Abstract: Examples of methods are described. In some examples, a method includes generating, using a first branch of a machine learning model, a first agent map based on a layer image. In some examples, the method includes generating, using a second branch of the machine learning model, a second agent map. In some examples, the first agent map and the second agent map indicate printing locations for different agents.

Abstract: Examples disclosed herein relate to 3D printer fresh and recycled powder supply management. In one implementation, a processor estimates fresh and recycled powder use and recycled powder creation by a 3D printer based on a set of print jobs. The processor may coordinate fresh and recycled powder resources based on the powder estimate and stored information about powder resources.

Abstract: Examples of methods are described. In some examples, a method includes determining objects corresponding to a manufacturing period of three dimensional (3D) printing. In some examples, the method includes packing build volumes based on the objects. In some examples, the method includes simulating manufacturing powder degradation based on the build volumes. In some examples, the method includes determining a quantity of manufacturing powder consumption based on the manufacturing powder degradation. In some examples, the method includes adjusting a manufacturing parameter based on the quantity of manufacturing powder consumption.

Abstract: Examples of methods are described. In some examples, a method includes determining a quantification of a spatial neighborhood of a voxel of a build volume. In some examples, the method includes predicting, using a machine learning model, a manufacturing powder degradation based on the quantification and a position of the voxel.

Abstract: Examples of methods are described. In some examples, a method includes determining, using a variational autoencoder model, a latent space representation based on a three-dimensional (3D) input. In some examples, the 3D input represents a build of manufacturing powder. In some examples, the method includes predicting manufacturing powder degradation based on the latent space representation.

Abstract: Systems and methods of predicting temperature during a build of a three-dimensional (3D) part include determining a temperature profile at a plurality of layers of a part based on geometric characteristics of the 3D part as defined by a 3D part file, and adjusting a process parameter of the build based on the determined temperature.

Abstract: A computing device comprising a controller is disclosed herein. The controller is to access print data of a virtual build volume including a 3D object to be generated by a 3D printer; modify the print data to include a 3D structure at a location within the build volume to encapsulate an amount of build material; receive powder degradation data corresponding to the powder degradation of the encapsulated amount of build material; and calibrate an additive manufacturing parameter based on the powder degradation data.

Abstract: Examples of methods for powder reclamation are described. In some examples, a method includes estimating powder degradation for voxels of a three-dimensional (3D) manufacturing build based on a simulation. In some examples, the method includes determining a quantity of reclamation powder. In some examples, the quantity of reclamation powder may be determined based on the estimated powder degradation.

Abstract: A method of custom print mode generation in a three-dimensional (3D) printing device may include printing a plurality of parts with a plurality of 3D printing devices, the parts each being printed using different process parameters, and capturing a plurality of images of the parts. The method may also include, with an image analysis module, analyzing the images to classify the parts into a plurality of defect gradings, and adjusting a number of the process parameters based on characteristics of the parts identified by a user as undesirable. The examiner may also include, with a recommending module, creating a custom print mode based on the parts defect gradings and adjusted process parameters.

Abstract: A computing device comprising a controller is disclosed herein. The controller is to access print data of a virtual build volume including a 3D object to be generated by a 3D printer; modify the print data to include a 3D structure at a location within the build volume to encapsulate an amount of build material; receive powder degradation data corresponding to the powder degradation of the encapsulated amount of build material; and calibrate an additive manufacturing parameter based on the powder degradation data.

Abstract: An example system includes a simulation engine to determine a plurality of thermal states that will be experienced by powder at a voxel of a three-dimensional print volume as a result of printing a particular build. Each thermal state corresponds to a time during the printing or cooling from the printing. The system includes a stress engine to calculate a stress to the powder at the voxel based on the plurality of thermal states. The system includes a degradation engine to determine an amount of degradation of the powder at the voxel based on the stress.

Abstract: Examples of apparatuses for agent map generation are described. In some examples, an apparatus includes a memory to store a layer image. In some examples, the apparatus includes a processor coupled to the memory. In some examples, the processor is to generate, using a machine learning model, an agent map based on the layer image.

Abstract: Examples of methods for thermal image generation are described. In some examples, a method may include determining a score map based on first features from a model, a simulated thermal image at a first resolution, and second features of the simulated thermal image. In some examples, the method may include generating a thermal image at a second resolution based on the score map, the first features, and the second features, where the second resolution may be greater than the first resolution.

Abstract: Examples of methods are described. In some examples, a method includes generating, using a first branch of a machine learning model, a first agent map based on a layer image. In some examples, the method includes generating, using a second branch of the machine learning model, a second agent map. In some examples, the first agent map and the second agent map indicate printing locations for different agents.

Abstract: Examples of methods are described. In some examples, a method includes determining, using a variational autoencoder model, a latent space representation. In some examples, the latent space representation is of object model data. In some examples, the method includes predicting manufacturing powder degradation. In some examples, predicting the manufacturing powder degradation is based on the latent space representation.

Abstract: A system for detecting three-dimensional (3D) part drag includes an infrared image capture device to capture a plurality of thermal images of a 3D part build region of a 3D printing device on which a part is built, and an image analysis module to detect drag of the part based on a difference image produced by subtracting a first thermal image from a second thermal image.

Abstract: In an example, an apparatus includes an image processing system, a print engine, and a vision system. The image processing system generates electronic signals based on a model of an object to be fabricated using an additive manufacturing process. The print engine performs the additive manufacturing process in a plurality of passes based on the electronic signals. The vision system acquires a plurality of thermal images of the plurality of passes and assigns individual passes to individual images based on data acquired during a build of a calibration object by the additive manufacturing process. The print engine may further include a material coater to spread a powder coating material, a plurality of fluid ejection devices to eject a fusing agent, and an emitter to emit energy to fuse the fusing agent and the powder coating material into a layer of the object to be fabricated.

Abstract: Examples disclosed herein relate to 3D printer fresh and recycled powder supply management. In one implementation, a processor estimates fresh and recycled powder use and recycled powder creation by a 3D printer based on a set of print jobs.

Abstract: Examples disclosed herein relate to 3D printer fresh and recycled powder supply management. In one implementation, a processor estimates fresh and recycled powder use and recycled powder creation by a 3D printer based on a set of print jobs. The processor may coordinate fresh and recycled powder resources based on the powder estimate and stored information about powder resources.

Abstract: A control and feedback technique to determine the proper location of powder control temperature regions includes detecting solid parts having specified shape and thermal characteristics. The solid parts are generated by a 3D printer. The solid parts include powder material fused on a layer by layer basis. A powder control temperature region is selected in each layer of powder material, wherein the region is located in unfused powder material adjacent to the solid parts, and wherein the region is selected based on characteristics of the shape of a build area and thermal imaging associated with any of a structural formation of the solid parts adjacent to the region, and an amount of thermal energy radiated by the solid parts adjacent to the region. The amount of thermal energy to cause fusing of the powder material on each subsequent layer is modified based on location of the powder control temperature region.