This paper addresses the challenge of supporting dynamic mechanical design processes in distributed, multi-user environments. Although mechanical CAD is a mature technology, assessing the functional and manufacturing impact of design changes remains a labour-intensive manual task. To address this issue, the authors introduce the concept of “3D CAD Change Detection and Classification” (3D-CDC) and present a novel dataset, “3D-GCD” (3D Geometric Change Dataset), consisting of pairs of 3D CAD components with a variety of technically plausible changes. The dataset represents 3D components in multiple 2D image formats (e.g., grayscale and comparative heatmaps) with renderings generated from single or multiple viewpoints. This dataset is used to train four deep learning architectures—Siamese Networks with convolutional encoders, transformer encoders, 3D Convolutional Networks (3D Conv), and Convolutional Recurrent Neural Networks (CRNN)—to classify the engineering significance of changes to a component’s design. The results indicate that while three architectures (Siamese Networks, transformer encoders, and CRNN) exhibit similar accuracy, the CRNN demonstrates greater consistency and efficiency, requiring fewer trainable parameters for 3D-CDC tasks.

It has been widely reported that the reuse of previously created components, or features, in new engineering designs will improve the efficiency of a company’s product development process. Although the reuse of engineering components has established metrics and methodologies, the reuse of specific design features (e.g. stiffening ribs, hole patterns or lubrication grooves, etc.) has received less attention in the literature. Typically, researchers have reported approaches to partial design reuse that identify patterns predominately in terms of geometrically similar shapes (i.e. a set of features) whose elements are adjacent, cohesive, and decoupled from the overall form of a component. In contrast, this paper defines a common design structure (CDS) as collections of frequently occurring features (e.g. holes) with common parametric values (e.g. diameters) in a CAD database (irrespective of their locations or spatial connectivity between other features on a component). By exploiting the established data-mining technology of association rules and item-sets the authors show how CDSs can be efficiently computed for hundreds of 3D CAD models. A case study, with hole data extracted from a publicly available dataset of hydraulic valves, is presented to illustrate how item-sets associated with CDS can be computed and used to support predictive design by identifying potentially ‘substitutable features’ during an interactive design process. This is done using a combination of association rules and geometric compatibility checks to ensure the system’s suggestion are implementable. The use of the Kullback–Leibler divergence to assess the degree of similarity between components is identified as a crucial step in the process of identifying the “best” suggestions. The results illustrate how the prototype implementation successfully mines the CDSs and identifies substitutable hole features in a dataset of industrial valve designs.

In recent years, the “power of the crowd” has been repeatedly demonstrated and various Internet platforms have been used to applied collaborative intelligence to areas that range from open innovation to conceptual design. However, crowdsourcing applications in the fields of design research and creative innovation have been much slower to emerge. Although reported work has been suggested that designs can be improved by applying the crowdsourced design methodology, such as Crowdsourced Design Framework, its evaluation process is still required to be analyzed. In this paper, the statistical analysis methods, normal probability plots, are used to validate the crowdsourced design evaluation results in the crowdsourced design case study of living room layout design task. The discussion about the ranking distributions and results suggested that in the living room layout design task, standard distribution trendlines and r2results proves the effects of the ranking evaluation method used in the case study. The distribution determines that the evaluation process and results matched a normal distribution. As the contribution to knowledge, this paper applies the statistical analysis method to measure and approve the crowdsourced design evaluation method and its results.

In recent years, the “power of the crowd” has been repeatedly demonstrated and various Internet platforms have been used to applied collaborative intelligence to areas that range from open innovation to image analysis. However, crowdsourcing applications in the fields of design research and creative innovation have been much slower to emerge. So, although there have been reports of systems and researchers using Internet crowdsourcing to carry out generative design, there are still many gaps in knowledge about the capability and limitations of the technology. For example on crowdsourcing platforms, like Amazon's Mechanical Turk, the relationship between remuneration and the final quality of designs has not been established, so it is unclear how much payment should be offered to ensure a particular standard of result. To investigate the relationship between the crowd's remuneration and the quality of their innovation, this paper reports how payment for a 2D interior design task (living room layout) was systematically varied and the quality of the output assessed by a ranking process that was also crowdsourced. Information about individual Mturk workers who participated in the study was also collected. The results suggest that while average design quality only slowly increases, the quality of the “best” design generated by the crowd improved dramatically with payment levels. In other words, increasing monetary rewards does not improve the average creativity of design but rather it increases the chance of an excellent solution being generated by an individual in the crowd.

