System and Methods for Additively Manufacturing Highly Customized Structures
Systems and methods in accordance with embodiments of the invention implement additive manufacturing processes to fabricate highly customized products tailored to unique situations. In one embodiment, a method of additively manufacturing a highly customized product tailored to a particular individual includes: obtaining relevant information pertaining to the particular individual; developing a robust anthropomorphic model of the particular individual; establishing a goal that the desired product is intended to achieve; determining a design variable for the product; simulating numerous instances of varying product designs; determining at least one algorithm for assessing the efficacy of each of the simulated product designs, the algorithm accounting for the developed robust model; assessing the efficacy of each of the simulated product designs using the determined at least one algorithm; determining a product design suitable for fabrication based on the assessment; and additively manufacturing the product in accordance with the determined product design.
Latest Patents:
The current application claims priority to U.S. Provisional Patent Application No. 61/936,263, filed Feb. 5, 2014, the disclosure of which is hereby incorporated by reference.
FIELD OF THE INVENTIONThe present invention generally relates to techniques for additively manufacturing highly customized structures.
BACKGROUND‘Additive manufacturing,’ or ‘3D Printing,’ is a term that typically describes a manufacturing process whereby a 3D model of an object to be fabricated is provided to an apparatus (e.g. a 3D printer), which then autonomously fabricates the object by gradually depositing, or otherwise forming, the constituent material in the shape of the object to be fabricated. For example, in many instances, successive layers of material that represent cross-sections of the object are deposited or otherwise formed; generally, the deposited layers of material fuse (or otherwise solidify) to form the final object. Because of their relative versatility, additive manufacturing techniques have generated much interest.
SUMMARY OF THE INVENTIONSystems and methods in accordance with embodiments of the invention implement additive manufacturing processes to fabricate highly customized products tailored to unique situations. In one embodiment, a method of additively manufacturing a highly customized product tailored to a particular individual includes: obtaining relevant information pertaining to the particular individual that the desired product is meant for; developing a robust anthropomorphic model of the particular individual using the obtained relevant information; establishing a goal that the desired product is intended to achieve; determining at least one design variable for the product based on the established goal; simulating numerous instances of varying product designs, the variation being based on the determined at least one design variable; determining at least one algorithm for assessing the efficacy of each of the simulated product designs, the algorithm accounting for the developed robust model; assessing the efficacy of each of the simulated product designs using the determined at least one algorithm; determining at least one product design suitable for fabrication based on the assessment; and additively manufacturing the product in accordance with the determined at least one product design.
In another embodiment, a method of additively manufacturing a highly customized product tailored to a particular individual includes: obtaining anthropomorphic information about the particular individual that the product is meant for; developing an anthropomorphic model of the particular individual that characterizes at least some aspect of the particular individual's body and how it moves in space using the obtained anthropomorphic information, using a computational system; establishing at least one goal on the computational system that the desired product is intended to achieve; determining on the computational device at least one design variable for the product based on the established goal; simulating numerous instances of varying product designs on the computational device, the variations being based on the established at least one design variable; determining at least one algorithm for assessing the efficacy of each of the simulated product designs on the computational device, the algorithm accounting for the developed robust model; assessing the efficacy of each of the simulated product designs using the determined at least one algorithm on the computational device; determining at least one product design suitable for fabrication based on the assessment; and additively manufacturing the product in accordance with the determined at least one product design.
In yet another embodiment, the computational device contains machine learning algorithms that retain data obtained from previously run methods of the additive manufacture of highly customized products tailored to particular individuals, and uses the data to inform at least one performed computation.
In still another embodiment, the developed anthropomorphic model characterizes how the particular individual's body deforms when pressure is applied to it.
In yet still another embodiment, establishing at least one goal on the computational system is achieved by having a human input the goal on the computational system.
In a further embodiment, the determination of at least one design variable is accomplished by a human inputting the determined at least one design variable into the computational system.
In a still further embodiment, the determination of at least one design variable is accomplished by the computational system.
In a yet further embodiment, the at least one design variable is the localized elastic characteristics of the desired product.
In a still yet further embodiment, the at least one design variable is the localized thickness characteristics of the desired product.
In another embodiment, the at least one design variable is the cell structure in any implemented Voronoi structures.
In yet another embodiment, thousands of instances of varying product designs are simulated on the computational system.
In still another embodiment, millions of instances of varying product designs are simulated on the computational system.
In still yet another embodiment, the numerous instances of varying product designs include redundant product designs.
In a further embodiment, the determination of at least one algorithm is accomplished by the computational system.
In a yet further embodiment, the determination of at least one algorithm is accomplished by a human inputting the determined algorithm on the computational system.
In a still further embodiment, additively manufacturing the product includes using an active deposition technique.
In a still yet further embodiment, at least one established goal is that the desired product implements a CPAP mask that is custom fitted for the particular individual.
In another embodiment, at least one established goal is a generative goal.
In still another embodiment, at least one established goal is that the desired product implements a shaper that is an article of wear that tightly conforms to the particular individual's body and is configured to motivate a predetermined figure.
In yet another embodiment, at least one established goal is that the desired product provide footwear for that can enhance the athletic performance of the particular individual.
In still yet another embodiment, at least one established goal is that the desired product provide comfortable footwear for the particular individual.
Turning now to the drawings, systems and methods for implementing additive manufacturing processes that fabricate highly customized products that are tailored for unique situations are illustrated. In many embodiments, a method of manufacturing a product tailored for a particular individual includes: obtaining relevant information regarding the particular individual that the desired product is meant for; developing a robust anthropomorphic model of the particular individual using the obtained relevant information; establishing a goal that the desired product is intended to achieve; determining at least one design variable for the product based on the established goal; numerous instances of varying product designs, the variation being based on the determined at least one design variable; determining at least one algorithm for assessing the efficacy of each of the simulated product designs, the algorithm accounting for the developed robust model; assessing the efficacy of each of the simulated product designs using the determined at least one algorithm; determining at least one product design suitable for fabrication based on the assessment; and additively manufacturing the product in accordance with the determined at least one product design. In numerous embodiments, the product that is manufactured is an article of wear for a particular individual. In a number of embodiments, the at least one design variable is the localized elastic characteristics within the product. In various embodiments, the established goal is based on treating sleep apnea for a particular individual, e.g. via providing continuous positive airway pressure (for instance, as conventional CPAP masks do). In many embodiments, where the product to be manufactured is a CPAP mask, the obtained relevant information includes the geometry of a target user's face. In many embodiments, where the product to be manufactured is a CPAP mask, the established goal is based on treating sleep apnea using a CPAP mask that is designed to reduce potential discomfort experienced by the user when the CPAP mask is worn. In further embodiments, the established goal is based on facilitating laminar air flow through CPAP mask.
