Abstract: In one embodiment, a modeling process adds unique numbers for field welds or flange bolt-up connections to a virtual construction model of an infrastructure construction project. Based on the virtual construction model and a fabrication data file that describes a portion of the infrastructure construction project, a QC drawing generation process generates a QC drawing for the portion of the infrastructure construction project that includes balloons linked to locations of field welds or flange bolt-up connections, each balloon including the unique number of a corresponding field weld or flange bolt-up connection. A copy of the QC drawing is provided to a field inspection worker. Subsequently, a marked-up copy of the QC drawing is received back that includes field inspection results associated with the unique number of each field weld or flange bolt-up connection. The marked-up copy may be maintained in a document control system.

Abstract: In example embodiments, techniques are provided to implement changeset conflict rebasing when performing conflict-detection and merging in an infrastructure modeling software architecture that uses an optimistic concurrency policy. Changeset conflict rebasing involves adjusting the pre-change values in a local changeset so they match post-change values of a remote version, rather an original base version, or removing changes from the local changeset entirely.

Abstract: In various embodiments, techniques are provided for coordinating the assignment of codes to infrastructure elements in an infrastructure modeling software architecture. A chain of authority is provided from a code authority local to a client to a cloud-based code authority service. Further, a code is structured in three parts, including a code specification part identifying a code specification that indicates what information is encoded into the code and how it is encoded, a code scope part based on boundaries over which the code authority guarantees the code is unique, and a code value part with a unique value.

Abstract: In example embodiments, techniques are provided for converting geometry of an infrastructure model represented as high-level geometric primitives having a given symbology to low-level primitives, and encoding these low-level primitives into tile contents suited for transmission, and ultimately display upon, on a client device. An architectural split between frontend applications and backend applications may allow conversion of the high level primitives to be performed remote from the client devices by backend applications. Backend applications may be executed on robust hardware devices that execute software in a multi-threaded environment on powerful CPUs. By performing CPU-intensive operations on backend applications, frontend applications on client devices may be required to do little more than submit low-level primitives for rendering on their GPU.

Abstract: In example embodiments, techniques are provided for frontend-driven, incremental HLOD sub-tree creation and dynamic generation of tiles. An infrastructure modeling frontend module (e.g., a iModel.js Frontend module) of a frontend application may incrementally create a local HLOD sub-tree, beginning with an initial HLOD sub-tree based on initial metadata provided without tile contents, and refined based on metadata provided with tile contents, from an infrastructure modeling backend module (e.g., a iModel.js Backend module) of a backend application. The local HLOD sub-tree of each infrastructure modeling frontend module may differ based on the individual tiles requested and the tile refinement strategies employed. The infrastructure modeling backend module may not maintain an HLOD tree, and simply generates tiles upon request, serving them and storing them in a tile cache incase they are needed again.

Abstract: In example embodiments, techniques are provided integrating 2D images into the display of a 3D reality mesh to recover lost context. A determinization is made whether there is level of detail (LOD) in the 3D reality mesh sufficient to provide full context for a desired view of the 3D reality mesh. When such a LOD does not exist, a subset of the source 2D images that intersect a view frustrum for the desired view is selected from the set of 2D images used to reconstruct the 3D reality mesh. The selected source 2D image is evaluated to determine if it would be visually appropriate for the desired view. If the selected source 2D image is determined to be visually appropriate, the 3D reality mesh is replaced with the selected source 2D image or a portion thereof.

Abstract: In an example embodiment, techniques are provided for locking a region of fully-connected large-scale multi-dimensional spatial data (e.g., a large-scale 3-D mesh) defined by a bounding box. A region is associated with a lock state (e.g., exclusive or sharable). Clients may access the fully-connected large-scale multi-dimensional spatial data based on a comparison of the bounding box of the requested spatial data to the bounding boxes of other client's locks and their lock state.

