Biohybrid device mimicking wild oat seeds

Abstract

A biohybrid device configured to self-bury into a substrate to transport a payload, the device comprising an artificial capsule (10) composed of an elongated and pointed body of biodegradable material, the artificial capsule being configured for a payload to be attached or incorporated thereto, and two humidity-driven motors (21, 22), each of which comprises a base beam (21a, 22a) connected to the artificial capsule (10) and capable of twisting in response to a change in humidity, and a tail (21b, 22b) connected to the base beam (21a, 22a) through a knee point (21c, 22c) and capable of bending, wherein the two humidity-driven motors are composed of respective awns of oat seeds attached to the artificial capsule (10) and capable of elastically engaging with each other.

Description

[0001] Biohybrid device mimicking wild oat seeds

[0002] The present invention relates in particular to the field of agricultural and forestry techniques.

[0003] In this field, solutions have been proposed that mimic the dispersion mechanisms of seeds of some plants to make devices that may be dispersed in difficult-to-reach geographic areas and are able to penetrate into the soil. Such a solution is disclosed, for example, in US 2022 / 0322599 Al, which describes a biodegradable device comprising a capsule that may be functionalized with a payload and a bio-inspired actuation system that reacts to variations in humidity. Specifically, this actuation system mimics the spiral mechanism which is present in the seeds of some plants of the genus Erodium, and which allows these seeds to penetrate slits present in the soil. If this mechanism is actually useful in stimulating the device to self-bury, it is not effective in achieving displacements of the device on the soil, in the event that burial sites are not available in the immediate vicinity of the area where the device is located. Furthermore, the device described in US 2022 / 0322599 Al appears massive and made of wood, and therefore could theoretically require a major variation in the humidity of the environment to stimulate the self-burial of the device.

[0004] The inventors therefore turned their attention to oat seeds, in particular those of wild oat (Avena sterilis). As is well known, it is one of the most widespread plant species due to its extremely effective dispersion mechanism which is based on repeated snap movements of two interacting twin awns, and which has allowed this species to colonize various habitats in Asia, Europe and Africa.

[0005] In that regard, the article by Lindtner T. et al, R. Repetitive hygroscopic snapping movements in awns of wild oats. Acta Biomaterialia 135, 483-492 (2021) analyzes the simultaneous movements of the two twin awns of wild oat seeds in response to variations in humidity, with particular attention to the accumulation of stress and the sudden release of such stress linked to the rotation of such awns. This release of stress induces jumps in the wild oat seed and changes in its direction of movement on the soil. The article also suggests the possibility of using this reversible jump mechanism for making jumping robotic systems, but does not provide guidance regarding the embodiments. One object of the present invention is to provide a solution capable of at least partially resolving the aforementioned drawbacks relating to the prior art.

[0006] In view of these objects, the subject matter of the invention is a biohybrid device configured to self-bury in a substrate for transporting a payload, the device comprising an artificial capsule composed of an elongated and pointed body of biodegradable material, said artificial capsule being configured for a payload to be attached or incorporated thereto, and at least two humidity-driven motors, each of which comprises a base beam connected to the artificial capsule and capable of twisting in response to a change in humidity, and a tail connected to the base beam through a knee point and capable of bending, wherein said at least two humidity-driven motors are composed of respective awns of oat seeds (more precisely, seeds of the Avena genus) attached to the artificial capsule and capable of elastically engaging with each other.

[0007] The aforesaid biohybrid device is capable of mimicking the dispersion movements and biomechanical performance of natural oat seeds, more specifically oat seeds of the genus Avena, and is likewise capable of functioning as a low-cost, scalable and sustainable selfdispersed system, paving the way to reforestation and precision agriculture applications.

[0008] Further features and advantages of the device according to the invention will be presented in the following detailed description which refers to the accompanying drawings, provided merely by way of non-limiting example, in which:

[0009] Fig. 1 is a graphical representation of a device according to the invention;

