Device and method to simulate microgravity

Abstract

Disclosed is a device (100) to simulate microgravity, the device comprises a cube-like outer frame (104) formed of a first closing frame (108a), a second closing frame (108b) and side frames (110a, 110b); an inner rotating frame (102) within the cube-like outer frame; a power and data transferring means; a sample holder (106) fitted in the inner rotating frame; and a sample holder rotating means (204), wherein the device has adaptable physical dimensions, featuring an outer frame in a range of 400x400x400-800x800x800 mm, and an inner sample area that provides an extensive volume in a range of 180x80x100-360x360x200 mm.

Description

[0001]DEVICE AND METHOD TO SIMULATE MICROGRAVITY TECHNICAL FIELD The present disclosure relates to devices to simulate microgravity. The present disclosure relates to methods to simulate microgravity using said devices. BACKGROUND Microgravity, often colloquially referred to as weightlessness or zero gravity, is the condition in which objects appear to be in a state of near- weightlessness. Conducting experiments in microgravity offers several advantages that are not possible or are difficult to achieve on Earth. These advantages contribute to advancements in scientific knowledge, technology development, and various fields of research. Specifically, microgravity arises as a family of conditions that are defined by various types of mechanical unloading. In particular, different types of movements (such as rotations), driven by mechanical control algorithmsare designed to produce specific motion, that can create variousmicrogravity conditions on a sample. Experiments conducted in microgravity have led to the discovery of potential drug targets for various diseases. For example, a drug thatpromotes bone formation, developed by understanding biologicalfunctions in microgravity, demonstrated how NELL-like molecule-1(NELL-1) induces bone formation in osteoporosis models and spaceflight conditions. This drug is in clinical development to address bone loss with implications for osteoporosis treatment on Earth. Additionally, microgravity has provided new insights into manufacturing of monoclonal antibodies, such as Pembrolizumab (i.e., Keytruda), an immunotherapy drug for cancer. Evidently, disease research and drug testing in microgravity accelerates discovery of new targets for drugs or drug resistance development pathways. Microgravity also provides a biomaterial production platform, as biological or other material growth or expansion is not affected by gravitational pulling forces, effectively enabling higher purity and more prominent roundness. This is particularly important in stem cell generation or organoid generation, where more volume and a more physiological structure are desired. However, experiments in space are constrained due to limited opportunities and resources. As a result, two-dimensional (2D) clinostats and three-dimensional (3D) clinostats, sometimes referred to as random positioning machines (RPMs), have been designed to simulate microgravity or near-zero gravity conditions. With the 2D clinostat, an object rotates around a fixed single axis perpendicular to gravitational force. The radial acceleration due to this rotation, for a point p on the sample, is directly proportional to a distance from a center of rotation and a square of angular velocity. This radial acceleration imposes practical constraints on the sample size (as points farther from the origin experience greater acceleration) and on the angular velocity (as excessive speeds can cause the machine to act as a centrifuge, generating high-g rather than low-g unloading). On the other hand, the RPM makes an object rotate around two separate axes, creating a "vectoraveraged gravity" environment. The 3D clinostat in static mode, rotatesobject around two axes but without changing the rotational directions.Microgravity-like and simulated microgravity-like (e.g., partial-g) conditions allow for new discoveries and for certain biomaterial production. However, running these experiments is costly and optimization can take years due to trial-and-error approach using internally developed ad-hoc industrial machinery in order to find optimal conditions for production. Importantly, present-day optimizations are done to find homogeneously distributed simulated microgravity paths considering the rotational configuration, but not providing a platform that instantly finds the best paths for the high variety of biological samples used, resulting in high inconsistency in experimental findings and reproducibility. SUMMARY The aim of the present disclosure is to provide a device and constraint- optimized methods to simulate microgravity in its manifold forms. The aim of the present disclosure is achieved by a device and constraint- optimized methods to simulate various microgravity conditions as defined in the appended independent claims to which reference is made to. Advantageous features are set out in the appended dependent claims. Additional aspects, advantages, features and objects of the present disclosure would be made apparent from the drawings and the detailed description of the illustrative embodiments construed in conjunction with the appended claims that follow. BRIEF DESCRIPTION OF THE DRAWINGS Embodiments of the present disclosure will now be described, by way of example only, with reference to the following diagrams wherein: FIG. 1 illustrates a device for simulating microgravity, in accordance with an embodiment of the present disclosure; FIG. 2a, FIG. 2b and FIG. 2c illustrate side views of a device, in accordance with an embodiment of the present disclosure; and FIG. 3 illustrates a method of training models for rotational constraints to achieve certain biological state, in accordance with an embodiment of the present disclosure; FIGs. 4A and 4B collectively illustrate a difference between unbiased and pole-biased random sampling for a device, in accordance with an embodiment of the present disclosure; FIG. 5 illustrates a moving frame of a sample tray, in accordance with an embodiment of the present disclosure; FIG. 6 illustrates a spacelike coverage condition of the device of FIG. 1, in accordance with an embodiment of the present disclosure; and FIG. 7 illustrates a flowchart depicting steps of a method to simulate microgravity, in accordance with an embodiment of the present disclosure. DETAILED DESCRIPTION OF EMBODIMENTS The following detailed description illustrates embodiments of the present disclosure and ways in which they can be implemented. Although some modes of carrying out the present disclosure have been disclosed, those skilled in the art would recognize that other embodiments for carrying out or practising the present disclosure are also possible. In a first aspect, an embodiment of the present disclosure provides a device to simulate microgravity, wherein the device comprises: a cube-like outer frame formed of a first closing frame, a second closing frame and side frames;- an inner rotating frame within the cube-like outer frame;- a power and data transferring means;- a sample holder fitted in the inner frame; and- a sample holder rotating means,wherein the device has adaptable physical dimensions, featuring an outer frame in a range of 400x400x400-800x800x800 mm, and an inner sample area that provides an extensive volume in a range of 180x80x100-360x360x200 mm. A device (namely, a machine) to stimulate microgravity according to the present disclosure is a laboratory machine having a modular cube like setup, which allows the devices to be racked on top of or against each other. For example, the top part of the device has four grooves to smoothly attach next machine. This allows direct serial connection of multiple machines vertically in “tower-like” fashion, and also several towers can be placed next to each other. Active vibration sensing and reduction elements can be further implemented in at least one of: the cube-like outer frame, the second closing frame, the side frames. This enables parallel multiplexed experiments in a space saving way. Also, easier to design control experiments with a gradient of different conditions. The device can deliver controlled mechanical unloading and stressing conditions for experiments. Mechanical unloading is a primary use case of such devices, though it may not directly replicate all aspects of microgravity. Furthermore, the mechanical unloading is also used to induce mechanical stress, with higher speeds. Thus, the device enables a wide range of discovery experiments related to the aforementioned explanation. The power and data transferring means are structured to convey power and data to the inner rotating frame without creating interference. The power and data transferring means enable all kinds of electronic equipment to be placed at a sample, including liquid exchange, sensors, microscopes, irradiator i.e., all types of modules that are frequently used for research. The power and data transferring means could be interference-free shielded slip rings, more specifically, sample holder electronics delivery slip-ring and an inner rotating frame electronics delivery slip ring. Alternatively, the power and data can be delivered wirelessly. The interference-free shielded slip rings allow a much higher bandwidth as general slip rings known in the prior art. Advantageously, there is continuous monitoring and modification of the sample for research purposes with a wide range of devices. The device does not require any battery. The interference shielding solves a problem created by motor driving and delivering data in other scenarios. Optionally, the sample holder comprises a space for electronic components connected to the device through the cube-like outer frame. In this regard, the sample holder is designed with a dedicated space for the electronic components, which are connected to the device throughthe cube-like outer frame. Such configuration integrates electroniccomponents directly into the sample holder. This integration simplifiesthe device design, reduces wiring complexity, and allows for more compact and efficient electronics placement. A technical effect of the aforementioned feature is that it enhances the device's modularity, ease of assembly, and overall performance. The sample holder rotating means could be a motor system comprising an inner rotating frame driver motor unit and an cube-like outer frame driver motor unit or air or liquid pressure inlets. Such design of the device enables to simulate a broad spectrum of mechanical unloading scenarios including, but not limited to, microgravity environments, through the implementation of a control algorithm that optimizes rotational paths in pursuit of the desired conditions, striving to prevent the generation of forces that might cause unwanted biological effects. As an exemplary scenario, researchers might want to do mechanical unloading profiles when examining for example, bone and muscle cells or solid tumors to see tumor drug efficiency changes in different scenarios. In this regard, required conditions