Abstract: Techniques are provided for maintaining a class hierarchy of an object-oriented data model in tables of a relational database that minimize performance impact (e.g., in terms of both processing resources and storage footprint). In one embodiment, an object-relational mapping system accesses a class hierarchy of an object-oriented data model. A plurality of tables of a relational database are generated, including a primary table, a joined table and an overflow table. The object-relational mapping system maps classes of the class hierarchy to the primary table, joined table and overflow table, wherein the mapping maps properties of classes to columns of the primary table and joined table until a column threshold is reached and then maps excess properties to columns of the overflow table, and the mapping shares at least one column of the primary table, joined table or overflow table among classes such that multiple different properties are mapped to the shared column.

Abstract: In various embodiments, techniques are provided for photogrammetric 3D model reconstruction that modify the optimization performed in bundle adjustment operations of an automatic SfM stage to apply a depth-aware weighting to reprojection error of each 3D point used in the optimization. The reprojection error of each 3D point may be weighted based on a function of distance, density of a cluster, or a combination of distance and density. A loss function may be scaled to account for the weighting, and normalizations applied. Such weighting may force consideration of 3D points on an object of interest in the foreground and improve convergence of the optimization to global optima. In such manner, accurate and complete 3D models may be reconstructed of even ill-textured or very thin objects in the foreground of a scene with a highly textured background, while not consuming excessive processing and storage resources or requiring tedious workflows.

Abstract: In example embodiments, techniques are provided for smoothing a mesh to remove unwanted bumpiness on regular surfaces. In one example embodiment, an editor determines a capture shape (e.g., a user-specified capture shape) that and then extracts a set of vertices of a multi-resolution mesh that include vertices that intersect the capture shape. The editor generates a fitted shape from the extracted set of vertices that more precisely defines the portion of the mesh to be smoothed, the fitted shape to have at least one of a different size or a different orientation than the capture shape. The editor then modifies the vertices that fall within or close to the fitted shape to change their coordinates to smooth the portion of the multi-resolution mesh. The modified vertices are persisted to a storage device for subsequent display or analysis.

Abstract: In an example embodiment, techniques are provided for concurrently editing fully-connected large-scale multi-dimensional spatial data (e.g., a large-scale 3-D mesh) by ensuring that edits performed by multiple clients are non-conflicting edits that are “trivially” mergeable (e.g. mergeable simply via cut-and-paste operations). Conflicting edits may be prevented by locks (e.g., region-based locks). Non-conflicting edits that require “non-trivial” merging may be prevented through the use of marked read-only boundaries.

Abstract: In one embodiment, an augmented reality application executing on an augmented reality device accesses a representation of the physical environment. The augmented reality application aligns information that describes near-ground features with the representation of the physical environment. After alignment, the augmented reality application generates an augmented reality view that is displayed on a display device by projecting the near-ground features onto the ground in the representation of the physical environment, generating a virtual plane above the ground that is parallel to the ground and separated therefrom by a distance, projecting the near-ground features onto the virtual plane above the ground, and showing an indicator of visual correspondence that connects at least a portion of the projection of near ground features on the ground and a corresponding portion of the projection of near-ground features on the virtual plane above the ground.

Abstract: In an example embodiment, techniques are provided for displaying contour lines on a multi-resolution mesh substantially in real-time. Contour lines may be computed on a per-tile basis, scaling for various resolutions. The mesh and computed contour lines from lower resolution tiles may be displayed as temporary (referred to hereinafter as “overview”) data while the mesh and contour lines for higher resolution tiles are obtained or computed, to enable substantially real-time update. The techniques may handle very large meshes and large numbers of contour lines, without unduly taxing hardware resources. The techniques may also be applicable to multiple types of meshes (e.g., 2-D, 2.5-D, 3-D, 4-D, etc.).