[0010] Fig. 2 shows: a, b) Images of a natural awn of Avena sterilis under a) dry and b) humid conditions. Following humidity absorption, the humidity-sensitive base beam rotates by a torsion angle, and the tail (not sensitive to humidity) bends accordingly (due to the structured connection at the knee point joining the base beam and the tail), c) Torsion angle of the base beam (9), measured at the knee point, as a function of time, for two twin awns (awm, awm) of Avena sterilis. Once sprayed with water vapor, both awns have continuously changed their configuration over time until they reached a steady state condition, with awm having reached a greater torsion interval than awn?. For both awns, the steady state was reached in approximately : considering a diffusivity coefficient D = IO-10÷ 10-9m2 / s, the humidity traveled a distance 0 pm in the tissue of the awn, which is consistent with the characteristic dimension of the awns considered in the transverse direction (=420[pm]). d) Model torsion angle (0, continuous curve) and associated torsion velocity (0, dashed curve) with respect to time, for the base beam of awm. e) Model torsion angle (0, continuous curve) and associated torsion velocity (0, dashed curve) with respect to time, for the base beam of awn2. For both rods, the characteristic time T of the actuation dynamics guided by the absorption of humidity and consequent swelling was quantified by fitting the experimental records (circles - except the first, to exclude the transient effects probably associated with the diffusion of humidity into the air) to the following pattern of the model: 0(t) = 0OO[1 — exp (— t / r)], where 0^ indicates the steady state value. The torsion velocity was defined by differentiating the aforesaid fitting progression. Consistent with previous observations, awm showed a slightly faster dynamic than awm, as described by the resulting times:1= 220 s and T2— 240 s, respectively for awm and awm;

[0011] Fig. 3 shows: A) Diagram showing the initial placement of the awns (base rods) with respect to the capsule. B) Effect of the angles of placement of the base rods ^cot and 0^oton the resulting number of contacts between the awns: given the hygroscopic behavior of the awns, their placement may be set, for example, in order to maximize the number of contacts;

[0012] Fig. 4 shows: a, b) Results of the resistance force tests of the capsule of natural seeds and biohybrid samples a) without and b) with rotation using different soils as substrates, including artificial and natural soil. The artificial soil is composed of 1 mm glass beads. The natural soil is composed of topsoil, sand and clay, c, d, e) c) Natural seeds and biohybrid samples in dry and humid conditions after the use of a humidifier, d) Contact forces exerted by the awns of the natural seeds and biohybrid samples on a load cell and e) associated torque. Three force levels were considered, namely 1 / 3 (F1 / 3), 2 / 3 (F2 / 3) and 3 / 3 (Fmax) of the corresponding maximum value, f) Illustrative results of the mathematical model that describes the dynamics of the awns, based on the experimentally recorded torsion and taking into account the flexibility of the awn and the mutual contact. A test case is considered with two contacts: for both contacts, the characteristic configurations of the awns (left) and the contact forces between the awns (right) are shown. The relatively low variability of the forces with respect to the only free parameter of the model (i.e. k in the force graphs) allows the model to be predictive, and the results obtained are in good agreement with the load cell experiments;

[0013] Fig. 5 is a graphical representation of two complementary molds that may be used to manufacture a device according to the invention.

[0014] With reference to Fig. 1, a biohybrid device according to the invention is shown, which is configured to self-bury into a substrate to transport a payload.

[0015] The biohybrid device in question comprises:

[0016] (I) a biodegradable capsule 10 which reproduces the shape of the head of an oat seed and is functionalized with a payload. 10a denotes a tip of the capsule 10;

[0017] (II) at least two interacting, humidity-guided oat seed awns 21, 22, which are connected to the capsule 10. Each of the awns 21, 22 comprises a base 21a, 22a connected to the capsule 10, and a tail 21b, 22b via a knee 21c, 22c.

[0018] The awns 21, 22 are composed of dead tissues and, when they change from dry to wet conditions, the cells of the shell of the base beam 21a, 22a tilt in their alignment about the longitudinal axis of the awn 21, 22, while the flexion of the tail 21b, 22b is guided by the structured connection in the knee point 21c, 22c, which joins the base beam and the tail. The resulting movement causes the tangling / contact between the twin awns and the surrounding environment, finally allowing the locomotion of the seed by means of repeated and reversible jumps on the soil and the penetration into the ground.

[0019] According to a preferred embodiment, the capsule 10 may also comprise a plurality of natural oat seed hairs 10b, which are manually fixed to the biodegradable capsule 10. Such hairs advantageously make the self-burial of the device into the soil non-reversible, as will be better explained below.

[0020] Preferably, the capsule 10 may be made of different molding materials, including (but not limited to):

[0021] 1) organic materials and compostable resins (e.g. earth; compost; biomass waste-based resins)

[0022] 2) thermoplastic materials (for example polycaprolactone; polylactic acid)

[0023] 3) dissolvable materials (e.g. sugar-derived materials; isomalt).