are determined at a site of the sample, or shear stresses resembling a tumor microenvironment pushed by the tumor, without adding in additional artifacts by too strong rotations or external forces. As an example, the required condition may be 0.3g perceived gravity Additionally, orientation constraints beyond the "sample normal" vector of each voxel within the sample may be applied. The device may have adaptable physical dimensions, featuring an outer frame in a range of 400x400x400-800x800x800 mm, and an inner sample area that provides an extensive volume in a range of 180x80x100-360x360x200 mm. This enables to fit the devices to most standard incubators and to customize said devices to support specific experimental apparatus and space requirements. The physical setup also makes it very easy to carry and to tower. Optionally, the sample holder comprises one or more sensing means. The one or more sensing means are important to further refine microgravity simulation quality and reduce the number of artifacts. Examples of the one or more sensing means may include, but are not limited to, a shear stress sensor or a force sensor, a positional sensor, a piezoelectric sensor, a device capable of delivering rapid anti-vibrational movements. In an implementation, when the one or more sensing means are implemented as a shear stress sensor, it enables real-time feedback for possible shear forces on the sample, to correct settings and rotations right away when they are too harsh. Optionally, the device further comprises an anti-vibrational element, wherein the anti-vibrational element reduces outside disturbances to the sample. In this regard, the anti-vibrational element acts as a barrier to external vibrations, reducing their impact on the sample. By minimizing external disturbances, the device ensures a stable and controlled environment for the sample, improving the accuracy and reliability of experimental results. A technical effect of anti-vibrational element is that it enhances the device's stability and precision, leading to improved experimental outcomes. The overall benefit is both precision and higher quality when mechanically unloading a sample (via both vector-averaging and directional coverage) without unwanted artefacts, e.g., jerking, jolting, low entropy paths and / or periodic or homoclinic behavior. Moreover, a sensor setup of the one or more sensing means are arranged to detect and report high-fidelity data on motion vectors, acceleration patterns, and optionally thermal fluctuations, offering the option for customization to incorporate additional sensor types as needed for specialized experimental demands. Optionally, the sample holder further comprises a well plate clip holder. The well plate clip holder can hold well plates, wherein the well plates are used to hold multiple samples in individual wells. This is achieved by a clip holder, that fits well plates lying in a range of 6 well plates (w) to 384w, and any other vessel within similar dimensions. The sample holder is thus configured to be universally compatible with a diversity of standard laboratory assay vessels, ensuring that a wide array of assay formats can be securely placed within the machine's inner rotating frame. This obviates a need of screwing or replacing the mount for samplecontainer, as in many cases of existing RPMs and custom-builtapparatuses. A technical effect of the well plate clip holder is that a wide range of well plate formats and custom vessels ae accommodated, which simplifies operation through easy loading and unloading, and improves efficiency by enabling rapid processing of multiple samples. Optionally, the device further comprises a communication means. The communication means enables the device to be capable of interfacing with both standard and advanced communication and power delivery protocols. Moreover, the communication means explicitly covers a voltage range from 3 volts (V) to 24V. Furthermore, the communication means incorporates interfaces, for example, such as Universal Serial Bus (USB), Inter-Integrated Circuit (I2C), serial bus connectivity, Ethernet, Bluetooth®, 802.11-type wireless networking, and similar wired and wireless networking, for enhanced data exchange and control. A technical effect of the communication means is that it enables versatile interfacing with various devices and protocols, thus supporting a wide voltage range and incorporating diverse connectivity options for efficient data exchange and control. Optionally, the device further comprises at least one drive, wherein the at least one drive for at least one of: the first closing frame, the second closing frame, the side frames, are operated by an ultrasonic piezo actuator. In particular, this encompasses a configuration where the ultrasonic piezo actuator is designed to be essentially non-magnetic. Herein, utilization of the ultrasonic piezo actuator, within a predefined performance range that is necessary to drive the frames as described, effectively eliminates mechanical vibration. This is achieved as the actuators operate at frequencies exceeding 80 kHz, which translate into mechanical energy without emitting perceptible vibrations. Currently, no existing microgravity simulators, clinostats, or similar devices for tissue engineering employ drive mechanisms that result in near-zero mechanical vibration or interference attributable to the drives. Optionally, the predefined performance range is 60 kiloHertz (kHz) to 200 kHz. The predefined performance range may, for example, lie in a range of 60, 62, 65, 70, 80, 95, 115, 145, or 195 kHz to 66, 100, 130, 160, 180, 190, 196, 198, or 200 kHz. Furthermore, these ultrasonic piezo actuators can be engineered to be entirely non-magnetic. This attribute is of significant industrial and research importance, as it enables the isolation of biological and inorganic phenomena influenced by microgravity from those influenced by magnetic fields. Traditional motors used in clinostat systems often emit magnetic fields, which can lead to artefacts that affect biological systems. In the exemplary scenario, in a field of cell manufacturing, where specific rotational paths may be employed to derive cells with different shapes or to cultivate stem cells, residual vibration from traditional drives may introduce artefacts. Thus, achieving a completely interference-free environment could be critical for the precise manufacturing of cells. The implementation of ultrasonic piezo actuators not only meets this requirement but also enhances a reliability and integrity of experimental and manufacturing processes. Examples of such drives may include, but are not limited to, a pneumatic drive, a hydraulic drive, a purely mechanical drive, a shielded magnetic drive, an electromagnetic drive. Herein, when the at least one drive is implemented as the electromagnetic drive, the device operation is smoothest when compared to conventional drives. A technical effect of using the ultrasonic piezo actuators as drives is that it eliminates mechanical vibration, thus enabling precise control of device motion without introducing artifacts. This approach provides a superior solution for microgravity simulation and tissue engineering applications, surpassing the limitations of traditional motor-based systems. In a second aspect, an embodiment of the present disclosure provides a method to simulate microgravity using the device according to the first aspect, wherein the method comprises:- mounting an accelerometer to the sample holder of the device;- collecting acceleration data by setting the accelerometer to recordreadouts for XYZ vectors at predetermined intervals continuously over a predetermined period;- calculating moving average of the XYZ vector readouts for eachpredetermined interval throughout the predetermined period;- summing up all the moving averages of the XYZ vectors recordedduring the predetermined period;- dividing the summed average by the normalized acceleration value(go) of 9.80665 m / s² to adjust for standard gravitational acceleration;- dividing the summed moving averages of XYZ vectors by go tocalculate the time-averaged perceived gravitational acceleration (g) at the sensor site over the summed average period; and- using the calculated time-averaged perceived gravitationalacceleration (g) to adjust the device's rotational path to simulate microgravity. Herein, the accelerometer is physically attached to the sample holder of the device to ensure it accurately measures the accelerations experienced by the sample during operation. This ensures that that the accelerometer is properly positioned to capture real-time acceleration data corresponding to the movements and rotations of the device. Theaccelerometer is configured to record acceleration values (i.e., readouts)along the X, Y, and Z axes (namely, XYZ vectors) at the predeterminedinterval over the predetermined period. Optionally, the predeterminedinterval lies in a range of 0.001 second to 60 seconds. The predeterminedinterval may, for example, lies in a range of 0.001 second, 0.01 second,0.1 second, 0.5 second, 1 second, 2 seconds, 5 seconds, 10 seconds, 20seconds, 40 seconds, or 50 seconds to 0.005 second, 0.05 second, 0.5second, 5 seconds, 10 seconds, 20 seconds, 40 seconds, 50 seconds, 55 seconds, 58 seconds, 59 seconds, 60 seconds. Moreover, the predetermined period, for example, may be, 1 minute, 15 minutes, 60 minutes, 2 hours, 6 hours, 24 hours, 2 days, 7 days, 2 weeks, and so forth. Such recording of the acceleration values generates a continuous data stream of acceleration readings that describe the device's motion and orientation throughout the predefined period. This allows for detailed analysis and enables the derivation of averaged acceleration values necessary for simulating microgravity conditions. Subsequently, the acceleration values are processed to calculate a moving average for the X, Y, and Z components over each interval. In this regard, computing moving averages reduces noise in the data and ensures stability in the calculations used to simulate microgravity. Thereafter, the moving averages for each XYZ component are summed over the entire duration of the predetermined period. Such summing of the moving averages provides a basis for normalizing the data and calculating the perceived gravitational acceleration. Subsequently, the aggregated moving averages are divided by go, i.e., 9.80665 m / s2, wherein 9.80665 m / s2is a standard value for Earth'sgravitational acceleration as defined by the General Conference onWeights and Measures. Such division is performed to normalize the data to account for gravitational effects, converting the acceleration values into a relative scale. Herein, normalizing the data ensures that the calculations are consistent with the standard gravitational framework, facilitating accurate microgravity simulation. The normalized summedaverages are divided again by go to compute the time-averagedperceived gravitational acceleration, which provides a single value representing the effective gravitational force experienced at the sensor site over the specified period. Such