Since its inception, additive manufacturing, or ‘3D Printing’, has generated much interest from manufacturing communities because of the seemingly unlimited potential that these fabrication techniques can offer. For example, these techniques have been demonstrated to produce any of a variety of distinct and intricate geometries, with the only input being the final shape of the object to be formed. In many instances, a 3D rendering of an object is provided electronically to a ‘3D Printer’, which then fabricates the object. Many times, a 3D Printer is provided with a CAD File, a 3D Model, or instructions (e.g. via G-code), and the 3D Printer thereby fabricates the object. Importantly, as can be inferred, these processing techniques can be used to avoid heritage manufacturing techniques that can be far more resource intensive and inefficient. The relative simplicity and versatility of these techniques can pragmatically be used in any of a variety of scenarios including for example to allow for rapid prototyping and/or to fabricate components that are highly customized for particular consumers. For example, shoes that are specifically adapted to fit a particular individual can be additively manufactured. Indeed, U.S. Provisional Patent Application No. 61/861,376 discloses the manufacture of customized medical devices and apparel using additive manufacturing techniques; U.S. Provisional Patent Application No. 61/861,376 and its progeny are hereby incorporated by reference in its entirety, especially as they pertain to the additive manufacture of customized articles. It should also be mentioned that the cost of 3D printers has recently noticeably decreased, thereby making additive manufacturing processes an even more viable fabrication methodology.
Given the demonstrated efficacy and versatility of additive manufacturing processes, their potential continues to be explored. For example, while the operation of many current generation additive manufacturing apparatuses is premised on the uniform deposition of a material in the shape of the desired object such that the material properties of the corresponding printed object are largely homogenous throughout its structure, in many instances it may be desirable to additively manufacture multi-material objects. Accordingly, additive manufacturing apparatuses and techniques have recently been developed that can selectively deposit any of a plurality of different materials during the buildup of a desired object such that the printed object can be made up of a plurality of different materials. For example, Stratasys is a 3D Printing Company that develops 3D printers that can deposit any of a plurality of materials during the buildup of a single printed object, i.e. the printed object can be printed to include a plurality of distinct materials. For instance, the Objet Connex line of printers developed by Stratasys is adept at such ‘multi-material printing.’ Incidentally, Stratasys also boasts of its PolyJet Technology which allows 3D printing resolutions as fine as 0.0006″ per layer of deposited material to be achieved. PolyJet technology essentially involves depositing a plurality of drops of liquid photopolymer onto a build tray, and instantly uniformly curing the deposited drops with UV light.
Moreover, U.S. patent application Ser. No. 14/321,046, applied for by MetaMason, Inc., discloses additive manufacturing processes whereby the constituent material of an object to be fabricated is actively manipulated prior to, or during, the deposition process, such that different portions of the deposited constituent material can be made to possess different material properties; these deposition techniques can be referred to as ‘active deposition.’ U.S. patent application Ser. No. 14/321,046 is hereby incorporated by reference in its entirety, especially as it pertains to additively manufacturing objects that possess non-homogenous materials properties.
Nonetheless, even with these laudable achievements, the state of the art can further benefit from efficient and efficacious additive manufacturing techniques that fabricate highly-customized products tailored to particular individuals, featuring particularly implemented design characteristics, e.g. customized elastic characteristics. Thus, the instant application discloses efficacious and efficient additive manufacturing techniques for fabricating such highly customized products tailored to particular individuals, which utilize computationally derived design characteristics. For example, in one disclosed embodiment, a process for fabricating a highly customized CPAP mask tailored to a particular individual is implemented. Tailored CPAP masks can offer a host of advantages relative to conventional off the shelf CPAP masks. For instance, they can be worn with relatively less discomfort compared to conventional CPAP masks, which in turn can promote more consistent usage by the individual, and promote his/her wellbeing. These processes are now discussed in greater detail below.
Methods for Additively Manufacturing Highly-Customized Products for Unique SituationsIn numerous embodiments of the invention, additive manufacturing processes are implemented that fabricate highly-customized products tailored to unique situations. In many embodiments, the additive manufacturing processes rely on rigorous computation to derive a set of desirable design characteristics. For example,
The method 100 further includes developing 104 a robust anthropomorphic model of the particular individual using the obtained 102 information. For example, where the desired product is a shaper (i.e. an article of wear that typically tightly conforms to an individual's body and is intended to motivate a particular figure) designed for a particular individual, the developed model can be an anthropomorphic model that characterizes the particular individual's body. Moreover, the anthropomorphic model can be robust insofar as it accounts for the way the particular individual's body changes as it is reoriented in space (e.g. it accurately reflects the shaping of the particular individual's body as he/she is standing, sitting, or walking). While the instant method 100 is described with respect to fabricating products tailored to particular individual, it should be emphasized that the method can be generalized and applied to any of a variety of unique situations. For example, relevant information about a unique situation that a product is meant for can be obtained, and a robust model of the unique situation can be developed based on the obtained relevant information in accordance with embodiments of the invention. Generally, embodiments of the invention are not limited to designing and fabricating products for particular individuals; embodiments of the invention include designing and fabricating products for unique situations. The development of a robust model is described in greater detail in subsequent sections of the application.