[0024] According to a preferred embodiment, the capsule is also edible.

[0025] According to an even more preferable embodiment, the capsule 10 is made of a flour-based material.

[0026] The shape of the capsule based on oat seeds was chosen as it represents the best profile in terms of self-burial performance.

[0027] The payload incorporated / attached to the biodegradable capsule may be composed of seeds or fruits of different plants, nutrients, pesticides, spores, sensors, etc. Preferably it may include seeds and fertilizers, in order to promote reforestation.

[0028] The biohybrid device described above mimics the dispersion movements and biomechanical performance of natural oat seeds, and may achieve, for example, resistance forces up to ~0.38 N and torques up to -100 pN m. The biohybrid device proposed herein functions as a low-cost, scalable and sustainable self-dispersed system, paving the way to applications in reforestation, precision agriculture.

[0029] As mentioned above, the proposed biohybrid device exploits the inter-awn contacts both to move on the soil (unlike the device illustrated in US 2022 / 0322599 Al) and to penetrate into the ground. To optimize the delivery of the load, it is therefore advisable to study the contacts between the awns quantitatively and to optimize the number thereof.

[0030] For a given pair of twin awns, the number of contacts between awns depends essentially on their initial placement with respect to the capsule. In fact, once the hygroscopic behavior of the awns has been characterized, it is reasonable to assume that exposure to different humidity conditions mainly affects their (almost static) torsion / flexion in terms of velocity, rather than the contact between the awns. Therefore, it is possible to exploit the experimentally recorded hygroscopic behavior of the awns (see Fig. 2) to estimate the number of contacts as a function of the initial placement of the awns. To this end and with reference to Fig. 3(A), each base beam 21a, 22a may be positioned in a sagittal plane of the capsule 10 defining an insertion point (k=1.2 in the example discussed here), the angle between the base beam 21a, 22a and the longitudinal axis of the capsule 10, and the angle of rotation 0 .ootabout the longitudinal axis of the base beam 21a, 22a (which determines the plane defined by said base beam and by the relevant tail 21b, 22b - the latter is not shown in the figure being considered). For example, assuming that P and are defined as in the natural oat seed model, it is possible to vary the number of contacts by varying 0^otand 0^ot. In particular, it is possible to model each awn as an arm with two- articulated elements, with the rotation angles of the base beam and the flexion angles of the tail defined by the recorded hygroscopic behavior, thus determining the number of contacts based on the kinematics of the rigid body. An illustrative and non-exhaustive result is shown in Fig. 3(B) (considering the hygroscopic behavior of the awns over a time interval of 300 s, after which no further contact occurs, assuming the hygroscopic characteristics shown in Fig. 2). It is therefore possible to exploit the proposed approach to position the awns so as to maximize the number of contacts. Furthermore, to determine quantitatively the dynamics of the awns upon engagement / disengagement of the contact, it is possible to fully describe the humidity-guided biomechanics of the awns using the mathematical model reported below.

[0031] Mathematical model

[0032] An analytical biomechanical model was introduced to quantitatively describe the dynamics of the twin awns (hereinafter labeled as aw and aw ), based on their experimentally recorded torsion (see Fig. 2). In fact, once the hygromorphic behavior of the individual (natural) awns has been characterized, a physics-based modeling framework may be used to perfect the design of the proposed biohybrid machines, based on the dynamics of the awns.

[0033] The current geometry of the capsule was considered, as well as the measured morphometric data of the awns and their relative placement with respect to the capsule (see Table 1 below).

[0034] Table 1

[0035] The measurements were taken on five pairs of twin awns

[0036] #The measurements were taken on one pair of twin awns

[0037] Each of the four segments involved (i.e. the base beam and the tail of awm and awm) was modeled as a flexible beam with a circular cross-section: each base beam was kinematically coupled to the corresponding tail (since the flexion of the tail, as a first approximation, is proportional to the torsion of the base beam - Lindtner T. et al, R. Repetitive hygroscopic snapping movements in awns of wild oats. Acta Biomaterialia 135, 483-492 (2021)), and contact was allowed between the two tails. Therefore, each awn was operated individually based on the torsion of the base beam determined by the humidity (Fig. 2), and the contact interaction was taken into account between the two tails.

