Abstract: One embodiment of the present invention sets forth a technique for performing machine learning. The technique includes applying one or more placement rules to a floorplan of a building to generate a set of candidate column locations in the floorplan. The technique also includes selecting, using a first reinforcement learning (RL) agent, one or more column locations from the set of candidate column locations based on a structural stability of the one or more column locations. The technique further includes outputting the floorplan that includes the one or more column locations as a structural design for the building.

Abstract: A method, apparatus, system, and computer program product provide the ability to dynamically generate a digital building information model. Design data for various designs is received. The design data for each design is encoded into a graph. A knowledge base is maintained and defines a model of design intent while processing and storing the graph. First user input of a goal or constraint is received. The knowledge base generates solutions based on the input. Second user input based on the solutions is received and used to iteratively train the knowledge base. The solutions are then output.

Abstract: In various embodiments, an intent-driven layout application automatically generates design for floor spaces. The intent-driven layout application generates a logic formula based on a statement of a design intent and at least one fuzzy geometric predicate. The intent-driven layout application computes, for a first spatial object, a set of desirability values for a set of candidate placements within a first design based on the logic formula. Based on the set of desirability values, the intent-driven layout application selects a first candidate placement from the set of candidate placements. Subsequently, the intent-driven layout application generates a second design based on the first design, where the first spatial object has the first candidate placement within the second design.

Abstract: A method, apparatus, system, and computer program product provide the ability to dynamically generate a digital building information model. Design data for various designs is received. The design data for each design is encoded into a graph. A knowledge base is maintained and defines a model of design intent while processing and storing the graph. First user input of a goal or constraint is received. The knowledge base generates solutions base don the input. Second user input based on the solutions is received and used to iteratively train the knowledge base. The solutions are then output.

Abstract: A design engine consolidates portions of a mechanical assembly design to reduce the number of components included in the design. The design engine analyzes the design to determine various criteria associated with the assembly. Then, the design engine identifies a group of components within the design to be consolidated. The design engine determines a volumetric region where the group of components resides and then subdivides the volumetric region. The design engine then initiates a generative design process based on the determined criteria to create geometry within each subdivision of the volumetric region. The newly generated geometry includes fewer components than the initial group of components. The design engine then replaces the group of components with the newly generated geometry, thereby consolidating the group and reducing the total number of components included in the design.

Abstract: One embodiment of the present invention sets forth a technique for performing machine learning. The technique includes applying one or more placement rules to a floorplan of a building to generate a set of candidate column locations in the floorplan. The technique also includes selecting, using a first reinforcement learning (RL) agent, one or more column locations from the set of candidate column locations based on a structural stability of the one or more column locations. The technique further includes outputting the floorplan that includes the one or more column locations as a structural design for the building.

Abstract: A method, apparatus, system, and computer program product provide the ability to dynamically generate a digital building information model. Design data for various designs is received. The design data for each design is encoded into a graph. A knowledge base (consisting of a collection of the design data, actions taken on the design data, and interpretations of the received design data) is maintained. The knowledge base processes and stores the graph, and indexes and provides access to design knowledge. The knowledge base is iteratively trained based on the graph and updates to the graph, and translates user input for new design projects into actionable design models, documentation, and analytical data. User input (e.g., a sketch or bubble diagram) is received. As the user input is received, a layout floorplan is generated and displayed in real-time (based on the user input and the knowledge base).

Abstract: A method and system provide the ability to optimize a structural engineering design. A dataset is synthesized by acquiring a structural skeleton design of an entire building. The skeleton defines locations and connectivities of bars that represent columns or beams. The skeleton design is represented as a structural graph with each bar represented as a graph node and edges connecting graph nodes. Structural simulation results are computed for the synthetic dataset based on the structural graph, various loads, and a structural analysis. A simulation model and a size optimization model are trained based on the structural simulation results with the size optimization model determining cross-section sizes for the bars to satisfy a building mass objective, building constraints, and output from the simulation model. The structural engineering design is output from the size optimization model.