calculation enables precise adjustments to the device's movement for simulating microgravity. Thecomputed value of ^ is used as an input to adjust the rotational dynamicsof the device, modifying its motion to achieve conditions mimicking microgravity. This ensures that the device operates in a way that replicates the reduced gravitational environment required for experiments. This helps achieve accurate microgravity simulation. It will be appreciated that a Random positioning machine (RPM) is favored in recent research because it offers a precise gravity control at a slow rotation speed than a two-dimensional (2D) clinostat, while ensuring comprehensive simulation of microgravitational effect on the samples. Herein, the RPM is considered a type of three-dimensional (3D) clinostat. However, not all 3D clinostats is necessarily RPMs. In 2D clinostats, artefacts are introduced when pulling vector towards axes of a motor which is not rotating. Herein, the axes of the motor may be u and v. This can lead to less genuine representation of averaging out of the vectors. The RPMs recapitulate microgravity better, as these RPMs continuously alter gravity vector and the object's relative position, especially when rotating faster than the duration the object perceives gravity. To prevent strong centrifugal forces, the RPM turns at a minimal angular speed, ensuring that the objects at its center remain mostly stationary. Although the RPMs cannot fully replicate space conditions, their simulated microgravity is beneficial for studies in fields like cancer research, stem cell treatments, tissue development, drug target identification and regenerative medicine. For sake of brevity, hereinafter the term "axes of the motor" is used interchangeably with the term "motor axes". In addition, use of 3D ground-based microgravity facilities offer an advantage over the use of traditional vessels and flasks, where complex fluid motions and extended culture areas could limit the availability of high quality observations of biological processes altered exclusively by microgravity effects. Notably, the present disclosure seeks to provide a device adept at simulating not just microgravity conditions but also a multitude of customized mechanical unloading within a research setting. The machine adjusts these conditions with unparalleled precision, ranging from nearly zero gravity (0.001 g) to fractional gravity settings (up to 0.9 g), and extends to any designated mechanical unloading and stressing profile. The cornerstone of this machine is its control algorithm, which is based on a mathematical model. This mathematical model and the control algorithm are tailored to ensure that operational paths of the device closely align with the requisite unloading and stress conditions, while minimizing unintended forces that could lead to adverse biological responses. Moreover, the control algorithm's desired output can be refined by additional fluid simulations, biological knowledge, antivibrational measurements and direct sensing of forces from the sample holder. The device exhibits a dual-frame structure as the cube-like outer frame is formed of the first closing frame, the second closing frame, and the side frames. The first closing frame and the second closing frame are opposite to each other. The first closing frame and the second closing frame are connected to each other via the side frames, wherein the side frames connect at each corner of the first closing frame and the second closing frame. The inner rotating frame that houses the sample holder is engineered to accommodate a wide range of standard laboratory assay vessels. The dual-frame structure ensures scalability, for example in the use of remote controlled laboratories. Multiple devices can be stacked on top of each other or next to each other in a tower-like, modular fashion to run up to hundreds of experiments in parallel. Moreover, power anddata are seamlessly transferred through data and power slip rings. Inparticular, data and power slip rings that are shielded and interference free, transmit data and power between the cube-like outer frame and theinner rotating frame. This helps prevent signal interference, therebymaintaining integrity in a sensitive experimental environment. The integrated sensor array allows for precise measurement of physical parameters, and the device comes with customizable communication and power options to fit diverse research protocols. Herein, the integrated sensor array comprises different sensors of physical parameters that contribute to monitoring and controlling of the device. Optionally, the device is not using motors to rotate the sample. In this scenario, the sample is a biological sample that is embedded into hydrogel mesh or other gelatinous material in liquid, and rotated in the center of a liquid container by air or liquid pressure inlets. An advantage of this is that the sample inside the hydrogel mesh will not encounter any shear stress from moving parts of the device, vibrational elements areminimal and rotation is smooth. An amount and frequency of inletpressure of the air or liquid pressure inlets to rotate could be controlled by the same mathematical model and / or the control algorithm that is controlling a motorized version of the device, as essentially, desired and undesired biomechanical forces inside the biological sample are modelled. It will be appreciated that there is a methodology of orienting an object with a fixed center. Herein, the object is, for example, a disc. The disc is oriented in such a manner so that samples placed on the disc experiencemicrogravity uses averaging operators ^^[−] applied to ^(⋅) (a vectorialpath in a configuration space) to define various forms of artificialmicrogravity. In this regard, the device is mathematically modelled withina 3D space in an (^, ^, ^) orientation. In a standard position, the devicehas a reference frame defined as ^^ = ^^, ^ = ^^, ^ = ^^^, where thecondition ^ = ^ × ^. Is always satisfied. This ensures that the normalvector of the sample holder is consistently a cross product of the motoraxes (i.e., ^, ^). Moreover, the motor axes move during device operation.Herein, basic and standard notions from mathematical mechanics, functional analysis, and probability theory are used to define various forms of artificial microgravity. In particular, notions of manifolds ^, tangent spaces ^^(^), Lie groups (namely, ^^(3)) and Lie algebras (^^(3)), orbits of points under group actions, probability theory, abstract vectorvalued integration, and integral kernels ^(−) are used explicitly.A non-standard notation that is important in this context are rotationmatrices ^^(^) ∈ ^^(3), which rotate (in the 3D space) ^ radians aboutvector ^. Typically, ^ is normalised, i.e., ^ ^ ^ = ^ / ||^|| to be of unitlength. A concept of amenable rotations, which is closely related todynamics of the device is defined as follows: A rotation ^ in ^^(3) isconsidered to be amenable if expressed by an exemplary expression (1), ^^^(^^)^^^(^^) (1) It is observed that for the expression (1), following commutation relation holds, as given by an exemplary equation (2) By unpacking the aforementioned definition, it is clear that these rotations, referred to as subordinate compositions, correspond precisely to those rotations that the device enacts for any path originating from standard position (for example, such as an upright sample tray) at (0,0) in angular coordinates on ^^) and terminating at any arbitrary point(^^, ^^). In summary, the amenable rotations are those rotations that areachievable by the device. Herein, the control algorithm that is described is based upon a gimbal- like mechanical description of configurations of the sample, which establishes a bidirectional relationship between a state of the motors and the orientation of the sample in the 3D space. Moreover, this mapping is smooth in a sense of differential topology, allowing paths in one space to be translated into corresponding paths in the other. This concept is formalized in a remainder of this section. In this regard, following spaces are used:- The torus ^^ ≅ ℝ / ℤ^ with angular coordinates as given by anexemplary notation (3), - The Amenable rotations are defined by an exemplary notation (4), - Configurations as given by an exemplary notation (5),Herein, points ^^, ^^ of ^ are occasionally referred to as "frames" and areinterpreted as tuples representing "coordinates" on ^^.It will be appreciated that the configurations, i.e., frames (namely, two vectors recording orientation), are used rather than a sample tray's normal vector ^. This choice is due to a presence of an additional hidden degree of freedom that arises, for example, when the sample tray's normal vector aligns with ^^. If, in this state, the cube-like outer frame undergoes rotation, then the normal vector remains stationary, yet the configuration of the device evolves in a non-trivial manner over in time. Herein, the mapping is provided by an F-map, wherein the F-mapcomprises internal state computations of the device that rely ondiffeomorphisms between sets{motor states} ↔ { amenable rotations} ↔ {configurations }.Symbolically, this relationship is expressed by an exemplary equation (6), wherein the maps ^ and ^ are defined in a direction from left to right. Acomposition of these maps is written using an exemplary notation (7), ^= ^ ∘ ^: ^^ → ^ (7)In this regard, element-wise formulas, i.e., definition of the maps, areprovided. The map ^: ^^ → ^^(3) sends a motor state (^^) anotherstate, wherein the another state is given by an exemplary equation (8), This computation can be carried out using a matrix exponential of infinitesimal generators in ^^(3), or alternatively, by a quaternionic method. Details on explicit computations are provided in detail below.When an action of F on a standard frame ^^^, ^^^ is considered, a relationis given by an exemplary equation (9),^(^^, ^^) ⋅ ^^^, ^^^ = ^^^^ (^^)^^^(^^)^^, ^^^^(^^)^^(^^)^^^(^^)^^^ (9) A simplified version of the equation (9) is given by an exemplary equation (10): By construction, the vectors u and v satisfy ^ ⊥ ^^, ^ ⊥ ^, ensuring thatfollowing exemplary equation (11) is satisfied, ^(^) ∈ ^ ⊆ ^^ × ^^ (11)Moreover, using a standard inner product, following exemplary equation (12) is obtained, and similarly following exemplary equation (13) is obtained, Thus ^(^) lies in ^ ⊆ ^^ × ^^, as required.An auxiliary map of importance, ^: ^ ⊆ ^^ × ^^ → ^^, is defined using a 3Dcross product ×. This map serves as a crucial tool for further analysis of microgravity and reduced gravity conditions. Such analysis is provided under the heading "MICROGRAVITY AND REDUCED GRAVITY CONDITIONS". MICROGRAVITY AND REDUCED GRAVITY CONDITIONS Before providing precise definitions, it is important to outline the fundamental aspects of microgravity-like conditions, which are characterized by two key properties of paths. The first property is averaging, wherein the mechanical unloading of forces results in an average force of zero, either within a finite time window or asymptoticallyover time. The second property is coverage, wherein