[0038] More specifically, considering the slenderness of the awns (whereby the shear / elongation effects are negligible with respect to those of flexion / torsion), each of the flexible segments considered was modeled using the following (classic) Cosserat / Kirchoff equations - Antman

[0039] S. S., "Problems in nonlinear elasticity. Nonlinear Problems of Elasticity", 513-584 (2005): where denotes the partial differentiation of (•) with respect to , s represents the arc length, x denotes the position (of a point of the rod, as tacitly understood hereinafter) represented in a global Cartesian system (for simplicity, normal font weight is used for both scalars and vectors, since the classical entities involved leave no room for ambiguity), q is the quaternion that determines the orientation of the local reference system (i.e. attached to the body) with respect to the global one, v is the velocity in the local system, co is the angular velocity in the global system, n is the internal force in the global system and m is the internal torque in the global system. Moreover, t denotes time, R = R(q) is the rotation matrix that maps the vector representations from the local to the global system, e3= [0,0, 1]T(the matrix formalism is used below for ease of reading), u represents the curvature in the local system (whereby the matrix fl = fl(u) is well defined), p denotes the density, A is the cross- sectional area, J represents the area moment of the inertia tensor, f is the external force per unit length, and the operator7maps a given vector z = [a, b, c] in the antisymmetric matrix z = [0, — c, b; c, 0, — a; — b, a, 0] (the semicolon indicates the division of the lines). For clarity, it is recalled that Eqs. (1-2) derive from the base geometry of the curves, Eqs. (3-4) derive from the relative mathematical constraints, while Eqs. (5-6) provide the balance of the linear and angular momentum, respectively, for the relevant rod. Finally, in Eq. (6) external moments do not appear per unit of length, because they are not relevant to the interawn contact considered.

[0040] In order to take into account the possible contacts between aw and awn?, a simple non- linear penalty sub-model was introduced, as shown below. Given a point P> on the taili and a point P on the taih, at relative distance , it is assumed that they are in contact if are the radii of the cross-section at, respectively, P and P^2and p = 1.5 allows a certain margin in the detection of the contact (with respect to a net limit, here given by Q = 1, which may cause instability during (12) numerical integration). The force exerted by the taili on the tail , which acts along the direction δ(12)ijconnecting to P(2)jwas then calculated as follows: f ,12'’ = k F(λ(12)ij) δ(12)ijwhere k is a free parameter of the model and F(y) = (max(y, 0))2. It is noted that the aforementioned quadratic progression is a simple modeling law that simultaneously allows for a soft contact start and a stiffening following the reduction of the relative distance. Furthermore, k is the only free parameter of the model. To the contact between and P2the contributions corresponding to the linear force density f were added, integrating the Eq. (5) numerically. In particular, -f / δs1and / δs2were added for the tail and the tail2, respectively, where δs1and δs2indicate the corresponding dimension of spatial discretization (see below).

[0041] Furthermore, to close the problem considered, the differential equations to the partial derivatives cited above were integrated with the following constituent relationship: where u* is a given reference curvature (which depends, in the general case, both on space and time), δ is a (positive) damping factor that takes into account the internal dissipation of the material (commensurate with the available experimental data, a scalar damping was assumed and, for simplicity, was set to a level low enough not to cause non-physical overdamping of the dynamics of the awn during numerical integration, see below), and K = (π r4 / 4)E[l,0,0; 0,1,0; 0,0,l / (l + v)] is a stiffness matrix that determines the elastic response of the material, based on Young’s modulus E and Poisson’s ratio v. In particular, for each of the four segments considered (i.e. the base beam and the tail of awm and awn?), the reference curvature was assigned on the basis of the humidity-guided torsion data, as described below. Considering that the tails of the awns are not sensitive to humidity, u* = [0,0, 0]Twas placed for both the tails of the awm and the awn?. Otherwise, for the base beam of awn1 and awn2, the torsion induced by the absorption of humidity was defined respectively by where the lengths of the base rods Lbland Lb2were known from the measured geometry and the torsion angles guided by humidity absorption, i.e. 0mland 0∞2. as well as the characteristic times for humidity absorption, i.e. and . were set based on the relative measurements (Fig. 2). Note that only the torsion angles guided by humidity, i.e. 1 (t) = , were used to set the reference curvature: the corresponding flexion angles of the tail (i.e., the angles between the initial portion of each tail and the extension of the final portion of the corresponding base rod) were used to set the boundary conditions of the problem, as illustrated below.

