no soap bubbles containing pollen grains were attached, no pollen grains or tubes were present at all (Figure S9). Essentially, the number of pollen grains on each pistil increased as the number of soap bubbles got larger. The number of pollen grains decreased and the tube lengths became shorter after more than 10 bubbles were shot, which could be attributed to the toxicity of the accumulated pollination solution on the pollen grains and/or pistils. Therefore, less than 10 bubbles were subsequently shot onto the pear flowers at the farms.
Here, we pollinated the flowers of P. pyrifolia var. culta (N = 50) in an orchard using a soap-bubble-mediated technique with an optimized solution (Figures 2E and S10). Surprisingly, after shooting the soap bubbles onto the targeted flowers, young fruits formed after 16 days at a volume that was almost the same as that of conventional hand pollination with a spherical feather brush. The volumes of pear fruits obtained over time increased steadily and exhibited the same swelling tendency as that of flowers pollinated after 16 days. The 2–10 soap bubbles that hit the flowers had no influence on pollination. Anyway, it is an amazing effect because even only a couple of bubbles can effectively achieve bearing fruits. Moreover, control flowers, which did not undergo any kind of pollination, yielded the smallest volume of young fruits, probably because of the influence of naturally pollinating insects as well as the non-specific and imperfect attachment of scattered pollen grains in the wind from other hand pollination work by farmers in the same orchard. Specifically, because of these environmental factors, the fruit-bearing rate of the control that did not undergo any pollination was only about 58% (Figure S11). Moreover, the rates of both soap-bubble-mediated pollination and hand pollination were approximately 95%, and no significant differences were observed between them. Apparently, such results demonstrated that soap-bubble-mediated pollination is effective not only for the expression of fertility of pollen grains but also for the substantial production of pear fruits.
Figure S12 and Table S2 show the number of pollen grains used in different pollination methods. For the tests, we used the inexpensive Lycopodium clavatum spore as a model to replace costly natural pollen grains. Conventional hand pollination with a spherical feather brush required approximately 1,747 mg pollen grains for one instance of pollination, whereas machine-based pollination required approximately 165 mg pollen grains. Moreover, a solution hand spray ejected 3.2 mg of pollen grains by one-push spraying. Based on our calculation, the soap-bubble-mediated method consumed only 0.06 mg pollen grains for essential pear pollination via two or three soap bubbles. Thus, the supersizing reduction effect of pollen grains by soap bubble-mediated pollination should be beneficial for practical cost cutting of labor and pollen manufacturing.
Robotic Pollination
Our final goal was to perform robotic pollination using a drone and soap bubbles. However, as drones often generate strong downstream wind from their propellers, the soap bubbles should be mechanically stabilized to ensure that they are correctly shot at the flowers while retaining their original structure. Indeed, the above-mentioned normal soap bubbles that were prepared by bubble formation tools vanished quickly. Herein, we loaded 2% HPMC into the solution to physically stabilize the soap film. Interestingly, soap bubbles containing 2% HPMC and 1% A-20AB turned out to be very stable self-standing bubbles with a diameter of approximately 2 cm, prepared using a bubble blowing tool (Figure 3A). Unexpectedly, most of the mechanically stabilized soap bubbles did not quickly vanish for at least 10 min at 25°C. What was more surprising was that some of them maintained their spherical shape for as long as 5 h, and they also could withstand compression as they were recorded to tolerate up to 0.03 N maximum proof compressive load (Figure S13). During the compression test, we heard faint pops as the soap bubbles burst, signifying their HPMC-improved mechanical stability.
Figure 3Robotic Pollination Using Mechanically Stabilized Soap Bubbles
(A) A photograph illustrating a commercial bubble wand (left) and a mechanically stabilized soap bubble on a glass slide (right).
(B) Number of adsorbed pollen grains per bubble for various flowers.
(C) Pollination success rate of different types of flowers by attachment of a single soap bubble containing pollen grains.
(D) Image of a robotic pollinator consisting of a UAV and a bubble maker.
(E) Robotic pollination using an autonomous robotic pollinator and L. japonicum flowers.
(F) Effect of the speed of a robotic pollinator on the success rate (hitting of bubbles onto flowers) of artificial pollination. Data are represented as mean ± SEM in (B, C, and F).