mechanical unloading of forces occurs in such a way that no particular direction is favored. The second property is especially significant because, without it, certain paths, such as those where the sample tray's normal vector continuously loops in a circle, would incorrectly be classified as "microgravity-like." While such paths are clinostatic and achievable using a single motor without advanced control algorithms, they do not encapsulate the essential features of microgravity-like conditions. The primary focus here is to present an overview of the broad spectrum of conditions related to gravity, rather than to define a narrowly specified family. Detailed considerations regarding measurable and integrable function spaces are intentionally omitted, as they would detract from the essential points of interest. Moreover, pathological curiosities do not arise in reality. To formalize these ideas, each causal integral kernel ^ (supp ^ ⊆ [0, +∞)could be convolved with a vectorial path ^: [0, +∞) → ^ in some vectorspace ^. (Of particular interest are paths on the sphere ^^ ⊆ ℝ^, wherethe convolution is defined as given by an exemplary equation (14), ^ A standard notion of average may be expressed as ^^ ∗ ^(^), where thekernel ^ is the Heaviside step function ^, defined as ^ = ^[^,^^).Examples of such kernels include simple moving average (SMA) and exponential moving average (EMA). For SMA of window size ^, the kernel is ^ = ^[^,^], which evaluates to 1 on [0,1] and 0 otherwise. For EMA withdamping parameter ^, the kernel is given by an exemplary equation (15), These concepts of the SMA and the EMA generalize to cases where integral kernels depend on two variables, reducing to the single variable case by considering a difference of their arguments: ^^(^, ^) = ^^(^ − ^).The average in this manner can be written as an exemplaryequation (16),wherein the kernel A^-machine is defined as a (possibly stateful) stochastic process ^(^)which continuously outputs a path in real inner-product space ^ ≅ ℝ^. Inother words, this is a ^-valued continuous time stochastic process, where values at future times are correlated to values in the previous times, for example, via integral operators. When ^ ↪ ^ is an embedded compactsubmanifold (with pullback metric), there is an obvious notion of ^- machines. The following conditions are considered for any realization of the process ^(^), rather than, for example, its expectation ^^(^) = ^[^(^)]).Herein, there are two averaging conditions, namely, a first averaging condition and a second averaging condition. The first averaging condition comprises a process called weakly ^-null, given by an exemplary equation (17), The second averaging condition is called strongly ^-null, given by an exemplary equation (18), |^ ∗ ^(^)| ≤ ^ for some bound ^ > 0 for all ^ ≫ 0 (18)Notably, strong nullity implies weak nullity but not vice versa. For example, ^^[^] may grow on the order of √^, in which case weak nullityholds, but strong nullity does not For compact Riemannian manifolds, following conditions are phrased in terms of uniform probability measures. The uniform distribution on such compact Riemannian manifolds is proportional to standard volume measure constructed from the metric. The ^-machine {^(^)}^^^is said to eventually converge to a probability distribution ^ if given access to anyfixed prior information ^ about the state(s) of the process (for example,such as bounded by time ^), the conditioned variables ^^|^ tend to towardbeing ^-distributed, i.e., converge in distribution to ^.For what follows processes ^(^) = ^^(^), ^(^)^ take values in ^, meaningthey are frame-valued. From such processes, normal process can be constructed as given by following exemplary equation (19), ^(^) : = ^(^) × ^(^) =: ^(^(^)) (19)where ^ is a cross-product map.To express coverage conditions clearly, mechanical unloading conditionson random variables ^^, ^^ in ^ are introduced instead of on at least oneprocess:- naive RPM uniformity: ^^^(^, ^) is uniformly distributed on ^^ =^^ × ^^, representing uniformly distributed random motor positions.- Normal uniformity: ^ = ^(^, ^) is uniformly distributed on ^^ where^ =×: ^ → ^^- Spacelike uniformity: ^ = ^(^, ^) is distributed uniformly on ^^along with a subordinate condition that, given ^ = ^ ∈ ^^^^(^^) (theprojection of ^ onto ^^), conditioned variables ^^, ^^|^ are distributeduniformly It is nontrivial that these conditions for abstract machines emitting paths are independent, i.e., either one condition can be satisfied regardless of the other. Finally, coverage conditions for the processes are defined as follows:- for normal coverage: eventually becomes normally uniform.- for spacelike coverage: eventually becomes spacelikeuniform, i.e., ^^becomes uniformly distributed on ^^.)For cases of partial gravity, say ^^-partial gravity with ^ ∈ (0,1), the firstaveraging condition given by equation (17) is modified to following exemplary equation (20) Moreover, the spacelike coverage condition is challenging to represent visually due to the six-dimensional nature of the problem. As the normalvector ^ moves closer to alignment with ^^, a number of feasible motorstates increases. It is important to note that not all abstract pathmachines will generate paths such that the frames ^^ ∣ ^^, where ^ =^ × ^, are equally probable. However, the claim is that the currentmachine is equipped with controls that effectively enable this novel statistical property, ensuring uniformity in the distribution of these frames. It will be appreciated that more refined conditions can be derived from these definitions by incorporating concepts such as convergence rates or window-based constraints. Similarly, more refined conditions can beproduced by requiring certain rates of convergence for any of thecoverage / eventuality conditions. Listed below are two examples, namely a first example and a second example. In both the first example and the second example, it may beassumed that ^ = ^, the Heaviside step function, ^ = ^^ or ^, ^ as theuniform distribution on ^, and ^(^) = ^^(^), ^(^)^ is a ^-machine.As the first example, a realistic and achievable microgravity-like condition may satisfy the second averaging condition (and consequently the first averaging condition), as well as normal and spacelike coverage conditions. This may correspond to the second averaging condition givenby the equation (18), ensuring that (| ∫^ ^^ (^(^)) ^^| ≤ ^ for some smallbound ^ > 0 for all large ^). In this case, the normal vector is eventuallyuniform, and moreover, the orientations are as uniformly distributed as possible. As the second example, the partial gravity may be given by replacing the equation (17) with the equation (20), and replacing the normal coverage^^^ , ^^^ with the distribution ^ that has a mean of −^^ ⋅ ^^ for some ^ ∈(0,1). It should be noted that there exists an infinite-dimensional degree of freedom in choosing the distribution ^. Moreover, on the configuration space ^, time varying) potentials is defined as ^: ^ → ℝ that are constructed at any instant ^ via ^ which maps a given history to a potential, i.e. for some "computation" ^: {histories} → {potentials}, wherein ^ is a path in configuration space.Herein, (−)|[^,^]denotes restriction of an infinite path to a finite history between time zero and time ^. With greater precision, this can be following exemplary equation (21) of the inverse of the map ^^ → ^, i.e., the differential along computes updated motor speeds via tangent vectors to ^.Since ∇^|^ : ^(^) → ^(^^), this provides a mechanism to produce motor update controls via computations based on historical state(s): ^(^^^) ∘∇^|^ : ^ → ^(^^). Methods for implementing this on a microcontroller are detailed below. Intuitively, "hills" in a potential repel machine configuration away from maxima; conversely, while "valleys" attract the dynamical state. This method enables sculpting the evolving machine state to desired outcomes. For instance, attractors set at the standard position cause the machine to spend more time there, thus enacting partial-^ types ofmicrogravity-like conditions. Since the configuration space is a submanifold of ℝ^, Gaussian functions in ^-space may then be restricted to ^, producing approximate Riemannian Gaussian functions on the curved manifold ^ withoutrequiring complexities of performing analysis on curved spaces. An entire state of the machine can be encoded using two angular coordinates ^^, ^ = 1,2, the dynamics are formulated in terms of a classicalfirst order (highly nonlinear) system. Writing ^ = (^^) = (^^(^)) with dotsfor denoting time derivatives, the system dynamics can be expressed by the following exemplary equation (22): ^̇ = ℱ(history of ^) (22)The dynamical update ℱ is computed via gradients of potentials. Byexpanding the system via substitution, the explicit dynamical equation is given by the following exemplary equation (23) before ^) providing a formula for the computation of ℱ. The next subsection delves into the computation of this in microcontrollers. A discrete version of this system is given by following exemplary equation (24) step window) (24)EXPLICIT COMPUTATION ON MICROCONTROLLERS It is well known how to compute explicit vectors tangent to the sphere ^^at any given point. Since the configuration space can be realized as a submanifold of ^^, constructing explicit bases for the tangent spaces ofpoints of ^ is straightforward. The forward direction of the map ^: ^^ → ^is computable. Since the potential ^ is defined on ^, tangents to ^ istransferred via the differential of the inverse of the map ^: ^ → ^, ^ ↦ ^^ ⋅^^, ^ ⋅ ^^^. It is important to note here that if at least the forward directionof a diffeomorphism is computable using an efficient algorithm, itbecomes possible to compute the differential ^(^^^) (a Jacobian ^^^^) ofthe inverse map ^^^in explicit coordinates / bases. This involves firstcomputing ^^ in one basis and then inverting resulting square matrix.This approach leverages the fact that the inverse of a smooth homeomorphism is itself smooth.To compute ^ for a particular motor configuration, the explicitconstruction of rotation matrices, as given by exemplary equation (25) can be used: wherein ^^^^is the Levi-Civita symbol. Since matrix exponential can be computed or defined via Taylor series expansion, this allows computation of rotation matrices on the microcontroller with minimal errors arising only from, numerical truncation and / or numerical chaos. Finally, since ^ is constructible from rotation matrices, this approach provides an efficient calculation for ^, and by the above. Furthermore, by computing ^^^^andcomposing it with ^^^^, the map ^ → ^(^^) is obtained.Since the tangent vectors to the torus are essentially the same as motor speeds, it is concluded that this method established an explicit framework for mapping configurations to amenable rotations simply via motor movements. This enables practical implementation of dynamical controls in a standard sense. MACHINE STATE The machine uses a stateful algorithm executed on a microprocessor with a limited memory capacity (for example, such