[0042] For each awn, the aforesaid governing equations for the base beam have been resolved, considering a curvilinear abscissa sbE [0, Lb], and the governing equations for the tail, considering a curvilinear abscissa stE [0, Lt]. The two sets of equations were coupled by the physical continuity at the knee point (identified by sb= Lband st= 0) and the kinematic constraint introduced by the flexion angle 0*(t) guided by the humidity between the base beam and the tail, again at the knee point. In particular, the flexion angles and respectively for awn1 and awn2, respectively, were provided by with based on previous measurements of the flexion angle - Lindtner T. et al

[0043] "Repetitive hygroscopic snapping movements in awns of wild oats", Acta Biomaterialia,

[0044] 2021. For each awn, the related boundary conditions are therefore as follows: with and which denotes the ordinary derivative of 0* with respect to t (in the present case, in fact, both 6* and (p* depend only on time). As mentioned above, the position x0and the rotation Rowere defined on the basis of the measured morphometric data of the awn: for convenience of definition, x0was chosen so as to lie on the sagittal plane which (approximately) bisects the capsule through the wings, and Rowas determined by the angle between the axis of the capsule (on the sagittal plane) and the axis of the base rod, once the angle 0rootwas assigned between the sagittal plane and the plane defined by the initial placement of the awn. Overall, 0rootmust be considered a fixed parameter of the model, corresponding to the effective degree of freedom that, in practical implementations, allows the initial placement of the awn to be rotated about the axis of the base rod, without altering the point of wedging of the awn on the capsule. It should be noted that, while Eqs. (8-11) and (18-19) respectively describe the static wedging of the base beam and the no-load condition of the tip of the tail, Eqs. (12-17) define a non-trivial position coupling at the knee point, where the time-variable flexion plays a fundamental role.

[0045] The governing equations were discretized numerically by adopting an approach based on the method of lines - Saucez P. et al, "Adaptive method of lines", CRC Press, 2001; J. W. Thomas, "Numerical partial differential equations: finite difference methods", Springer Science & Business Media, 2013, vol. 22.

[0046] In particular, the semi-discrete formulation was obtained by discretizing earlier in time with a backward differentiation formula BDF2 (which also allows the reformulation of Eq. (7) as an explicit relationship for the semi-discrete variable corresponding to it). The spatial integration was then performed for (the semi-discrete counterpart of) Eqs. (1-6) using a classic Runge-Kutta pattern: the unknowns {x, q, v, a>, n, m} were integrated photographing on n(sj, = 0) and m(sb= 0), and adopting as a target the no-load condition given by Eqs. (18-19) (considered both for aw and for aw ). More specifically, the governing differential problem has been reformulated in non-dimensional terms, and the aforesaid numerical integration was implemented on a common scientific computing environment (Matlab, MathWorks, USA), refining the discretization until results that are independent of the discretization were obtained. Finally, the following literature values were adopted for the biomechanical parameters involved: p = 600 kg / m3Rusch F. et al (SciELO Brasil, 2019)), E = 5.5 ■ 109Pa (Elbaum R. et al, The role of wheat awns in the seed dispersal unit. Science 316, 884-886 (2007)), v = 1 / 5 (Molenda M. et al, Determination of the elastic constants of cereal grains in a uniaxial compression test. International Agrophysics 16, (2002)). As noted above, the only free parameter of the model is k: its range is physically delimited by the fact that, for values that are too low, the contact is not physically reached (i.e., the resulting interaction force is so low that the tails end up crossing in a non-physical manner), while for values that are too high, a non-physical vibrating / jumping contact is obtained, caused by reactions of the impulsive type already at the start of the contact. The “physical” interval was then explored for k through numerical sweeps, considering in particular the test case associated with Fig. 4(f), wherein two contacts occurred between the awns (with — 3.2 and 0 / oot= —4.1 rad). It was found that physically representative behaviors were achieved for 0.8 < k < 1.4 (the non-dimensional formulation was directly used to set k, so that it is consistently presented as non-dimensional also in the aforementioned figure) and, even more remarkable, the numerical results were weakly dependent on the specific value of k in this range (Fig. 4(f)). This, in turn, made it possible to consider the model as predictive, and the force results obtained by the model were found to be in good agreement with the related experimental force measurements.