The method 100 further includes establishing 106 a goal that the desired product is intended to achieve; any appropriate goal can be established in accordance with many embodiments of the invention. For example, in many embodiments, where the desired product is a shaper customized for a particular individual, the goal that is established 106 is the goal of having the shaper be capable of contorting the individual's body into a predefined figure in any of a variety of stances—e.g. the shaper should be capable of motivating a particular figure in the target individual regardless of whether the individual is standing or sitting. The establishing of a goal is described in greater detail in subsequent sections of the application.
The method 100 further includes determining 108 at least one product design variable based on the established goal. For example, where the desired product is a shaper, the at least one product design variable can be the localized elastic characteristics of the shaper. Additionally, the determination 108 can be carried out in any suitable way. For example, the determination 108 of the at least one product design variable can be conducted computationally, or it can be established by a human. In general, any suitable way of determining at least one product design variable based on the established goal can be implemented in accordance with embodiments of the invention. The determining of at least one product design variable based on the established goal is described in greater detail in subsequent sections of the application.
The method 100 further includes simulating 110 numerous instances of varying product designs, the variations being based on the determined at least one design variable. For example, where a shaper is the product to be manufactured, numerous instances of the shaper can be computationally simulated, the simulated numerous instances including shapers having various elastic characteristics. Moreover, in many embodiments, a vast number of such instances are simulated. For example, in some embodiments thousands of instances are simulated; in a number of embodiments, hundreds of thousands of instances are simulated; in many embodiments, millions of instances are simulated. Accordingly, in many embodiments, the simulations can be computationally intensive. In general, any number of instances can be simulated in accordance with embodiments of the invention. The simulation of numerous instances of varying product designs is described in greater detail in subsequent sections of the application.
The method 100 further includes determining 112 at least one algorithm for assessing the efficacy of the simulated product designs, the algorithm accounting for the developed robust model. Thus, for instance, where the desired product is a shaper that contorts a particular individual's body into a certain figure, the determined algorithm can measure how well each of the simulated designs, as applied to the robust model, contorts body into the desired shape. The determination 112 of the algorithm can be achieved in any suitable way in accordance with certain embodiments of the invention. For example, in many embodiments, the determination is achieved computationally. In several embodiments, the determination 112 is made by a human. The determination of at least one algorithm for assessing the efficacy of each of the simulated product designs is described in greater detail in subsequent sections of the application.
The method 100 further includes assessing 114 the efficacy of each of the simulated product designs using the determined at least one algorithm. In many embodiments, algorithm is applied to each of the simulated product designs as applied to the developed robust model, and the efficacy of each of the simulated product designs is thereby assessed. For example, where the desired product is a shaper, the algorithm can be applied to assess the efficacy of each of the simulated shapers (e.g. in view of the established goal) as applied to the robust model. The assessing of the efficacy of each of the simulated product designs using the determined at least one algorithm is described in greater detail in subsequent sections of the application.
The method 100 further includes determining 116 at least one product design suitable for fabrication based on the assessment. Thus, in some embodiments, a single product design is computationally determined as the algorithm deemed this product design more efficacious relative to the other product designs. In some embodiments, several product designs are determined, and offered to a human for final determination as to which of the several designs are to be fabricated. The determination of at least one product design suitable for fabrication is described in greater detail in subsequent sections of the application.
The method 100 further includes additively manufacturing 118 the desired product in accordance with the determined at least one product design. Any of a variety of additive manufacturing techniques can be implemented. For example, in some embodiments, For example, the investment molding techniques described in U.S. patent application Ser. No. 14/173,549, applied for by MetaMason, Inc., and PCT Patent Application No. PCT/US2014/049481, also applied for by Metamason, Inc., can be implemented in accordance with embodiments of the invention. The disclosures of U.S. patent application Ser. No. 14/173,549 and PCT Patent Application No. PCT/US2014/049481 are hereby incorporated by reference in their entirety, especially as they regard investment molding techniques. Additionally, as can be appreciated, in many embodiments, it is desirable that the highly customized products have varying materials properties throughout their structure. Thus, in many embodiments, the implemented additive manufacture techniques utilize active deposition techniques as disclosed in U.S. patent application Ser. No. 14/321,046, incorporated by reference above, which are well-suited to implementing structures having varying materials properties. But of course, it should be clear that any of a variety of deposition techniques can be implemented. The additive manufacture of the desired product in accordance with the at least one product design is described in greater detail in subsequent sections of the application.
As can be appreciated, the above-described method is general and can be applied in any of a variety of contexts, and can be used to fabricate any of a variety of products in accordance with embodiments of the invention. While certain examples have been mentioned, the described examples should not be construed as limiting the scope of the described methods. Note that the described methods can be achieved in any suitable way. In many embodiments, the many of the described aspects of the method are substantially performed on a device capable of computation including, but not limited to: a personal computer; a supercomputer; a tablet computer; and a cell phone. In general, any suitable device capable of such computations can be implemented. In many embodiments, many of the described aspects are substantially performed in a cloud server context. In numerous embodiments, such computational systems are well configured to handle ‘big data’, and implement associated algorithms, e.g. HADOOP and MAPREDUCE. Moreover, where the described processes are implemented on a computational system, the data passed back and forth can be encrypted. For example, where the product to be fabricated is a therapeutic device, and compliance with HIPAA privacy laws is desired, encryption techniques that satisfy the HIPAA privacy laws can be implemented.
Furthermore, it should be understood that the described aspects of the described methods can occur in any of a variety of sequences in accordance with embodiments of the invention. The order of the description is not meant to necessarily imply a particular order of operation. For example, in many embodiments, establishing a goal that the desired product is intended to achieve occurs prior to obtaining relevant information regarding the unique situation that the desired product is meant for.