Abstract: In various embodiments, techniques are provided for clipping and displaying a multi-resolution textured mesh using asynchronous incremental on-demand marking of spatial index nodes to allow for substantially real-time display refresh after a change is made to clip geometry. Timestamps may be added to spatial index nodes and an upper bound placed on the number of operations performed such that an index in an intermediate (unfinished) state may be produced. Further, operations may be focused on tiles required for display and not simply all tiles affected by the change to the clip geometry. A display process may use the spatial index in the intermediate (unfinished) state to produce a substantially real-time display, without waiting for all operations to complete.

Abstract: In one embodiment, techniques are provided for improving perception of representations of subsurface features (e.g., virtual paint markings) in augmented reality. An input image of a terrain surface is accessed. An augmentation stencil image aligned with the input image is created and represented utilizing HSL color space. The input image is converted to the HSL color space. The technique creates and displays an augmented image that, for each pixel that falls outside of the representation subsurface features, has pixel values based on a hue value, a saturation value and a lightness value of the input image and for each pixel that coincides with the representation subsurface features has pixel values based on a hue value and a saturation value of the augmentation stencil image and a lightness value based on the input image.

Abstract: In one embodiment, techniques are used to create a customized 3-D menu for controlling an application (e.g., a CAD application) leveraging the application's native storage format to describe the menu's appearance and function. The menu is defined using elements of a model to which commands are associated within the storage format. The elements, upon later display, operate as menu items, whose selection activates the associated command. The application's (e.g., the CAD application's) visualization capabilities may be used to rotate, zoom, or otherwise manipulated the model in in 3-D space to create a desired appearance. A view of the model is then stored within the storage format to defined an initial appearance of the 3-D menu. Upon later display, the view of the 3-D menu is loaded to provide the defined initial appearance.

Abstract: In one embodiment, techniques are provided for capturing accurate information describing the location of subsurface features (e.g., subsurface utilities such as water pipes, sewer pipes, electrical conduits, etc.) usable in providing an augmented reality view. A set of images is captured with a camera rig coupled to a mobile portion (e.g., the boom) of a piece of heavy construction equipment (e.g., an excavator) being used by workers to conduct an excavation that exposes the subsurface features. The set of images is provided to a structure-from-motion (SfM) photogrammetry that generates a 3D reality mesh. Relative and/or absolute locations of the subsurface features are calculated based on the 3D reality mesh and provided to an augmented reality application executing on an augmented reality device for use in providing an augmented reality view.

Abstract: In one embodiment, techniques are provided for projecting information (e.g., describing subsurface features such as subsurface utilities) onto a surface (e.g., road surface) of the physical environment using pre-captured topography (e.g., determined by structure-from-motion (SfM) photogrammetry) and pre-projection of the information onto the pre-captured topography. The pre-projected information is subsequently combined with a live view of the physical environment to produce an augmented reality view that is displayed to a user.

Abstract: A computer-implemented land planning system is designed to generate at least one conceptual fit solution to a user-defined land development problem. The system employs a computer readable medium and a computer program encoded on the medium. The computer program is operable, when executed on a computer, for electronically creating at least one candidate solution to the land development problem. The candidate solution incorporates a plurality of engineering measurements applicable in development of an undeveloped land site. Existing GIS data is collected for the selected undeveloped land site. A fitness function quantitatively evaluates the candidate solution based on its cost. A heuristic problem-solving strategy manipulates the engineering measurements of the candidate solution to achieve a more quantitatively fit solution to the land development problem. Documentation illustrating the fit solution to the land development problem is delivered to the user.

Abstract: In one embodiment, at an initial time, a capture system records a 3D video stream and an audio stream of a first user interacting with the physical environment. A processing device receives the 3D video stream and the audio stream, isolates the first user from the physical environment, and stores at least a portion of the isolated representation and a portion of the audio stream. At a subsequent time, a second user uses a camera of an augmented reality device to capture a scene. The augmented reality device loads the portion of the isolated representation of the first user and the portion of the audio stream, aligns the 3D video stream with the captured scene, and calculates a pose of the augmented reality device. The augmented reality device then produces an augmented scene, which is displayed to the second user while the portion of the audio stream is played back.