[0047] To conclude, a number of observations should be added. The model introduced describes the dynamic interaction between the twin awns, also taking into account their flexibility, based on humidity-induced placement changes (which modulate the torsional curvatures of the base rods, also modulating the relative flexion angles between the base rods and the tails). A physically representative modeling framework was therefore introduced which, notwithstanding the simplifications intrinsically introduced by the analytical approach (including its numerical integration), could be used both to quantitatively study natural Avena sterilis seeds and to design the proposed biohybrid device. For example, the proposed model allows richer design spaces to be explored (e.g., allowing a wider set of relative positions between the awns and the capsule), also based on the fact that the number of contacts may be estimated in advance from a simplified rigid kinematic and non-contact model, as discussed above. At the same time, the proposed model could be further improved by explicitly describing the diffusion dynamics of the humidity, so as to be able to adapt the response of the awns to specific environments and (controlled) humidity absorption conditions. Manufacturing

[0048] By means of computerized micro-tomography, a detailed 3D reconstruction of the oval Avena sterilis L. capsule was obtained, which was then used to design two complementary miniaturized molds Ml, M2 for the artificial capsules and the junction connections with the awns (see Fig. 5). The molds were directly made by 3D printing with two-photon lithography, allowing high resolution and reproducibility even at the level of the tip. The cavities of the complementary molds Ml, M2, which cooperate to reproduce the shape of the head of an Avena sterilis seed, are respectively denoted by Cl and C2.

[0049] Samples of the device according to the invention were made by positioning a pair of natural Avena sterilis awns in seats S1M1, S2M1, S2M1, S2M2 obtained in the molds Ml, M2, and casting a flour-based material so as to overmold it over an end portion of the base rods of the awns. Subsequently, the artificial capsule thus obtained was coated with a 10% solution of ethyl cellulose to allow the adhesion of natural Avena sterilis hairs.

[0050] A morphometric analysis of the artificial capsules and the twin awns of the samples obtained was carried out (see Table 2 below).

[0051] Table 2

[0052] The measurements were taken on three pairs of twin awns

[0053] The molds made with 3D micro-printing may be reused several times, allowing for the creation of multiple samples of the device according to the invention, using sustainable and low-cost materials.

[0054] Biomechanical characterization

[0055] A biomechanical analysis of the parts of the capsule and the awns both in the natural seeds and in samples of the device described above was performed (Fig. 4).

[0056] First, the rigidity and resistance forces of the natural and artificial capsules were characterized. To study the rigidity of natural and artificial capsules, a nanoindentation system was used with a miniaturized probe, similar to the methods used to characterize the capsule material in Geraniaceae (Fio ello I. et al in 2022 IEEE 5th International Conference on Soft Robotics (RoboSoft). 255-260 (IEEE)) or the materials of the Poaceae (Elbaum R. et al The role of wheat awns in the seed dispersal unit. Science 316, 884-886 (2007)). The natural capsules of Avena sterilis show a local average accumulation modulus (E) of 7.7+2.6 GPa, similar to that of artificial capsules produced with flour (E= 6.4+3.3 GPa), demonstrating that the biodegradable and low-cost flour-based material has adequate rigidity to be chosen as the optimal casting material in the device according to the invention. To study the forces necessary for natural seeds and biohybrid devices to dig into the soil, the resistance forces of natural and artificial capsules were measured without and with rotation during their vertical penetration into different types of granular substrates, including artificial and natural dry soils with different densities and particle sizes (Fig. 4(a), (b)). In the experimental tests without rotation, the resistance forces in the natural capsules ranged from lower values of 0.06+0.01 N in artificial glass beads to higher values of 0.37+0.07 N in natural sandy soil, similarly to artificial capsules where the resistance forces increased from 0.08+0.01 N in artificial glass beads to 0.38+0.12 N in natural sandy soil (Fig. 4(a)). To verify the reduction of the resistance of the natural and artificial capsules through the rotary movement, the resistance forces of the natural and artificial capsules in the different substrates during rotation were measured (Fig. 4(b)). In both natural and artificial capsules, a decrease in resistance forces was observed in all substrates (Fig. 4(b)), demonstrating that natural Avena seeds and biohybrid devices may reduce their granular resistance with the rotary movement. Furthermore, a reduction in resistance was observed with the rotation in natural and artificial capsules not only in dry environments, but also in humid soils. Overall, the resistance forces observed in the natural and artificial Avena capsules are approximately ten times greater than those observed in the previously reported self-burying seeds, such as the Erodium and Pelargonium seeds, probably due to the considerable differences in the dimensions, mechanics and shapes of the capsules.

