Abstract: A computing system may include an initial design space engine and an active region adaptation engine. The initial design space engine may be configured to identify a design domain for which to optimize a topology based on an objective function and determine an active region. The active region adaptation engine may be configured to iteratively adapt the active region until an optimization ending criterion is satisfied. Iterative adaptation of the active region may include expanding the design domain to include branch design elements, performing finite element analysis (FEA) on the expanded design domain, and determining an adapted active region by activating some of the branch design elements based on an active sensitivity threshold and deactivating some of the active design element based on design variable value changes.

Abstract: A system and method for accelerated simulation setup includes receiving a description of a new problem for simulation, extracting input data and output data of previous simulation results, generating a representation of data based on the extracted input data and output data, and quantifying similarities between the new problem and the extracted input data and output data to identify a candidate simulation for the new problem. A machine learning component infers a solution output for the new problem based on extrapolation or interpolation of outputs of the candidate simulation, thereby conserving resources by eliminating a simulation generation and execution. Alternatively, an efficient simulation setup can be generated using the queried knowledge, input variables, and input parameters corresponding to the candidate simulation.

Abstract: Examples of techniques for machine learning-based system architecture determination are described herein. An aspect includes receiving a system architecture specification corresponding to a system design, and a plurality of topological variants of the system architecture specification. Another aspect includes determining a system architecture graph based on the system architecture specification. Another aspect includes classifying, by a neural network-based classifier, each of the topological variants as a feasible architecture or an infeasible architecture based on the system architecture graph. Another aspect includes identifying a subset of the feasible architectures as system design candidates based on performance predictions.

Abstract: There are disclosed systems and methods for managing object configurations within a structure. For one embodiment, a fixed object profile including dimensions and locations of ventilation of a building space and a non-fixed object profile including dimensions and a location of a non-fixed object are identified. One or more contaminant risk locations of the building space are determined in response to determining the air flow within the building space associated with an HVAC unit. The air flow is determined based on incoming air flow streams to the building space and outgoing air flow streams away from the building space. One or more object positions are provided at an output device based on the contaminant risk location(s). For another embodiment, the contaminant risk location(s) are selected from possible people locations. A person position is displayed at an output device, in proximity to the building space, based on the contaminant risk location(s).

Abstract: A computer-implemented method is provided for predicting a fatigue response of a material. The method includes receiving a user input specifying one or more surface roughness parameters that characterize a surface of a material for which fatigue life is to be predicted. The method further includes generating at least one realistic virtual surface profile from the specified one or more surface roughness parameters. The method further includes predicting fatigue life of the material in dependence of a stress field applied to the generated virtual surface profile. In accordance with specific embodiments, the prediction of the fatigue life may be carried out using finite element analysis based simulations, machine learning methods, or combinations thereof.

Abstract: Methods for modeling of parts with lattice structures and corresponding systems and computer-readable mediums. A method includes receiving a model of an object to be manufactured. The method includes receiving a user specification of a void region within the model to create a lattice. The method includes performing a trimming operation to create a trimmed lattice by tessellating void surfaces and grouping together at least one row of connected rods to be treated as a single entity.

Abstract: A system for autonomous generative design in a system having a digital twin graph a requirements distillation tool for receiving requirements documents of a system in human-readable format and importing useful information contained in the requirements documents into the digital twin graph, and a synthesis and analysis tool in communication with the digital twin graph, wherein the synthesis and analysis tool generates a set of design alternatives based on the captured interactions of the user with the design tool and the imported useful information from the requirements documents. The system may include includes a design tool with an observer for capturing interactions of a user with the design tool, In addition to the observer, an insighter in communication with the design tool and with the digital twin graph receives design alternatives from the digital twin graph and present the receive design alternatives to a user via design tool.

Abstract: A computing system may include a design access engine and a design processing engine. The design access engine may be configured to access an object design to be constructed through additive manufacturing. The design processing engine may be configured to represent the object design as a combination of coarse geometric elements and high-resolution lattice elements and process the object design based on both the coarse geometric elements and the high-resolution lattice elements. Processing of the object design may include generation of lattice infills, lattice simulations, or a combination of both.

Abstract: A method of optimizing an additive manufacturing (AM) process includes receiving at least one design parameter of the AM process, receiving information relating to uncertainty in at least one other parameter of the AM process, performing uncertainty quantification in the optimization processor based on the at least one design parameters and uncertainty information to identify a shape error in an object being produced, updating the at least one design parameter of the AM process and utilizing the updated at least one design parameter in the AM process. A system for optimizing an AM process includes a design processor to produce at least one design parameter for an object to be manufactured, and an optimization processor to receive the at least one design parameter and uncertainty information to identify a shape error in the object to be manufactured and update the design parameters based on the shape error, prior or during the manufacturing process.

Abstract: A method includes receiving, via a first component in a production environment, a sensor measurement corresponding to a second component in the production environment. A first digital twin corresponding to the first component is identified, and a perception algorithm is applied to identify a component type associated with the second component. A second digital twin is selected based on the component type, and a third digital twin is selected that models interactions between the first digital twin and the second digital twin. The third digital twin is used to generate instructions for the first component that allow the first component to interact with the second component. The instructions may then be delivered to the first component.