as 4 kilobytes (kB)). This stateful algorithm is designed in such a way so that it does not require powerful or fast processing, or large amounts of memory. A primary use of state is for tracking historical data and / or for storing results of integral operators (computed via, for example, integer accumulator registers). Moreover, storage requirements for historical machine trajectories are minimal. Instead of storing 3D vectors, such as the normal vector of the sample tray normal, it is sufficient to store only the motor positions(^^, ^^). Alternatively, discrete differentials (^^̇, ^^̇) could be stored, whichgiven same number of bits per element, provide higher fidelity. So, for an entire day of recording path data, an exemplary data usage, given by an exemplary equation (26), may be determined: ^^ ms ^^ ms × 3600000 ms = 11520000 b = 1.4 MB (26)which is, interestingly, approximately equivalent to a capacity of a 1.44 megabyte (MB)) floppy disk. Moreover, the coverage-type conditions require that internal algorithm injects entropy into computation of the next state. Various forms of entropy could be considered for future state variables (that are conditioned on bits of history). This can involve both classical Shannon entropy and computational measures such as HILL entropy, with well- defined constraints. Either is achievable, respectively, via harvesting of true entropy, e.g., from sensor noise (namely, from accelerometers), or from pseudorandom number generators (PRNGs). Herein, the sensor noise plays a critical role in functioning of abstract machines, particularly in scenarios involving microgravity-like conditions or entropy harvesting. Abstract machines often rely on stochastic processes to drive state transitions or generate randomness for trajectory diversification. In this context, the sensor noise, such as that derived from accelerometers, gyroscopes, or magnetometers, serves as a natural source of entropy. By leveraging inherent unpredictability of sensor measurements, the abstract machines can inject randomness into their computations, enabling compliance with conditions like uniform coverage or stochastic state evolution. For example, accelerometer noise, whenappropriately processed, provides a practical approximation to truerandomness, which can be used to seed pseudorandom number generators (PRNGs) or directly influence state updates. From a theoretical perspective, the sensor noise aligns with the concept of entropy in the abstract machines by ensuring that future states are not deterministically linked to past trajectories. This randomness is crucial for meeting coverage-type conditions, as it prevents the system from becoming trapped in repetitive or preferential patterns. Furthermore, noise-induced entropy ensures robustness against modeling imperfections or external disturbances, thereby enhancing the device's ability to explore the configuration space uniformly. The abstract machines that incorporate the sensor noise into their computational framework thus achieve a balance between deterministic control and stochastic adaptability, which is essential for applications in dynamic and uncertain environments.In this regard, orbits of each point ^ in the sample area (or even in theball ^, where ∂^ = ^^) are defined relative to any path ^(^) in motorspace^^. Given any point ^ in the solid ball, as given by following exemplaryequation (27),^ = ^^(0) (∂^ = ^^) (27)and any path ^(⋅) in ^, the orbit trajectory of the point is given by followingexemplary equation (28), The orbit map extends naturally to all ^ ∈ ^^(^) as well. (for example, forframes). If ^ and ^ ∈ ^^(^) are considered, then any matrix ^ ∈ ^^(3)sends this structure to ^ ⋅ ^ and ^ ⋅ ^ ∈ ^^⋅^(^), respectively. Since ^: ^^→^ is defined, this extension applies to paths in the motor space ^^ (in ^^coordinates) as well, associating an initial point ^ in the sample with atrajectory in ^.Now any (pointwise) constraint ^ may be phrased as given by followingexemplary equation (29), ^: ∂^^ ^[^](^) ≤ ℎ(^, ^) (29)for some function ℎ(^, ^), where ℎ represents a time-varying field or profileover the sample region. Typical computational fluid dynamics computermechanics systems allow representation of ℎ via values meshes or grids,etc. Examples are as listed below:- Bounding centripetal force | ∂^^^^[^](^)| ≤ ^,- Non-bounding: non-pointwise constraints are extremely importantin the field of microgravity research. Herein, the main condition is anintegral constraint | ∫^ ^^ (^)up^ ^^| < ^, where up^ ∈ ^^(^) is the upwardnormal vector at the origin (the central normal vector of the sample tray), and- Stochastic conditions may be required if ^ is a randomly selected(non-terminating) path in ^ then {^^[^](^)}^^^ is a random process.Moreover, distributional eventualities for the random variables ^^[^](^) forfixed large ^, etc. may also be considered. Notably, these samples are conceptually aligned with Proportional- Integral-Derivative (PID) controllers, involving proportionality, integrals, and derivative components. This can be phrased as Integro-Differentialoperators acting on the paths ^(^) of points under the action of themachine. TRANSITION FROM CONTINUOUS TO DISCRETE CASES The results are applicable in in both continuous time and continuous manifold coordinates but can be phrased in discrete time. Before dealing with discretization of the underlying manifolds, a method of converting continuous time calculations, in particular for motors, into discrete pulses is provided. The framework for converting between abstract paths and pulse trains is given below.In the framework, the following notations are used: ^^ = ^^^, ^^,^ =[^^^^, ^^] (time window), length ^^ =[^^, ^^^^], length(^^) = ^^, ^ ^^ : = ^^^,^, ^^ ∈ {−1,0, +1} ≈ ^^pulse value at epoch ^, ^^ epoch length (between updates), ^^ step angle,(grid points), and #(^^,^ ∩ ℤ^^) = #{^ − ^, … , ^ − 1, ^} = ^ + 1Moreover, following assumptions are made for the framework: -At each time step ^ (during interval ^^,^) motor is any one of: still,moves forward, moves backward by one step, thus providing a pulse train step, during time length ^^, the angle changes by ^^^^, andsteps, i.e., one ^-window, the mean angular velocity is averages are approximately equal to (given by an exemplary equation (30)): wherein when manipulating the equation (30), the following exemplary equation (31) is derived Thus, following exemplary equation (32) is set Triangulations of 2-dimensional spheres, such as those based on pyramids and icosahedron, can be used to discretize configuration spaces. Since ^ is a subspace of ^^ × ^^, these triangulations can provideinherited coordinates for ^, facilitating computations in discrete settings. It will be appreciated that discrete differential geometry analogues of all of the above mathematics (differentials, tangents, potential, etc.) are all constructible and considered standard techniques in control theory. Moreover, instances of the control algorithm may be provided by anexemplary pseudocode, as provided in the APPENDIX.Optionally, the method further comprises harvesting entropy from sensor data streams to provide element of unpredictability in trajectories. Such element of unpredictability prevents "learning-like" biological processes from leveraging predictability when forming various cellular structuresand tissues, etc. In this regard, randomness is harvested with a goal ofpreventing biological samples from exploiting patterns. Optionally, the method further comprises enacting partial gravity. Optionally, the method further comprises moving at least one of: a biological sample, a chemical sample, to achieve at least one of: morphological properties, crystallographic properties. The at least one of: morphological properties, crystallographic properties are similar to at least one of: morphological properties, crystallographic properties when the at least one of: a biological sample, a chemical sample, is moved when exposed to actual microgravity. In one aspect, an embodiment of the present disclosure, unlike in the state of art, is carefully considering biological and chemical properties and effects during rotational execution, not simply achieving a certain acceleration average. In fact, reaching a certain acceleration average while avoiding unwanted artifacts is the main differentiator. This approach can be simplified and computationally can be made easier by adopting a mathematical model described in the present disclosure. A FIRST EXAMPLE EMBODIMENT Stem cells for regenerative therapy should grow uniformly, resembling a three-dimensional sphere. Before implantation, this structure should be kept. Certain stem cells in microgravity and rotating bioreactors may grow resembling physiological shape, however stem cells grown in standard gravity in cultures are often "flattened out". Even in microgravity and rotating bioreactors and microgravity simulators not all the conditions may be met to fulfil the best growth scenario. Usually, because this device may only consider the rotational parameters. In the present disclosure, rotational paths may be carefully generated with an algorithm considering previous simulation data, path data from previous experiment, biological requirements, modelled state of fluid dynamics in real time to predict and to in-situ correct for the best possible outcome. A SECOND EXAMPLE EMBODIMENT In another use case, user may want to see if a certain drug has less or more effect in mechanical unloading or mechanical stress model in a cell culture. Then, control algorithm may take in account the datasets discussed in the previous use case and able to produce paths of a gradient of mechanical unloading (0-1g) or mechanical stress, hypergravity. Then, researcher analyses the conditions and finds out that in one unloading model the drug was more effective. Optionally, the method further comprises applying a transfer function for correlating one or more settings of the device. Optionally, the method further comprises applying a transfer function for correlating one or more settings of the device with improved biological outcomes. The method enables to optimize biomanufacturing processes. The transfer function is used to establish a relationship between the one or more settings and output, wherein the output could be a biological response. A technicaleffect of applying the transfer function in such a manner is it reducingtrial-and-error, and improving process consistency, efficiency, and product yield and quality. Predicted configuration settings, derived from (paths driven by the rotation and biological setup parameters (for example, such as medium type, vessel type) that has high likelihood of yielding similar results to previously successful experiments. This reduces a need for trial and error. The correlation between one or more settings of the device and improvedbiological outcomes includes, but is not limited to, increased cell viability,enhanced proliferation rates, and superior cellular product quality.Optionally, the method further comprises minimizing cellular andmicrobial aggregation through precise manipulation of