Furthermore, while certain methods have been described, it should be understood that any of a variety of methods for fabricating highly customized products tailored to unique situations can be implemented. For example, any of a variety of computational methods can be implemented to derive a design for a desired product that can subsequently be additively manufactured. Moreover, in many embodiments, the described methods are modified so that they are used only to determine a desirable set of design characteristics for a desired product for a unique situation; they are not used to additively manufacture the product in accordance with the derived design characteristics. Thus, in many embodiments, the aspect of additively manufacturing the desired product is omitted. In several embodiments, the described methods further implement machine-learning algorithms that retain data from previously implemented methods and utilize this data to inform the operation of further methods. In general, as can be appreciated, the above-described methods can be implemented/modified in any of a variety of ways in accordance with embodiments of the invention. Thus, the description should not be construed as limiting. The aspects of the above described methods are now discussed in greater detail below.
Obtaining Relevant Information Regarding a Unique Situation that a Desired Product is Meant for
In many embodiments, relevant information regarding the unique situation that a desired product is to be additively manufactured for is obtained. As alluded to above, any of a variety of pieces of information can be deemed relevant and obtained; the particular pieces of information that are to be obtained can be largely context dependent. For example, in many embodiments, relevant information concerning the particular individual (e.g. his/her anatomical features) that the product is meant for is obtained, as many embodiments are implemented to fabricate products designed to fit particular individuals. For instance, in many embodiments, an additive manufacturing process for fabricating a highly-customized CPAP mask tailored for a particular individual is implemented. Accordingly, in many embodiments, the facial structure, including the particular nostril structure, of the user is obtained. This information can allow the CPAP mask to be additively manufactured so that it closely conforms to the particular user's facial contours, and thereby allows for comfortable wear. Moreover, with information concerning the individual's nostril structure, the CPAP mask can be additively manufactured such that it tightly conforms to the individual's nostrils, and thereby creates a tight seal, which can reduce undesirable leakage and allow for more effective mask operation.
Furthermore, in a number of embodiments, information concerning the subcutaneous structure (e.g. the physiological makeup of the facial structure under the skin) of the user's face is obtained. This information can be used to further inform the design of the CPAP mask. For example, information concerning the subcutaneous structure can be used to determine whether, and where, to make certain portions of the mask more or less elastic. For instance, wherever bone structure is detected, the CPAP mask can be designed such that it is more elastic at those regions to accommodate the rigid bone structure. In many embodiments, information regarding the subcutaneous structure can be used to allow the CPAP mask to be designed such that it includes moduli of elasticity that correspond with those of the associated localized regions of the face. This can allow the CPAP mask to retain its orientation when worn (e.g. if the individual is moving while sleeping). While several examples of relevant information are given for the additive manufacture of a CPAP mask, it should be clear that any of a variety of pieces of information can be deemed relevant and obtained in accordance with embodiments of the invention. For example, in many embodiments, information related to the health of the individual can be obtained, including but not limited to: the age of the individual, the individual's health history, information related to individual's lung capacity and/or typical breathing pattern; information concerning pulse oximetry, blood pressure, blood type, glucose levels, heart rate, EKG, MRI, PET, CT, IMS, genomics/binomics, and/or dermatological information can be obtained.
Additionally, as mentioned previously, it should be clear that the methodologies discussed in this application can be used to fabricate any of a variety of objects, and not just CPAP masks. Accordingly, it can be appreciated that the relevant information will depend on the particular product that is being additively manufactured, and its intended use. Thus, for instance, in many embodiments of the invention, the above-described methodologies are used to fabricate highly customized footwear tailored to a particular individual. Accordingly, the relevant information that is obtained can include (but is not limited to) the individual's foot structure, his/her typical walking/running/movement pattern, as well as the individual's principal intended use of the footwear (e.g. active wear or casual wear). As can be appreciated, each of these pieces of information can be informative in determining desirable design characteristics. In effect, it is seen that the above-described methodologies are general in nature and can be used to additively manufacture any of a variety of highly customized products for unique situations; accordingly, any of a variety of pieces of information can be deemed relevant and obtained so as to inform the design characteristics of the desired product.
The identified relevant information can also be obtained using any of a variety of techniques in accordance with embodiments of the invention. For example, in many embodiments a highly customized article of wear tailored for a unique individual is additively manufactured, and the relevant information that is obtained includes the geometry of the individual's body part(s) that the article of wear is to fit over. Accordingly, a 3D scanner can be used in accordance with embodiments of the invention to obtain the geometry information. Similarly, where a CPAP mask is to be additively manufactured, a 3D scanner can be used to obtain the geometry of the target user's facial structure. Of course, it should be understood that any suitable 3D scanner can be implemented. For example, in many embodiments, a FUEL 3D or Structure.IO scanner is used. Although, it should be reiterated that any suitable 3D scanner can be used, and more generally, any suitable way of obtaining the relevant information can be implemented. 3D scanning techniques are particularly effective insofar as the obtained data can be immediately provided to a device capable of computation (e.g. a computer) for subsequent computation.
Additionally, as discussed above, in many embodiments, subcutaneous information is considered relevant and is obtained. Similar to before, any suitable way for obtaining the subcutaneous structure information can be implemented. For instance, in many embodiments, thermal imaging techniques are used to obtain information concerning subcutaneous structure. Additionally, where stride and gait information are relevant and obtained, e.g. for the additive manufacture of highly customized footwear, this information can be obtained by video recording the particular individual's movement pattern. In several embodiments, a 3D video recording of the particular individual's movement pattern is obtained. Similarly, the pressure experienced by the individual's feet while standing can be relevant and obtained. For instance, the individual can be instructed to stand on a pressure-sensitive platform that measures localized pressure; all of this information can be used to inform the design of tailored additively manufactured footwear.
In general, as can be appreciated from the above discussion, any of a variety of pieces of information can be deemed to be relevant, and can be obtained in any of a variety of appropriate ways in accordance with embodiments of the invention. For instance, in some embodiments acoustic data is considered to be relevant, and obtained via an stereo recording. This obtained information can then be used to inform the final design characteristics of the desired product. In many embodiments, this obtained relevant information is provided as inputs for computation/simulation, and computations are performed so as to derive desirable design characteristics for the product to be manufactured. The development of a robust model of the unique situation using the obtained relevant information is now discussed below.