Abstract: An iterative design environment performs an iterative design process of a product by implementing usage feedback of the product when utilized under real-world conditions. Sensors are installed on the physical product and collect data about the behavior of the product under real-world conditions. The sensor data comprise usage feedback implemented to inform and produce a design problem statement and one or more design solutions. The sensor data is received by a problem statement engine to produce a problem statement based, at least in part, on the sensor data. A design engine then produces one or more design solutions for the problem statement and one of the design solutions is fabricated to produce a new physical product. Sensors are then installed onto the new physical product and the iterative design process may be performed again. The iterative design process may be performed multiple times until a satisfactory physical product is achieved.

Abstract: A method and system provide the ability to parametrize a sketch. A sketch is acquired and includes raster lines that define a raster image based floor-plan sketch. Vectorized geometry is generated from the sketch dynamically in real time based on raster lines. A parametric model that is optimizable is generated from the vectorized geometry. The parametric model is generated dynamically in real time, and the raster lines are represented in the parametric model as three-dimensional walls. The parametric model is displayed and edited. Upon editing a parameter of a three-dimensional wall, other parameters in the parametric model are autonomously updated.

Abstract: A method and system provide the ability to parametrize a sketch. A sketch is acquired and includes raster lines that define a raster image based floor-plan sketch. Vectorized geometry is generated from the sketch dynamically in real time based on raster lines. A parametric model that is optimizable is generated from the vectorized geometry. The parametric model is generated dynamically in real time, and the raster lines are represented in the parametric model as three-dimensional walls. The parametric model is displayed and edited. Upon editing a parameter of a three-dimensional wall, other parameters in the parametric model are autonomously updated.

Abstract: A design engine consolidates portions of a mechanical assembly design to reduce the number of components included in the design. The design engine analyzes the design to determine various criteria associated with the assembly. Then, the design engine identifies a group of components within the design to be consolidated. The design engine determines a volumetric region where the group of components resides and then subdivides the volumetric region. The design engine then initiates a generative design process based on the determined criteria to create geometry within each subdivision of the volumetric region. The newly generated geometry includes fewer components than the initial group of components. The design engine then replaces the group of components with the newly generated geometry, thereby consolidating the group and reducing the total number of components included in the design.

Abstract: A method, apparatus, system, and computer program product provide the ability to dynamically generate a digital building information model. Design data for various designs is received. The design data for each design is encoded into a graph. A knowledge base (consisting of a collection of the design data, actions taken on the design data, and interpretations of the received design data) is maintained. The knowledge base processes and stores the graph, and indexes and provides access to design knowledge. The knowledge base is iteratively trained based on the graph and updates to the graph, and translates user input for new design projects into actionable design models, documentation, and analytical data. User input (e.g., a sketch or bubble diagram) is received. As the user input is received, a layout floorplan is generated and displayed in real-time (based on the user input and the knowledge base).

Abstract: One embodiment of the present invention sets forth a technique for generating one or more designs for a structural frame, the method comprising: receiving an input frame and an optimization objective that indicates a design goal for the one or more designs; based on the input frame, generating multiple candidate frames via a divergent search algorithm; based on the optimization objective, generating a different solution frame for each candidate frame via a convergent search algorithm; and determining a quality factor for each solution frame that enables a quantitative comparison with respect to the optimization objective of the solution frame with each other solution frame.

Abstract: An iterative design environment performs an iterative design process of a product by implementing usage feedback of the product when utilized under real-world conditions. Sensors are installed on the physical product and collect data about the behavior of the product under real-world conditions. The sensor data comprise usage feedback implemented to inform and produce a design problem statement and one or more design solutions. The sensor data is received by a problem statement engine to produce a problem statement based, at least in part, on the sensor data. A design engine then produces one or more design solutions for the problem statement and one of the design solutions is fabricated to produce a new physical product. Sensors are then installed onto the new physical product and the iterative design process may be performed again. The iterative design process may be performed multiple times until a satisfactory physical product is achieved.