The ability to interpret and reason about shapes is a peculiarly human capability that has proven difficult to reproduce algorithmically. So despite the fact that geometric modeling technology has made significant advances in the representation, display and modification of shapes, there have only been incremental advances in geometric reasoning. For example, although today's CAD systems can confidently identify isolated cylindrical holes, they struggle with more ambiguous tasks such as the identification of partial symmetries or similarities in arbitrary geometries. Even well defined problems such as 2D shape nesting or 3D packing generally resist elegant solution and rely instead on brute force explorations of a subset of the many possible solutions.

Near net shape (NNS) manufacturing refers to the production of products that require a finishing operation of some kind. NNS manufacturing is important because it enables a significant reduction in: machining work, raw material usage, production time, and energy consumption. This paper presents an integrated system for the production of near net shape components based on the Octree decomposition of 3-D models. The Octree representation is used to automatically decompose and approximate the 3-D models, and to generate the robot instructions required to create assemblies of blocks secured by adhesive. Not only is the system capable of producing shapes of variable precision and complexity (including overhanging or reentrant shapes) from a variety of materials, but it also requires no production tooling (e.g., molds, dies, jigs, or fixtures). This paper details how a number of well-known Octree algorithms for subdivision, neighbor findings, and tree traversal have been modified to support this novel application. This paper ends by reporting the construction of two mechanical components in the prototype cell, and discussing the overall feasibility of the system.

This paper investigates the feasibility of predicting the attenuation of AE signals travelling within a complex solid body. Such AE can occur due to external stimulation (e.g. impact) or internal events (i.e. crack propagation). The attenuation of these signals is affected not only by material properties but also by the geometry of the object. For example, wave propagation on a typical engine block is complex because of its intricate shape with variations and discontinuities in thickness and surface curvature. In contrast to much of the reported work in computer graphics (CG) literature that models the transmission of sound in rooms and buildings, this paper reports the development of a ray firing procedure to model the transmission of rays both across the surface and through the interior of a complex mechanical solid. The attenuation rate is proportional to the AE path length so the challenge of modelling the physical phenomena involves calculating the path lengths between the sensor and the source. The results of the computational simulation are compared with those obtained by experiment and found to be in good agreement.

Numerous applications in mechanical CAD/CAM need robust algorithms for the identification of protrusion and depression features (DP-features) on geometric models with free-form (B-Spline) surfaces. This paper reports a partitioning algorithm that first identifies the boundary edges of DP-features and then creates a surface patch to cover the depressions or isolate the protrusions. The novelty of the method lies in the use of tangent continuity between edge segments to identify DP-feature boundaries that cross multiple faces and geometries.

This work presents a feature recognition (FR) technique that separates protrusions bounded by freeform surface geometry. The method can also generate complementary male/female assembly features at the interface between components. A heuristic that uses edge curvature continuity is applied to direct the search and the construction of protrusion boundaries. The result is represented as a cellular model. The objective is to manufacture complex prototype assemblies on multiaxis machining centres. After describing the algorithm and presenting some examples of its application the paper concludes by discussing the approach's limitations.

Motivated by the needs of mould and die manufacturers, this paper presents a novel approach to recognizing shape features on geometric models composed of both simple and complex ruled surfaces. The algorithm described uses a network of adjacent 2D-laminae (i.e., bounded surfaces) derived from a component's CAD model to both locate and create generic protrusion and depression feature volumes. The approach also enables the automatic generation of alternative feature descriptions and requires no predefined feature libraries.

This paper presents an overview of recent developments by the authors in geometric feature recognition using object-oriented graph-based methodologies. Graph-based feature recognition is reviewed in the context of developed graph and tree class hierarchies and the FeatureFinder software program.