Developing a Robust Model of the Unique Situation Using the Obtained Relevant InformationIn many embodiments, a robust model of the unique situation using the obtained relevant information is developed. In many embodiments, the methodologies are used to fabricate products tailored to particular individuals; accordingly, in many embodiments, robust anthropomorphic models of the particular individuals are developed. The development can occur in any of a number of ways. In many embodiments a device capable of computation is used to develop the robust model. Thus, for example, where the desired product is a shaper for a particular individual, the robust model that can be developed is an anthropomorphic model of the particular individual. In many embodiments, the anthropomorphic model is sufficiently intricate that it accounts for the deformability of the individual, e.g. how which aspects of the individual are soft and deformable (e.g. flesh) and which aspects are hard and rigid (e.g. bony structure). In several embodiments, the anthropomorphic model accounts for the kinematic abilities of the particular individual (e.g. how the individual moves in space). The anthropomorphic model can be developed by aggregating the obtained relevant information—e.g. in the form of computer data, and interpreting it appropriately. For example, obtained relevant data including the geometry of the individual (e.g. obtained via a 3D scan), the subcutaneous structure of the individual (e.g. obtained via thermal imaging), and kinematic data (e.g. obtained via 3D recording of the individual in motion), can be computationally aggregated and used to develop the robust anthropomorphic model. As can be appreciated, the associated computations can be computationally intensive, and may be performed on an appropriately robust computationally capable system. For example, in many embodiments, the computations are performed on a supercomputer. In several embodiments, the computations are achieved in a cloud based system. As can be appreciated, the computationally capable systems can be configured to handle “big data”—e.g. they can be configured to implement HADOOP and/or MAPREDUCE algorithms where necessary. It should of course be appreciated that the computations can occur on any suitable system in accordance with embodiments of the invention, and are not limited to the described systems.
In many embodiments, the complexity of the robust model correspond to the desired fidelity in the determination of the final design characteristics for the desired part. Thus, where greater fidelity in the determined design characteristics is desired, the robust model is made to be more complex, and vice versa.
Note that while the development of an anthropomorphic model corresponding to a particular individual has been discussed, it should be clear that the development of a robust model can adopt any of a variety of forms in accordance with embodiments of the invention. For example, in some instances, the above-described methods are used to fabricate a CPAP mask, and the development of a robust model in this instance regards modeling only the facial structure. In general, any suitable robust model that can facilitate the derivation of desirable design characteristics can be developed in accordance with embodiments of the invention. Moreover, while the development of robust models regarding humans is discussed, the above-described techniques are broad and can be applied in any of a variety of situations. For instance, in some embodiments, the methods are applied to products for animals, and robust models of the animals are developed. More generally, the methods can be applied to any situation, not just those concerning humans/animals. Any of a variety of unique situations can be modeled in accordance with embodiments of the invention.
In many embodiments, at least one goal for the product is established, and this aspect is now discussed below.
Establishing a Goal that the Desired Product is Intended to Achieve
In many embodiments, a goal that the desired product is intended to achieve is established. In a number of embodiments, measurable criterion can be determined to assess to what extent the goal is being achieved. Thus for instance, where a CPAP mask is to be fabricated, the established goal can be to ‘implement a CPAP mask that can deliver sealed airflow through the nasal passages.’ Note that the generality of the goal statement can be varied based on the tolerability of the design variation. Thus for instance, in some embodiments, the established goal is more generally phrased as ‘expose a particular individual's nasal passage to a positive air pressure so as to maintain the clearance of the nasal passages.’ In general, the level of generality of the goal statement can be scaled in accordance with the tolerance of the flexibility of the design constraints: where the design is allowed to be more flexible, the established goal can be phrased more generally, and vice versa. For example, in some embodiments, the established goal is most generally phrased as ‘resolve a particular individual's sleep apnea.’ Such a broad goal statement can for example allow a computer to computationally derive an adequate solution. As can be appreciated, such a method would be best implemented on a sufficiently powerful computing system.
In general, the establishment of a goal can be accomplished by specifying what the desired product is, and can also be accomplished by specifying what the desired outcome is. Where the establishment of a goal is accomplished by specifying what the desired product is, the goal establishment can be thought of as ‘deterministic’ (e.g. it is ‘pre-determined’ what the final product is intended to be) Where the establishment of a goal is accomplished by specifying what the desired outcome is, the goal establishment can be thought of as ‘generative’ (e.g. the final product that will accomplish the goal will be ‘generated.’) The establishment of ‘generative’ goal statements can be somewhat thought of as allowing a solution to evolve. In other words, there is no bias toward a particular design; rather, a device capable of computation (e.g. a super-computer) can derive a suitable solution by simulating the testing of a vast number of designs, and converging on those that are most efficacious. This process can be thought of as being bio-mimetic.' In a number of embodiments, both deterministic goal statements are established and generative goal statements are established. In general, any of a variety of goal statements can be established.
It should be appreciated that the goal statement is substantially context-dependent, and can be manifested in any of a variety of ways. For example, where a shaper is to be implemented, the establishment of the goal can pertain to contorting a particular individual to a specified form. Where footwear is to be implemented, the establishment of the goal can pertain to, for example improving athletic performance for a particular individual and/or providing comfort for walking for a particular user. Notably, in many instances, a plurality of goals are established, and the derived design characteristics are derived so as to accommodate each of them to the extent possible. Any number of goals can be established in accordance with embodiments of the invention.
Based on the established goal statement, suitable design criterion that can be varied can be determined, and this aspect is now discussed in greater detail below.