[0057] The kinematics and kinetics of the awns were then characterized experimentally. In particular, after hydration, the torsion of the base beam and the flexion of the tail for both natural seeds and biohybrid samples (Fig. 4(c)) were recorded, with a slightly different torsion velocity between the two base rods which is consistent with the previous observations. For both natural seeds and biohybrid samples, the tails exerted a contact force (on a load cell) on the order of 1 mN (Fig. 4(d)), with a maximum torque on the order of 100 mN* mm (Fig. 4(e)). The maximum average torque exerted by the awns in the natural Avena seeds was 81.9+29 mN* mm, which is comparable to the forces obtained with the biohybrid samples, for which the maximum torque was 100.4+16.6 mN*mm (Fig. 4(e)). It is noted that the measured torque is approximately five times greater than that previously recorded for other species, such as Erodium and Pelargonium. Considering that torque is essential to increase the possibility of successfully penetrating into the soil, these results highlight the potential of the proposed Avena-based biohybrid device to interact effectively with the unstructured environments of the real world and achieve the sought-after functions. Finally, in order to quantitatively study the dynamics of the awns, the recorded base beam torsion velocities were used as input for the mathematical model described above (Fig. 4(f)). The model provides for a contact force between the awns on the order of 1 mN, in good agreement with the experiments carried out with the load cells (also considering that the contact between the awns is more yielding, and generally more localized, than that between a single awn and the rigid plane of the load cell). Releasing a payload

[0058] The artificial flour capsule of the samples made was functionalized with a fertilizer, such as biochar. Furthermore, tomato or chicory seeds were inserted into the capsule of the various samples. After 18-20 days, 50% of the tomato seeds germinated in a growth chamber and the roots that grew from the functionalized capsule were visible under an optical microscope. It has been noted that only one or both tomato seeds may germinate from the same capsule. Moreover, 33.3% of chicory seeds germinated within 15-23 days. The biohybrid device described above has great potential to be used to enrich the ground and promote germination, with potential applications in reforestation and precision agriculture. For example, so-called “seed balls” have already gained increasing interest in Africa, where favorable fertilizers are mixed with soil and seeds to promote germination in infertile soil. The device described above may act as an autonomous technology similar to a seed ball to promote germination. Furthermore, it may be used to plant seeds to help bees, promoting the germination of favorable flowers used by bees for pollination.

Claims

CLAIMS1. Biohybrid device configured to self-bury into a substrate for transporting a payload, the device comprising an artificial capsule (10) composed of an elongated, pointed body of biodegradable material, said artificial capsule being configured for a payload to be attached or incorporated thereto, and at least two humidity-driven motors (21, 22), each of which comprises a base beam (21a, 22a) connected to the artificial capsule (10) and capable of twisting in response to a change in humidity, and a tail (21b, 22b) connected to the base beam (21a, 22a) through a knee point (21c, 22c) and capable of bending, wherein said at least two humidity-driven motors are composed of respective awns of oat seeds attached to the artificial capsule (10) and capable of elastically engaging with each other.

2. Biohybrid device according to claim 1, wherein one end of the awns of oat seeds is incorporated into the artificial capsule (10).

3. Biohybrid device according to claim 1 or 2, comprising a plurality of natural hairs (10b) attached to the artificial capsule (10).

4. Biohybrid device according to claim 3, wherein said natural hairs are hairs of oat seeds.

5. Biohybrid device according to any of the preceding claims, wherein the body of the artificial capsule (10) is shaped as the head of an oat seed.

6. Biohybrid device according to any of the preceding claims, wherein said biodegradable material is selected from the group consisting of organic materials and compostable resins, thermoplastic materials, water dissolvable materials.

7. Biohybrid device according to any of the preceding claims, wherein said payload is selected from the group consisting of plant seeds, fruits, nutrients, fertilizers, pesticides, spores, sensors.

8. Process for fabricating a biohybrid device according to any of the preceding claims, comprising: collecting said awns (21, 22) from oat seeds, and placing said awns into a mold (Ml, M2) and arranging said biodegradable material into the mold (Ml, M2) to produce the artificial capsule (10) with the awns (21, 22) attached thereto.

9. Process according to claim 8, wherein said mold is shaped to reproduce the shape of the head of an oat seed.

10. Process according to claim 8 or 9, wherein said payload is mixed with said biodegradable material into the mold (Ml, M2).

Images