the mechanical environment. For example, in certain fermentation processes it is beneficial to constantly rotate the bioreactor, however speed and directions are critical for effect. This reduces product loss and improves yield consistency in biomanufacturing applications.Optionally, the method further comprises providing selective mechanicalstress profiles. Optionally, the method further comprises facilitating the directed evolution or strain generation of cell and microbial lines by providing selective mechanical stress profiles that guide adaptation towards desired phenotypes or genotypes. Certain microbial organisms, such as yeast strains and bacteria can produce products even in reduced growth conditions. This can be confined growth area, radiation, increased or decreased mechanical stress. Databases and the predictive model are used to refine control of the rotation and advise on e.g. biological parameters to gain an adapted yeast for the user’s need. This enables again high-throughput and multiplexed generation of multiple yeast strains that very quickly could be evaluated for continuation or not and enhances the robustness or productivity of these lines. A technical effect of providing selective mechanical stress profiles is that it enables generation of adapted microbial strains by applying selective mechanical stress profiles, in conjunction with database and predictive modeling. This approach accelerates strain development, enhances strain robustness and productivity, and optimizes biomanufacturing processes.Optionally, the method further comprises calibrating rotational paths foroptimizing biochemical reactions and bioprocesses. The algorithm enables fine tuning of the paths and usable volume for the production increase, while reducing unwanted effects. This enables increase efficiency in the production of biological to biopharmaceutical products.Optionally, the method further comprises customizing mechanicalunloading profiles to promote favorable micro-environments for cellular morphology, matrix deposition, crystallisation and enhanced biological function. By customizing mechanical unloading profiles to promote favorable micro-environments for cellular morphology, matrix deposition, crystallisation and enhanced biological function the method enables to refine manufacturing workflows in the production of biomaterials, materials (including crystals) or tissue engineering constructs using the device according to the present disclosure. This can be achieved for real time sensing of growth conditions, either with sensors (pH, micronutrients, direct miniature microscope monitoring) that feedbacks real time information to the algorithm to evaluate the prediction and if needed, correct. For crystallization, long processes (more than a few hours are suitable here). This enables better output and multiplexed high- throughput discovery without wasting samples in trial and error. Optionally, based on the feedback from sensor data, the method comprises adjustments to the control algorithm based on feedback from sensor data, allowing for real-time process optimization that directly translates to improvements in both the structural and functional characteristics of manufactured biomaterials or tissue constructs. This is more specifically related to organoid production. For example, in addition to the control fine tuning methods discussed in the previous features, organoid can be further observed with a miniature microscope / camera to direct control to a favorable growth. From a database, image of correctly grown organoid is inferred and over time is analyzed if the actual sample is growing in the right way. A technical effect of the aforementioned feature is that it enables precise control of device motion by computing trajectories using Lie algebra- based techniques and sensor feedback. This approach allows for dynamic adjustment of device orientation to maintain artificial microgravity conditions, thus optimizing experimental outcomes. Optionally, the method further comprises dynamically modifying paths via deformation of the equations of motion. Dynamically modifying the paths via deformation of the equations of motion can be performed in a manner akin to providing a nonzero potential function so as to cause the paths to have a tunable amount of attraction to spending time in the upright standard position, so the quantity s(T) / T behaves asymptotically like the constant function f(T) = B. A technical effect of the aforementioned feature is that it enables control of device motion by deforming the equations of motion, allowing for tunable attraction to a standard upright position. This approach enhances flexibility and precision in device operation. Optionally, the method further comprises modifying the device behavior so that when a user defined constraint system is used that comprises inequalities involving or closely related to at least one of: sheer Stresses, Momenta, Forces, etc., the device performs any one of: before moving alerts the user that the user defined constraint system is over-constrained and emits an error, begins moving and dynamically computespaths satisfying these user defined constraint system. Herein, the device behavior is modified by dynamically computing paths that satisfy user- defined constraint systems (namely, constraint system) involving stress, moment, and force. The device can either alert the user to over- constrained systems or dynamically adjust its motion to adhere to theconstraints. This feature enhances the device's ability to perform complex tasks in constrained environments, ensuring safe and efficient operation. A technical effect of the aforementioned feature is that it improves the device's adaptability and robustness by enabling it to handle complex, user-defined constraints, optimizing performance and minimizing the risk of errors or damage. Optionally, the method further comprises using sensors of the device as true random number generators and transporting entropy supplied by sensor reading randomness to trajectories. Optionally, the method further comprises using sensors of the device as true random number generators and transporting entropy supplied by sensor reading randomness to trajectories with an intent of avoiding biological samples from "learning" of trajectory timings and shape. Such transportation of the entropy is performed with an intent of avoiding biological samples from learning of trajectory timings and shape. Current random positioning machines are using random generation based on software methods. These can have repeating patterns. In the present disclosure, hardware derived true random generators and entropy harnessing, for example sensor noise, are being used to combine the repeating patterns for true randomness. This true randomness can be utilized for the trajectory generation, inside the constraint system, to have improved random positioning. Thereafter, such improved random positioning will be constrained to fulfill biological needs from constrains. A technical effect of the aforementioned feature is that it improves randomness and security of device operations by incorporating hardware- based true random number generation. This leads to more effective and reliable experimental results. Optionally, the method further comprises consolidating inputs from a multitude of data types and incorporating at least one of: strategies for integrating data derived from fluid dynamics, cell mechanics, and biochemistry; dead reckoning processes for precise motion prediction; an optimization model that precludes the formation of motion paths unfavorable to maintaining ideal unloading states and tissue integrity. Strategies for integrating data derived fluid dynamics could be inferred from actual shear stress sensor, but also modelled on the fly or through predicted modeling of fluid dynamics. If a given cell type would experience overstress within the predicted fluid motions, the computing algorithm adjusts accordingly. This approach reduces trial and error and manual work, improves understanding, and enables high throughput multiplex experiment setups. Moreover, the dead reckoning processes for precise motion prediction can be estimated using current measurements for the motor and closed loop feedback data from circuits of the motors, combined with predicted outcome for the feature described. This enables a sensor-free operation in scenarios where minimal disturbance is desired, for example, such as for electromagnetically sensitive cells, or the simplest possible setup is required. Optionally, the method further comprises fine-tuning adjustments in response to real-time stress and shear force data garnered from at least one sensor of the device, wherein the at least one sensor comprises: shear stress detectors, microfluidic channels, capacitive arrays, piezoelectric transducers. Data considering these detectors are fed into the mathematical model and a corresponding implementation of algorithm that sends go / not go / go back step control to the motors to re- adjust paths. Herein, the real-time sensor data is utilized from the shear stress detectors, the microfluidic channels, the capacitive arrays, and the piezoelectric transducers to fine-tune motion of the device. The sensor data is fed into a mathematical model and / or the control algorithm, which generates control signals to adjust motor behavior and modifytrajectories. This approach enables real-time optimization of devicemotion, ensuring that the device operates within safe and efficient parameters, and minimizing the risk of damage to samples or the device itself. A technical effect of the aforementioned feature is that the precision and reliability of device operation is enhanced by incorporating real-time sensor feedback and adaptive control strategies. This leads to improved experimental outcomes and reduced operational costs. Optionally, the method further comprises dynamically modifying paths via deformation of equations of motion by providing a non-zero potential function to attract or repel a dynamical state of the device to a required configuration. In this regard, intuitively, "hills" in a potential repel machine configuration away from maxima; conversely, while "valleys" attract the dynamical state. This method enables sculpting the evolving machine state to desired outcomes. For instance, attractors set at the standard position cause the machine to spend more time there, thusenacting partial-^ types of microgravity-like conditions. A technical effectof the aforementioned feature is that it enables dynamic control of device motion by deforming the equations of motion through potential functions. This allows for precise manipulation of device trajectories, enabling the creation of custom microgravity-like conditions and other desired configurations. APPENDIX(1) A process may emit either a point in ^ or a null value. Byintroducing perturbations to the potential coming from these emissions, for example, such as time-decaying Gaussian functions, entropy can be injected into the system. The following pseudocode outlines an example implementation: while (1): rotate ring buffer; v<- sample-accelerometer();pt <- [Tess, 0]; / / randomly select point from tessellation / / above operation can be weighted for partial-g use Kalman filtering (OR amplify noise) to output a dead position (using v) d<- evaluate the gradient of the ("multimodal") potentiale <- inv (dF)(d)apply e to motors; Sampling such a random process to perturb the potential can beleveraged to achieve non-standard motions, for example, partial-^ andother microgravity-like conditions by incorporating attractors into specific parts of the configuration space. This is achieved by judicious choice of the distribution algorithm and / or sampling algorithm). DETAILED DESCRIPTION OF THE DRAWINGSReferring to FIG. 1, illustrated is a device 100 for simulatingmicrogravity, in accordance with an embodiment of the presentdisclosure. The device exhibits a dual-frame structure, with an innerrotating frame 102 and a cube-like outer frame 104. The inner rotatingframe 102 houses a sample holder 106 engineered to accommodate awide range of standard laboratory assay vessels. The cube-like outerframe 104 has a first closing frame 108a and a second closing frame108b and two side frames 110a, 110b. The second closing frame 108bhas stand pegs 112 with vibrational reducing material on one side andmatching grooves (not shown) for the stand pegs on the opposite first closing frame 108a. This allows to stack devices on each other or next to each other in a tower like, modular fashion to run up to hundreds of experiments in parallel. Optionally, the device comprises axis in x orientation (depicted as X-axis), y orientation (depicted as Y-axis), and z orientation (depicted as Z-axis). FIG. 1 is merely an example, which should not unduly limit the scope of the claims herein. A person skilled in the art will recognize many variations, alternatives, and modifications of embodiments of the present disclosure. Figures FIG. 2a, FIG. 2b and FIG. 2c illustrate side views of a device, in accordance with an embodiment of the present disclosure. FIG. 2a illustrates a side view of the first closing frame 108a, and the second closing frame 108b. FIG. 2b illustrates a side view of the side frames 110a-b. FIG. 2c illustrates a side view of which shows the connection ofthe side frames 110a-b and the first closing frame 108a, and the secondclosing frame 108b. In FIG. 2a, the inner rotating frame 102 with asample holder 106 is shown. In FIGs. 2a- c, the second closing frame108b has stand pegs 112. In FIG. 2c, matching grooves 202 for thestand pegs 112 are shown on the opposite first closing frame 108a. InFIG. 2c, sample holder rotating means 204 is shown. The sample holderrotating means 204, is for example, depicted as a motor unit. In FIG. 2a,inner rotating frame driver motor units 206 are shown.FIGs. 2A-C are merely examples, which should not unduly limit the scope of the claims herein. A person skilled in the art will recognize many variations, alternatives, and modifications of embodiments of the present disclosure.FIG. 3 illustrates a method 300 of training models for rotationalconstraints to achieve certain biological state, in accordance with anembodiment of the present disclosure. 302 refers to user interface usedfor obtaining information from the user and 304 refers to the experimentinformation from the user. 306 refers to previous and simulated datawhere previous data is reported data from publications exported asconfiguration to a local database and simulated data is on the fly or pre- experiment simulation, generating constraints for the specific sample type, e.g. to do not rotate too fast.Constraint set 318 is generated from the CFD data 308, path data 310,biological data 312, on-the fly simulation data 314 and radiation data316. CFD data 308 refers to sample state and is a modelling module,reporting the fluid / shear forces acting on the sample according to pre-run or on the run simulation. CDF data 308 could be converted tobiologically relevant data. Biological data is published data about what happens to specific cell lines under different shear stresses. Path data310 can be both training data and previous path control / position data tofurther re-fine, learn or to re-run. Biological data 312 is a relevant datafor certain cell types from database, including unwanted and optimaleffects within different settings. On the fly simulation data 314 is a reporton the perceived shear stresses, liquid dynamics or forces acting on thesample. Radiation data 316 is a raw radiation data translated toequivalent irradiation for the radiation module, or radiation dose input by the user. The aforementioned steps are only illustrative and other alternatives can also be provided where one or more steps are added, one or more steps are removed, or one or more steps are provided in a different sequence without departing from the scope of the claims herein. Referring to FIGs. 4A and 4B collectively, illustrated is a difference between unbiased and pole-biased random sampling for a device, in accordance with an embodiment of the present disclosure. In this regard, the device is, for example, a Random Positioning Machine (RPM). In bothFIGs. 4A-B, X axis, Y axis, and Z axis denote spatial coordinates withina normalized unit sphere. Moreover, each point represents a sample position or a configuration. Herein, each point provides X, Y, and Z values collectively to describe an orientation or position in a three-dimensional (3D) space, which is crucial for ensuring uniform coverage or avoiding bias in the RPM's operation. Additionally, both FIGs. 4A-B comprises poles, wherein the poles refer to regions along the Z axis. FIG. 4A is aligned to the left, and FIG. 4B is aligned to the right. This alignment emphasizes A visual and conceptual distinction between unbiased and biased sampling. In FIG. 4A, results of unbiased sampling using a proper algorithm has been depicted. The points are distributed uniformly across the surface of the unit sphere, ensuring equal probability for all orientations. This even distribution demonstrates the effectiveness of the algorithm in eliminating any preferential bias towards specific regions, such as poles. This figure represents an ideal sampling process that adheres to the desired statistical properties for RPM operation. In FIG. 4B, pole biasing is depicted that arises due to improper samplingdirectly from the motor space. Herein, a density of points near the polesis noticeably higher, indicating a non-uniform distribution. This pole biasing occurs when sampling algorithm does not account for the geometric transformation between motor space and 3D space, resulting in overrepresentation near the poles. FIGs. 4A-B are merely examples, which should not unduly limit the scope of the claims herein. A person skilled in the art will recognize many variations, alternatives, and modifications of embodiments of the present disclosure. Referring to FIG. 5, illustrated is a moving frame of a sample tray 502, in accordance with an embodiment of the present disclosure. Herein, thesample tray 502 is abstractly depicted as a three-dimensionally (3D)rotated) rectangle. Herein, axes ^, ^, and ^ are depicted as unit vectorsin three-dimensional space. In this regard, the normal vector ^ iscalculated as a cross product ^ = ^ x ^. Moreover, ^ and ^ represent motoraxes, which are subject to movement during operation of a device. hese axes correspond to specific rotational angles. The outer frame rotation angle, denoted as ^^, determines orientation of ^, while an additional inner frame rotation angle ^^(not shown for sake of brevity) governs deviations from a standard position. In the standard position, the vectorsare aligned as ^ =^1, ^=^2, and ^=^3, thus forming an orthogonal basisin three-dimensional space. FIG. 5 is merely an example, which should not unduly limit the scope of the claims herein. A person skilled in the art will recognize many variations, alternatives, and modifications of embodiments of the present disclosure.Referring to FIG. 6, illustrates a spacelike coverage condition of thedevice 100 (not shown for sake of brevity) of FIG. 1, in accordance withan embodiment of the present disclosure. Herein, two regions 602A and602B represent possible spatial configurations of a normal vector ^, whilean overlaid delta symbol 604, indicates a differential. The overlaid deltasymbol 604 points to a horizontal line 606 between the two regions602A and 602B that are parallel sections of surface of rotation, whichrepresents a partial coverage. The spacelike coverage condition isintrinsically complex due to the six-dimensional nature of the device 100, which encompasses both spatial and rotational degrees of freedom. Asthe normal vector ^ moves farther toward alignment with ^1, a numberof feasible motor states increases. This expanded set of motor states corresponds to the increased variability in configurations, making it challenging to maintain uniformity in the distribution of paths. Not all abstract path machines are capable of producing outputs such that the frames ^^,^^, where ^=^×^, are equally probable across the configuration space. This figure underscores the claim that the current machine incorporates advanced controls capable of supporting this novel statistical feature. These controls ensure that the spacelike coverage condition is maintained, allowing for uniform representation of configurations, even as the normal vector shifts toward extreme orientations. FIG. 6 is merely an example, which should not unduly limit the scope of the claims herein. A person skilled in the art will recognize many variations, alternatives, and modifications of embodiments of the present disclosure. Referring to FIG. 7, illustrated is a flowchart depicting steps of a method to simulate microgravity, in accordance with an embodiment of the present disclosure. At step 702, an accelerometer is mounted to the sample holder of the device. At step 704, acceleration data is collected by setting the accelerometer to record readouts for XYZ vectors at predetermined intervals continuously over a predetermined period. At step 706, moving average of the XYZ vector readouts are calculated for each predetermined interval throughout the predetermined period. At step 708, all the moving averages of the XYZ vectors recorded during the predetermined period are summed up. At step 710, the summedaverage is divided by the normalized acceleration value (go) of 9.80665m / s² to adjust for standard gravitational acceleration. At step 712, thesummed moving averages of XYZ vectors is divided by go to calculate thetime-averaged perceived gravitational acceleration (g) at a sensor site over the summed average period. At step 714, the calculated time- averaged perceived gravitational acceleration (g) is used to adjust the device's rotational path to simulate microgravity. The aforementioned steps are only illustrative and other alternatives can also be provided where one or more steps are added, one or more steps are removed, or one or more steps are provided in a different sequence without departing from the scope of the claims herein.