Determination of Product Design VariablesIn many embodiments, based on the established goal, suitable product design variables are determined. Suitable design variables can be determined in any appropriate way in accordance with embodiments of the invention. For example, they can be computationally determined. In numerous embodiments, the suitable design variables are manually determined (e.g. determined by a human). This aspect can be thought of as parameterization. For example, in many embodiments, the design variables include at least one of: multi-material cell lattice sizing; material property control; printing/fabrication process & materials selection; tool pathing, controlling microstructure; cross-sectional structure; surface features, e.g. ribs, contours, tread. In other words, cell structures (e.g. Voronoi structures) can be implemented to form the material, and the sizing and composition of the cell structures can be varied. Additionally, the material properties can be directly varied (e.g. via active deposition techniques). The additive manufacturing process and material selection can be varied. The tool pathing, which can impact the design characteristics of the structure, can be varied. The cross-section of the deposited constituent material can be varied. And the surface features can be varied. Of course, as can be appreciated, any of a variety of suitable design variables can be established. For example, where a shaper is to be fabricated, the design variables can include the localized elastic characteristics. Where a CPAP mask is to be fabricated, the design characteristics can include the localized elastic characteristics as well as the structure of the inner walls of any passages related to the airflow. Where a shoe is to be fabricated, the product design variables can include the elastic characteristics of the sole of the shoe, as implemented via a Voronoi structure. As can be appreciated, any of a variety of suitable product design variables can be established.
Based on the determined product design variable(s), numerous instances embodying various design configurations of the product can be generated, and this aspect is now discussed in greater detail below.
Simulating Varying Product DesignsIn many embodiments, numerous instances embodying varying product designs are simulated. In many embodiments, the number of simulations is voluminous, and the simulations are meant to encompass a substantial portion of the universe of possible designs. The simulated product designs can vary based on the determined product design variables. In many embodiments, each simulation is performed on a separate ‘server blade’, and is assigned a unique identifier. In numerous embodiments, redundant simulations are also established, e.g. consistent with notions underlying the APACHE HADOOP framework. For example, in many embodiments at least three simulations of the identical simulated product are instantiated. In general, many embodiments are compatible with conventional understanding of “Big Data” frameworks. In many embodiments, a naming node is also created, and is replicated to avoid losing data.
In a number of embodiments separate sets of simulations are established for each of the generative goal statements and the deterministic goal statements. In this way, the efficacy of generative solutions can be compared with the efficacy of pre-determined form-factors. As one example, in embodiments where a solution for sleep apnea are to be additively manufactured, simulations that assess the efficacy of the customized CPAP masks are vetted by the described methods can be compared to those simulations that assess the efficacy of the solutions that are derived from the generative goal statements. Importantly, machine learning can be harnessed to enhance the efficacy of the described techniques over time. For example, where the described techniques are implemented by a computer, the computer can retain data from previous implementations of the technique and use any of a variety of known machine learning algorithms to become more efficient at the described techniques moving forward.
In numerous embodiments, algorithms for assessing the efficacy of the simulated designs are determined, and this aspect is now discussed below.
Determining the Algorithms for Assessing the Efficacy of Each of the SimulationsIn many embodiments, algorithms for assessing the efficacy of each of the numerous simulations are determined. In many embodiments, the algorithms account for the developed robust model. In many embodiments, the algorithms output an efficacy score. In numerous embodiments, the algorithms themselves are scored in terms of relative importance. In a number of embodiments, each algorithm is assigned a unique identifier, and assigned to an agent. The algorithms are then delivered to each simulation. As alluded to above, in some instances certain of the simulations can fail; the established redundancies can thereby help buffer against such lost data. With the efficacy algorithms delivered to each simulation, the efficacy algorithms can then be implemented, and this aspect is now discussed below.
Assessing the Efficacy of the Simulated DesignsIn many embodiments, the efficacy of each of the simulations is assessed using the determined algorithms. As mentioned above, in many embodiments, the efficacy algorithms output a numerical score corresponding to the measured efficacy. Accordingly, each algorithm can be implemented on each simulation. Where there are multiple algorithms, each algorithm can be assigned a relative weight, so that an overall score for each simulation can be computed. In this way, an overall efficacy score can be computed for each simulation. Of course, it should be understood that while efficacy algorithms are mentioned as a way of evaluating the simulated designs, any suitable method for measuring the efficacy of the simulated designs can be implemented. Having scored the efficacy of the designs, a narrowed set of designs can then be determined and this aspect is now discussed below.
Converging on a DesignIn many embodiments, at least one design is established as the most viable. The at least one design can be based on the efficacy assessments. In several embodiments, a set of designs are established; when a set of designs are established, a human can elect which of the set is most desirable. In many embodiments, designs that are concluded from ‘generative’ goal statements are deemed entirely unviable/unrealistic, and a filter is applied to eliminate these designs. The filter can be computational, or it can be a human. In some embodiments, at least one deterministic design is established, and at least one generative design is established. A human can then elect which of the at least one deterministic design and the at least one generative design is most viable. In some embodiments, none of the designs are established as sufficiently viable, and the method is re-run; the re-run method can take advantage of any applied machine-learning algorithms.
Once a sufficiently viable design is converged upon, the product can be additively manufactured in accordance with the design characteristics, and this aspect is now discussed below.
Additively Manufacturing the Desired Product in Accordance with the Computed Desired Design Characteristics
In many embodiments, the desired product is additively manufactured in accordance with the computed design characteristics. Any of a variety of additive manufacturing techniques can be implemented in accordance with embodiments of the invention. Thus, for instance, in some embodiments, conventional additive manufacturing techniques are implemented whereby successive layers of material that represent cross-sections of the object are deposited, or otherwise formed; generally, the deposited layers of material fuse (or otherwise solidify) to form the final object. In many embodiments, additive manufacturing techniques that allow for greater versatility in manufacture are implemented. For instance, in many embodiments, 6-axis 3D Printers are used to additively manufacture the desired object. 6-axis 3D printers are characterized in that they can deposit constituent material at any of a variety of angles. In other words, whereas conventional additive manufacturing apparatuses typically have build heads (responsible for the deposition of the constituent material) that only have 3 degrees of freedom, the build head of a 6-axis 3D printer is characterized by 6 degrees of freedom. This greater versatility can allow for the desired product to be more efficiently manufactured, and can allow the final product to be more structurally integral. For instance, 6-axis 3D printers can deposit strands of material in a continuous fashion, at any angle, as opposed to having to repeatedly raster across a build surface to build up the strand by iteratively depositing cross-sections of the strand.