Accordingly, mechanically stabilized soap bubbles exhibited a membrane thickness of 4.1 μm, thicker than the above-mentioned conventional soap bubbles (2.4 μm), probably because of the formation of a thick polymer film due to the highly condensed HPMC.
Besides, we confirmed that the pear pollen grains still demonstrated strong activity even after being pollinated with soap bubbles containing 2% HPMC (Figure S14). The kinematic viscosity and density of the prepared soap bubble solution were 7,530 cSt and 1.023 g/cm, respectively. The relatively high viscosity of the solution was also useful for pollen dispersion (Figure S15); in fact, various pollen grains, such as those of L. japonicum, Rhododendron pulchrum, and Campanula persicifolia, were dispersed and suspended well in the highly viscous soap bubble solution (Figure S15A). Moreover, we observed a Tyndall effect (
) in L. japonicum pollen colloidal particles in a quartz cuvette at 650 nm laser irradiation (Figure S15B). Optical microscopy directly revealed that all types of pollen grains were individually and well dispersed in the solution (Figure S16). We counted the number of pollen grains on each soap bubble (L. japonicum, ca. 269 particles; R. pulchrum, ca. 304 particles; C. persicifolia, ca. 312 particles) using optical microscopy after smashing it between two cover glasses (Figures 3B and S17). Subsequently, we confirmed the concentration dependence of spore particles on loading number onto the soap bubble solution (Figure S18). The mechanically stabilized soap bubbles were also observed to stay on the pistil of each flower for a while, ranging from 10 s up to 5 min (Figure S19), and then they vanished depending on the pistil size and the amount of sticky pollenkitt (
) on the pistil. More particularly, the soap bubbles vanished shortly after attaching to the largest pistil in the flower of L. japonicum, which had lots of viscous pollenkitt when compared with other flowers. Interestingly, even after the pistil of each flower was hit only by one soap bubble containing pollen grains, followed by overnight incubation, we observed the growth of fibrous pollen tubes, indicating successful pollen fertilization (Figure S20). The control flower, which was not hit by any soap bubble including pollen grains, did not show any pollen grain adsorption or pollen tube growth at all, for all types of flowers (Figure S21). In essence, single-soap-bubble-mediated pollination reached a maximum success rate of 90% for the flowers of L. japonicum (Figure 3C), which was higher compared with the flowers of R. pulchrum and C. persicifolia, probably because of the bigger pistil size and large amount of pollenkitt (Figure 3C). Pollination success rate (PSR) was calculated as
, where N and Nt denote the number of actual pollen tubes observed under fluorescence microscopy and the total number of observations (10), respectively.
An automatic bubble maker was likewise used in combination with a drone (Figure 3D). This bubble maker continuously produced approximately 5,000 mechanically stabilized soap bubbles per minute (Figure S22 and Video S2). To demonstrate the use of robots in pollination, we utilized a commercially available, fully automatically controllable UAV as a robotic pollinator. Here, a bubble maker was attached to the body of the UAV (Figure 3D). The movement of the robotic pollinator was controlled using a fully automatic operation system (
) equipped with a global navigation satellite system (GNSS) (Figures S23A–S23C). Then, soap bubbles were incessantly shot onto the targeted fake L. japonicum flowers from a height of about 2 m at an angle of about 70–80° (Figures 3E and S23D, and Video S3). Apparently, the physical impact caused by the downstream wind from the UAV contributed to the immediate destruction of the soap bubbles after hitting the pistil of the L. japonicum flower (Figure S24). The UAV exhibited a downstream wind speed of about 4.5 m/s as it hovered steadily in the air; nonetheless, this downstream wind speed under the UAV increased to 5.8 or 8.7 m/s when it moved at a constant speed of 2 or 4 m/s, respectively. Based on the results of the robotic pollination experiment, the success rate of soap bubbles hitting the flowers depended on the speed of the UAV (Figure 3F). Although the isolated single soap bubble was able to hit a pistil on a lily flower as shown in Figure S24, most soap bubbles were formed from the UAV as a bunch of bubbles seemed like “cluster.” Thus, the number of the soap bubble cluster, which could be accurately struck to each flower, was recorded for figuring out the average success rate when the UAV moved over the flowers at a different speed (2–10 m/s). After all, we achieved a higher than 90% success rate at a velocity of 2 m/s; more importantly, we observed the growth of natural pollen tubes of real L. japonicum flowers (Figure S25). These results are a clear indication that a robotic pollinator equipped with a soap bubble maker would successfully promote effective flower pollination.