Abstract: A system and method for accelerating topology optimization of a design includes a topology optimization module configured to determine state variables of the topology using a two-scale topology optimization using design variables for a coarse-scale mesh and a fine-scale mesh for a number of optimization steps. A machine learning module includes a fully connected deep neural network having a tunable number of hidden layers configured to execute an initial training of a machine learning-based model using the history data, determine a predicted sensitivity value related to the design variables using the trained machine learning model, execute an online update of the machine learning-based model using updated history data, and update the design variables based on the predicted sensitivity value. The model predictions reduce the number of two-scale optimizations for each optimization step to occur only for initial training and for online model updates.

Abstract: A system and method is provided that facilitates optimizing tool paths based on thermal/structural simulations of a part produced via a 3D-printer. A processor may carry out a first simulation of the part being additively produced according to a first set of tool paths that correspond to instructions usable to drive the 3D-printer to produce the part. The first simulation may include: determining a hexahedral mesh of the part that includes a plurality of hexahedron elements; determining an order of the elements of the mesh to deposit for additively producing the part based on the first set of tool paths; and simulating an incremental deposit of each of the elements of the mesh in the order that the elements are determined to be deposited.

Abstract: Methods and systems are disclosed for generation of cellular lattice kernels optimized by multiple objectives for highly specific targeted properties of geometry and topology rather than state of the art methods that rely on a predefined kernel library. Using a characterization of virtual kernel features, bulk material properties can be predicted using approximations from the virtual kernel rather than having to rely solely on experimental finite element simulations of lattice structures.

Abstract: A computing system may include an initial design space engine and an active region adaptation engine. The initial design space engine may be configured to identify a design domain for which to optimize a topology based on an objective function and determine an active region. The active region adaptation engine may be configured to iteratively adapt the active region until an optimization ending criterion is satisfied. Iterative adaptation of the active region may include expanding the design domain to include branch design elements, performing finite element analysis (FEA) on the expanded design domain, and determining an adapted active region by activating some of the branch design elements based on an active sensitivity threshold and deactivating some of the active design element based on design variable value changes.

Abstract: A system and method are provided for adaptive domain reduction for thermo-structural simulation of an additive manufacturing process. The system may include a processor configured to carry out a simulation of a part being additively produced according to a set of tool paths. The simulation may include determining an original mesh of the part; determining an order of the elements of the original mesh to deposit; and simulating an incremental deposit of each of the elements of the original mesh for a material in the order that the elements are determined to be deposited. For each incremental deposit of an additional respective element the processor may determine thermal characteristics and structural deformation characteristics of the deposited elements.

Abstract: A method of optimizing an additive manufacturing (AM) process includes receiving at least one design parameter of the AM process, receiving information relating to uncertainty in at least one other parameter of the AM process, performing uncertainty quantification in the optimization processor based on the at least one design parameters and uncertainty information to identify a shape error in an object being produced, updating the at least one design parameter of the AM process and utilizing the updated at least one design parameter in the AM process. A system for optimizing an AM process includes a design processor to produce at least one design parameter for an object to be manufactured, and an optimization processor to receive the at least one design parameter and uncertainty information to identify a shape error in the object to be manufactured and update the design parameters based on the shape error, prior or during the manufacturing process.

Abstract: A system for autonomous generative design in a system having a digital twin graph a requirements distillation tool for receiving requirements documents of a system in human-readable format and importing useful information contained in the requirements documents into the digital twin graph, and a synthesis and analysis tool in communication with the digital twin graph, wherein the synthesis and analysis tool generates a set of design alternatives based on the captured interactions of the user with the design tool and the imported useful information from the requirements documents. The system may include includes a design tool with an observer for capturing interactions of a user with the design tool, In addition to the observer, an insighter in communication with the design tool and with the digital twin graph receives design alternatives from the digital twin graph and present the receive design alternatives to a user via design tool.

Abstract: A system and method are provided for adaptive domain reduction for thermo-structural simulation of an additive manufacturing process. The system may include a processor configured to carry out a simulation of a part being additively produced according to a set of tool paths. The simulation may include determining an original mesh of the part; determining an order of the elements of the original mesh to deposit; and simulating an incremental deposit of each of the elements of the original mesh for a material in the order that the elements are determined to be deposited. For each incremental deposit of an additional respective element the processor may determine thermal characteristics and structural deformation characteristics of the deposited elements.

Abstract: Systems (500) and methods (400) for an interactive system for automatic generation, analysis and exploration of composable system of systems based on knowledge graphs. A method (400) includes receiving (405) a scenario (110) and a domain ontology (111); determining (410) structures (132), attributes (133), and capabilities (131) from the domain ontology; generating (415) design alternatives (146) based on the scenario using the structures, attributes, and capabilities; performing (430) an evaluation (159) of the design alternatives based on the scenario; generating (445) an SoS design (300) based on the evaluation; and displaying the SoS design to a user.

Abstract: A system and method for accelerated simulation setup includes receiving a description of a new problem for simulation, extracting input data and output data of previous simulation results, generating a representation of data based on the extracted input data and output data, and quantifying similarities between the new problem and the extracted input data and output data to identify a candidate simulation for the new problem. A machine learning component infers a solution output for the new problem based on extrapolation or interpolation of outputs of the candidate simulation, thereby conserving resources by eliminating a simulation generation and execution. Alternatively, an efficient simulation setup can be generated using the queried knowledge, input variables, and input parameters corresponding to the candidate simulation.