For example, the use of 6-axis 3D printing techniques can be particularly suitable where the desired product is characterized by a Voronoi structure' that includes linking elements that adopt any of a variety of orientations. A Voronoi structure can be understood to be a 3D embodiment of a lattice based on a 3D Voronoi diagram including links, nodes, and optional interstitial material. Voronoi structures can be implemented to fabricate any of a variety of products, and they can be implemented so as to confer any of a variety of structural characteristics. For example, Voronoi structures can be made to implement a variation in the regional elasticity of a product to be fabricated. For instance,
For example, as mentioned above, the active deposition techniques described in U.S. patent application Ser. No. 14/321,046 (“the '046 application”), incorporated by reference above in its entirety, can be implemented so as to additively manufacture the desired product. These techniques can be particularly appropriate where the product is meant to include varied materials properties within its structure, as active deposition techniques can efficiently and effectively allow for the fabrication of structures having varying localized material characteristics. Notably, active deposition techniques can be used to vary any of a variety of materials properties. Importantly, any of the active deposition techniques disclosed in the '046 application, including those characterized by claims 20-30 of the as-filed application, can be implemented in accordance with embodiments of the invention. Briefly, the active deposition techniques described in the '046 application generally involve fabricating an object by: progressively depositing constituent material onto a surface to form the shape of the object to be fabricated in accordance with an additive manufacturing process; and manipulating the material properties of at least some portion of the constituent material that is deposited onto a surface such that at least some portion of the deposited constituent material possess different material properties than at least some other portion of the deposited constituent material; where manipulating the material properties of the at least some portion of the constituent material begins prior to, or concurrently with, its deposition onto a surface. As more thoroughly discussed in the '046 application, the material properties can be manipulated using any suitable technique such as one of: subjecting the at least some portion of the constituent material to electromagnetic waves; magnetizing the at least some portion of the constituent material; subjecting the at least some portion of the constituent material to a gas; vibrating the at least some portion of the constituent material; and heating the at least some portion of the constituent material. In some embodiments, implemented active deposition techniques can involve manipulating the cross-section of the deposited material as it is being deposited. Again, these techniques are more thoroughly elaborated on in the '046 application, which is incorporated by reference in its entirety, especially as it pertains to the disclosure of active deposition techniques.
In many embodiments, the investment molding techniques described in U.S. patent application Ser. No. 14/173,549 (“the '549 application”), incorporated by reference above, are used to additively manufacture the desired product in accordance with the computed design characteristics. These techniques can allow for the efficient fabrication of more intricate geometries. Briefly, investment molding fabrication techniques generally involve: fabricating a subassembly comprising a plurality of volumes; where each volume is defined by the homogenous presence or absence of a material; where fabricating the subassembly comprises using an additive manufacturing process; where at least one of the plurality of volumes defines a shape that is to exist in the object to be fabricated; where at least a first of the plurality of volumes comprises a first dissolvable material; dissolving the first dissolvable material; where the dissolution of the first dissolvable material does not dissolve at least one other material within the subassembly; forming at least one cavity within the subassembly; and introducing an additive material into the at least one cavity. These techniques can be advantageously used to fabricate any of a variety of structures such as a CPAP coupling as seen in
Examples of the fabrication of customized products in accordance with embodiments of the invention are now discussed below.
Additively Manufacturing a Customized ShoeIn many embodiments, techniques for additively manufacturing a highly customized shoe are implemented. Thus, for example, the structure of the shoe can be tailored so as to account for a particular individual's foot structure, the intended use of the shoe (e.g. athletics or casual wear), the individual's desired support/comfort balance, and the intended user's stride/gait. Accordingly, this information can be obtained prior to the additive manufacture of the shoe and can be used to inform the final design characteristics of the shoe. For instance, a pressure map can be obtained that characterizes the localized pressure exerted/felt by the user through his/her feet as he/she is standing, e.g. by having the individual stand on a platform that is designed to characterize the pressure exerted.
The obtained pressure map information can then be used to compute the desired design characteristics of shoe, e.g. using the above-described processes. In many embodiments, the end product is manifested via a Voronoi structure. Where a Voronoi structure is to be implemented, a computer can compute the Voronoi design—e.g. where the links/nodes are to exist—using the provided data. Thus,
As can be appreciated the above described techniques for fabricating highly customized products are general and can be implemented in any of a variety of ways, and can be used to fabricate any of a variety of products and not just shoes. Thus, for instance, the techniques can be used to additively manufacture a highly customized ‘shaper,’ and this aspect is now discussed below.
Additively Manufacturing a ShaperIn many embodiments, techniques for additively manufacturing a highly customized shaper are implemented. For example, in a number of embodiments, highly customized muscle shirts are fabricated that are meant to accentuate the contours of an individual's muscles. In several embodiments, articles of wear are fabricated that are meant to deform an individual's body into a more aesthetically pleasing form. As can be gleaned from the above discussion, the above described generalized techniques can be implemented in the fabrication of these articles of wear. Thus, for instance, where a highly customized ‘muscle shirt’ is desired, a 3D scan of the intended wearer's body can be obtained and used in the computation of the muscle shirt's final design characteristics. Additionally, in many embodiments, the individual's subcutaneous structure is also obtained, e.g. via thermal imaging. This information can also be used to compute how best to accentuate the individual's muscular structure.
With the obtained information, the design characteristics of the muscle shirt can be computed, e.g. the above described computation techniques can be implemented. For example, the shirt can be fabricated so as to encourage the repositioning of fat tissue into areas of the body associated with muscular definition.
Again, it should be reiterated that the above-described techniques are general and can be implemented in any of a variety of ways. Thus, for example, while a muscle shirt has been discussed above, in many embodiments, a woman's bra is fabricated. Thus,
While articles of wear that rely on customized elastic characteristics to promote a more aesthetically pleasing figure have been discussed, in many embodiments the ability to vary the elasticity across an article of wear is used for other purposes. For instance, in some embodiments, an athletic shirt having varying elastic characteristics that allows the wearer to advantageously harness the elastic energy that is stored when the shirt is stretched. Thus for instance, such an athletic shirt can be customized for a baseball pitcher, such that when the baseball pitcher ‘winds up’ to pitch the ball, the highly customized elastic characteristics of the athletic shirt will encourage the proper form in releasing the pitch. In some embodiments, an article of wear is customized to add elastic resistance to any of a variety of predefined movements. In effect, the article of wear can thereby facilitate resistance training. In a number of embodiments, an article of wear that provides for customized padding is implemented.