Video S3. Fully Automatic Pollination by a UAV Equipped with a Mechanically Stabilized Soap Bubble Maker, Related to Figure 3
Discussion
In this study, we designed and prepared chemically functionalized soap bubbles for the artificial pollination of flowers using various bubble-making devices. Herein, these chemically functionalized soap bubbles demonstrated interesting properties, such as the delivery of pollen particles, enhancement of pollen activity, excellent mechanical stability, suitable size and geometry of flower pistils, softness and high flexibility to prevent damage to flowers, adhesiveness on flowers for effective pollination, and simple ejection ability on a large number of soap bubbles from the devices. In addition, we successfully pollinated the flowers of P. pyrifolia var. culta in an orchard by soap bubbles using a bubble gun, which consequently formed young pear fruits. We also integrated soap bubbles within a drone for fully automatic pollination of L. japonicum flowers. This study is the first exploring the unique properties of soap bubbles as a material used for the artificial pollination of flowers using different types of bubble-making tools and an autonomous controllable drone. We expect our multidisciplinary approach combining soap bubbles and drone technology to lead to innovative developments in the field of agricultural engineering. We also assume that our findings will pave the way for discovering artificial pollination methods that can address relevant global issues such as the decline in pollinator insects, the heavy labor involved in artificial pollination, and the soaring costs of pollen grains.
Limitations of the Study
The surfactants used in this study are biocompatible, but their elimination in the environment might cause their accumulation and difficult degradation. Therefore, we are trying to use eco-friendlier and edible surfactants for future practical pollination. Investigations of the soap bubble-mediated pollination using the automated robotic drone at field or orchard scale are also future challenges because the use of a prototype artificial pollinator to spray bubbles caused a lot of waste as most bubbles would miss the flower. Further innovative technologies, such as state-of-the-art localization and mapping, visual perception, path planning, motion control, and manipulation techniques, would be essential for developing autonomous precision robotic pollination.
Resource Availability
Lead Contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Eijiro Miyako ([email protected]).
Materials Availability
This study did not generate new unique reagents.
Data and Code Availability
The data that support the findings of this study are available from the corresponding author on reasonable request.
This work was supported by a Japan Society for the Promotion of Science (JSPS) KAKENHI Grant-in-Aid for Scientific Research (A) (Grant number 19H00857); JSPS KAKENHI Grant-in-Aid for Scientific Research (B) (Grant number 16H03834); and JSPS KAKENHI Fund for the Promotion of Joint International Research (Fostering Joint International Research) (Grant number 16KK0117). The author also thanks Kao Chemical Co., Ltd. for providing the surfactants. The authors also thank National Institute of Advanced Industrial Science and Technology for their partial support with some experiments.
Author Contributions
E.M. conceived and designed the experiments and prepared the manuscript; E.M. and X.Y. performed the experiments and analyzed the data. All the authors discussed the results and contributed to the writing of the manuscript.
Elamvazhuthi, K., and Berman, S. (2015). Optimal control of stochastic coverage strategies for robotic swarms. in: 2015 IEEE Int. Conf. Robot. Autom., IEEE 1822–1829.
Strader, J., Nguyen, J., Tatsch, C., Du, Y., Lassak, K., Buzzo, B., Watson, R., Cerbone, H., Ohi, N., Yang, C., and Gu, Y. (2019). Flower interaction subsystem for a precision pollination robot. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2019, Macau, China.
30373-4&id=fx1.jpg&_x_tr_sl=en&_x_tr_tl=es&_x_tr_hl=es-419&_x_tr_pto=sc){kind=link}
30373-4&id=gr1.jpg&_x_tr_sl=en&_x_tr_tl=es&_x_tr_hl=es-419&_x_tr_pto=sc){kind=link}
30373-4&id=gr2.jpg&_x_tr_sl=en&_x_tr_tl=es&_x_tr_hl=es-419&_x_tr_pto=sc){kind=link}
30373-4&id=gr3.jpg&_x_tr_sl=en&_x_tr_tl=es&_x_tr_hl=es-419&_x_tr_pto=sc){kind=link}