Claims

CLAIMS1. A device (100) to simulate microgravity, wherein the devicecomprises:- a cube-like outer frame (104) formed of a first closing frame(108a), a second closing frame (108b) and side frames (110a, 110b);- an inner rotating frame (102) within the cube-like outer frame;- a power and data transferring means;- a sample holder (106) fitted in the inner rotating frame; and- a sample holder rotating means (204),wherein the device has adaptable physical dimensions, featuring an outer frame in a range of 400x400x400-800x800x800 mm, and an inner sample area that provides an extensive volume in a range of 180x80x100-360x360x200 mm.

2. The device (100) of claim 1, wherein the sample holder comprisesone or more sensing means.

3. The device (100) of claim 1 or 2, wherein the sample holdercomprises a well plate clip holder.

4. The device (100) according to any of the preceding claims, whereinthe device further comprises a communication means.

5. The device (100) according to any of the preceding claims, furthercomprising at least one drive, wherein the at least one drive for at least one of: the first closing frame (108a), the second closing frame (108b), the side frames (110a, 110b), are operated by an ultrasonic piezo actuator.

6. A method to stimulate microgravity using the device (100)according to any of claims 1-5, the method comprises:- mounting an accelerometer to the sample holder of the device;- collecting acceleration data by setting the accelerometer to recordreadouts for XYZ vectors at predetermined intervals continuously over a predetermined period;- calculating moving average of the XYZ vector readouts for eachpredetermined interval throughout the predetermined period;- summing up all the moving averages of the XYZ vectors recordedduring the predetermined period;- dividing the summed average by the normalized acceleration value(go) of 9.80665 m / s² to adjust for standard gravitational acceleration;- dividing the summed moving averages of XYZ vectors by thenormalized acceleration value (go) to calculate the time-averagedperceived gravitational acceleration (g) at the sensor site over the summed average period;- using the calculated time-averaged perceived gravitational acceleration(g) to adjust the device’s rotational path to simulate microgravity.

7. The method according to claim 6, wherein the method comprisesapplying a transfer function for correlating one or more settings of the device.

8. The method of any of claims 6-7, wherein the method furthercomprises providing selective mechanical stress profiles.

9. The method according to any of claims 6-8, wherein methodcomprises computing trajectories.

10. The method according to any of claims 6-9, wherein the methodfurther comprises dynamically modifying paths via deformation of the equations of motion.

11. The method according to any of claims 6-10, wherein the methodfurther comprises modifying the device behavior so that when a user defined constraint system is used that comprises inequalities involving orclosely related to at least one of: sheer Stresses, Momenta, Forces, the device performs any one of: before moving alerts the user that the user defined constraint system is over-constrained and emits an error, begins moving and dynamically computes paths satisfying these user defined constraint system.

12. The method according to any of claims 6-11, wherein the methodfurther comprises using sensors of the device as true random number generators and transporting entropy supplied by sensor reading randomness to trajectories.

13. The method according to any of claims 6-12, wherein the methodfurther comprises fine-tuning adjustments in response to real-time stress and shear force data garnered from at least one sensor of the device, wherein the at least one sensor comprises: shear stress detectors, microfluidic channels, capacitive arrays, piezoelectric transducers.

14. The method according to any of the claims 6-13, further comprisingdynamically modifying paths via deformation of equations of motion by providing a non-zero potential function to attract or repel a dynamical state of the device to a required configuration.

Images