Again, as can be appreciated, the above described techniques are general and can be used to fabricate any of a variety of shapers. More generally, the techniques can be used to fabricate any of a variety of products. The fabrication of a highly customized CPAP mask tailored to a particular individual is now described in greater detail below.
Additively Manufacturing Highly Customized CPAP MasksIn many embodiments, the above-described generalized techniques are implemented so as to fabricate a highly-customized CPAP mask tailored for a particular individual. Discomfort is often cited as an issue when people elect not to don prescribed CPAP masks. Accordingly, many embodiments provide for highly tailored CPAP masks to make their use more comfortable and promote an individual's wellbeing. Accordingly, in many embodiments, a 3D structure of the target individual's facial geometry is obtained. As alluded to above, this can be obtained using any of a number of techniques, including 3D scanning. In numerous embodiments, the subcutaneous structure of the target individual's facial structure is also obtained. This can be obtained, e.g. via thermal imaging.
As before, the obtained data can be used to derive the desired design characteristics. For example, the subcutaneous structure information can be used to configure the coupling so as to minimize discomfort. Additionally, the facial/nostril geometry can be used to facilitate a laminar flow through the mask. The CPAP mask can then be fabricated so as to implement the desired design characteristics. Any suitable additive manufacturing technique can be implemented including any of the above-described techniques.
While the above description is relatively generalized, it should of course be realized that the generalized techniques can be implemented in any of a variety of ways in accordance with embodiments of the invention. Thus, it is discussed below, a process listing some of the nuances that can be implemented in accordance with certain embodiments of the invention is listed below.
Process for Additively Manufacturing CustomizedAs can be appreciated, the discussed methodologies for fabricating highly customized products using computational methods can be implemented in any of a variety of ways in accordance with embodiments of the invention. For example,
More generally, as can be inferred from the above discussion, any of the above-mentioned concepts can be implemented in a variety of arrangements in accordance with embodiments of the invention. For example, as can be appreciated by one of ordinary skill in the art, the sequence of many of the techniques applied in the above-described methods can be varied in any of a variety of ways. Accordingly, although the present invention has been described in certain specific aspects, many additional modifications and variations would be apparent to those skilled in the art. It is therefore to be understood that the present invention may be practiced otherwise than specifically described. Thus, embodiments of the present invention should be considered in all respects as illustrative and not restrictive.
Claims
1. A method of additively manufacturing a highly customized product tailored to a particular individual comprising:
- obtaining anthropomorphic information about the particular individual that the product is meant for;
- developing an anthropomorphic model of the particular individual that characterizes at least some aspect of the particular individual's body and how it moves in space using the obtained anthropomorphic information, using a computational system;
- establishing at least one goal on the computational system that the desired product is intended to achieve;
- determining on the computational device at least one design variable for the product based on the established goal;
- simulating numerous instances of varying product designs on the computational device, the variations being based on the established at least one design variable;
- determining at least one algorithm for assessing the efficacy of each of the simulated product designs on the computational device, the algorithm accounting for the developed robust model;
- assessing the efficacy of each of the simulated product designs using the determined at least one algorithm on the computational device;
- determining at least one product design suitable for fabrication based on the assessment; and
- additively manufacturing the product in accordance with the determined at least one product design.
2. The method of claim 1, wherein the computational device contains machine learning algorithms that retain data obtained from previously run methods of the additive manufacture of highly customized products tailored to particular individuals, and uses the data to inform at least one performed computation.
3. The method of claim 1, wherein the developed anthropomorphic model characterizes how the particular individual's body deforms when pressure is applied to it.
4. The method of claim 1, wherein establishing at least one goal on the computational system is achieved by having a human input the goal on the computational system.
5. The method of claim 1, wherein the determination of at least one design variable is accomplished by a human inputting the determined at least one design variable into the computational system.
6. The method of claim 1, wherein the determination of at least one design variable is accomplished by the computational system.
7. The method of claim 1, wherein the at least one design variable is the localized elastic characteristics of the desired product.
8. The method of claim 1, wherein the at least one design variable is the localized thickness characteristics of the desired product.
9. The method of claim 1, wherein the at least one design variable is the cell structure in any implemented Voronoi structures.
10. The method of claim 1, wherein thousands of instances of varying product designs are simulated on the computational system.
11. The method of claim 1, wherein millions of instances of varying product designs are simulated on the computational system.
12. The method of claim 1, wherein the numerous instances of varying product designs include redundant product designs.
13. The method of claim 1, wherein the determination of at least one algorithm is accomplished by the computational system.
14. The method of claim 1, wherein the determination of at least one algorithm is accomplished by a human inputting the determined algorithm on the computational system.
15. The method of claim 1, wherein additively manufacturing the product comprises using an active deposition technique.
16. The method of claim 1 wherein at least one established goal is that the desired product implements a CPAP mask that is custom fitted for the particular individual.
17. The method of claim 1, wherein at least one established goal is a generative goal.
18. The method of claim 1, wherein at least one established goal is that the desired product implements a shaper that is an article of wear that tightly conforms to the particular individual's body and is configured to motivate a predetermined figure.
19. The method of claim 1, wherein at least one established goal is that the desired product provide footwear for that can enhance the athletic performance of the particular individual.
20. The method of claim 1, wherein at least one established goal is that the desired product provide comfortable footwear for the particular individual.
Type: Application
Filed: Feb 5, 2015
Publication Date: Aug 6, 2015
Applicant:
Inventors: Leslie Oliver Karpas (Pasadena, CA), Aaron M. Ryan (Los Angeles, CA)
Application Number: 14/615,389
