Wavelets based quantitative design of B-spline curves

In order to achieve a better control of deformations and a more accurate modeling, this paper first introduces a new concept called “reference index of modeling (RIOM)” by wavelet technology. Next, according to RIOM, a quantitative modeling algorithm of B-spline curves is presented. Compared with the traditional modeling methods, the present algorithm employs the RIOM as the objective function and the objective shape can be evaluated quantitatively by the value of the RIOM based objective function. Meanwhile, the new algorithm can preserve the overall shapes during the deformation. Several examples are given to demonstrate the effectiveness of this approach.     

Fast variational design of multiresolution curves and surfaces with B-spline wavelets

Multiresolution modeling provides a powerful tool for complex shape editing. To achieve a better control of deformations and a more intuitive interface, variational principles have been used in such multiresolution models. However, when handling multiresolution constraints, the existing methods often result in solving large optimization systems. Hence, the computational time may become too excessive to satisfy the requirements for interactive design in CAD. In this paper, we present a fast approach for interactive variational design of multiresolution models. By converting all constraints at different levels to a target level, the optimization problem is formulated and solved at the lower level. Thus, the unknown coefficients of the optimization system are significantly reduced. This improves the efficiency of variational design. Meanwhile, to avoid smoothing out the details of the shape in variational modeling, we optimize the change in the deformation energy instead of the total energy of the deformed shape. Several examples and the experimental results are given to demonstrate the effectiveness and efficiency of this approach.     

Modifying the shape of B-spline surface with geometric constraints

The B-spline surface is one of the most commonly used parametric surface in computer aided geometric design and computer graphics. To develop more convenient techniques for designing and modifying B-spline surface is an important problem. A new method for the shape modification of B-spline surface with geometric constraints is presented in this paper. The deformation energy of the physically based B-spline surface is minimized based on finite element method, while geometric constraints including point, curve and boundary continuity constraints can be imposed to control the modified shape. By setting the external force zero, the surface is modified by the constraints solely. This lead to a simplified linear system to be solved and to eliminate the need of internal energy that could convert the iteration process of finite element function to a faster change for control points vector, such that the modified surface satisfies the given constraints. Practical examples are also given.     

Generative design in architecture using an expert system

The mathematician-architect Christopher Alexander has devised a theory of objective architectural design. He believes that all architectural forms can be described as interacting patterns, all possible relationships of which are governed by generative rules. These form a ‘pattern language’ capable of generating forms appropriate for a given environmental context. The complexity of interaction among these rules leads to difficulties in their representation by conventional methods. This paper presents a Prolog-based expert system which implements Alexander's design methodology to produce perspective views of partially and fully differentiated 3-dimensional architectural forms.     

Interactive surface representation system using a B-spline formulation with interpolation capability

An interactive surface representation system is described which uses a parametric uniform bicubic B-spline formulation which can describe a surface initially defined to interpolate a specified network of points.     

Interactive design exploration for constrained meshes

In architectural design, surface shapes are commonly subject to geometric constraints imposed by material, fabrication or assembly. Rationalization algorithms can convert a freeform design into a form feasible for production, but often require design modifications that might not comply with the design intent. In addition, they only offer limited support for exploring alternative feasible shapes, due to the high complexity of the optimization algorithm.We address these shortcomings and present a computational framework for interactive shape exploration of discrete geometric structures in the context of freeform architectural design. Our method is formulated as a mesh optimization subject to shape constraints. Our formulation can enforce soft constraints and hard constraints at the same time, and handles equality constraints and inequality constraints in a unified way. We propose a novel numerical solver that splits the optimization into a sequence of simple subproblems that can be solved efficiently and accurately.Based on this algorithm, we develop a system that allows the user to explore designs satisfying geometric constraints. Our system offers full control over the exploration process, by providing direct access to the specification of the design space. At the same time, the complexity of the underlying optimization is hidden from the user, who communicates with the system through intuitive interfaces.     

Generative methods and the design process: A design tool for conceptual settlement planning

This paper implores the possible intervention of computers in the generative (concept) stage of settlement planning. The objective was to capture the complexity and character of naturally grown fishing settlements through simple rules and incorporate them in the process of design. A design tool was developed for this purpose. This design tool used a generative evolutionary design technique, which is based on multidisciplinary methods. Facets of designing addressed in this research are:

• •allocation of each design element's space and geometry,
• •defining the rules, constraints and relationships governing the elements of design,
• •the purposeful search for better alternative solutions,
• •quantitative evaluation of the solution based on spatial, comfort, complexity criterions to ensure the needed complexity, usability in the solutions.

Generative design methods such as geometric optimization, shape grammars and genetic algorithms have been combined for achieving the above purposes.The allocation of space has been achieved by geometric optimization techniques, which allocate spaces by proliferation of a simple shape unit. This research conducts an analysis of various naturally grown fishing settlements and identifies the features that would be essential to recreate such an environment. Features such as the essential elements, their relationships, hierarchy, and order in the settlement pattern, which resulted due to the occupational and cultural demands of the fisher folk, are analysed. The random but ordered growth of the settlement is captured as rules and relations. These rules propel and guide the whole process of design generation.These rules and certain constraints, restrictions control the random arrangement of the shape units. This research limits itself to conducting exhaustive search in the prescribed solution search space defined a priori by the rules and relationships. This search within a bounded space can be compared to the purposeful, constrained decision making process involved in designing.The generated solutions use the evolutionary concept of genetic algorithms to deduce solutions within the predefined design solution search space. Simple evolutionary concepts such as reproduction, crossover and mutation aid this search process. These concepts transform by swapping/interchanging the genetic properties (the constituent data/material making up the solution) of two generated solutions to produce alternate solutions. Thus the genetic algorithm finds a series of new solutions. With such a tool in hand various possibilities of design solutions could be analysed and compared. A thorough search of possible solutions ensures a deeper probe essential for a good design.The spatial quality, comfort quality of the solutions are compared and graded (fitness value) against the standard stipulations. These parameters look at the solution in the context of the whole and not as parts and most of these parameters could be improved only at the expense of another. The tool is able to produce multiple equally good solutions to the same problem, possibly with one candidate solution optimizing one parameter and another candidate optimizing a different one. The final choice of the suitable solution is made based on the user's preferences and objectives.The tool is tested for an existing fishing settlement. This was done to check for its credibility and to see if better alternatives evolved. The existing settlement is analysed based on the evaluation parameters used in the tool and compared with the generated solutions. The results of the tool has proved that simple rules when applied recursively within constraints would provide solutions that are unpredictable and also would resonate the qualities of the knowledge from which the rules were distilled from. The complex whole generated has often exhibited emergent properties and thus opens up new avenues of thinking.     

Generative learning of visual concepts using multiobjective genetic programming

This paper introduces a novel method of visual learning based on genetic programming, which evolves a population of individuals (image analysis programs) that process attributed visual primitives derived from raw raster images. The goal is to evolve an image analysis program that correctly recognizes the training concept (shape). The approach uses generative evaluation scheme: individuals are rewarded for reproducing the shape of the object being recognized using graphical primitives and elementary background knowledge encoded in predefined operators. Evolutionary run is driven by a multiobjective fitness function to prevent premature convergence and enable effective exploration of the space of solutions. We present the method in detail and verify it experimentally on the task of learning two visual concepts from examples.     

Feature design on deformation of freeform surfaces using range control

This research is to develop a freeform surface design method which uses a feature-based design approach to process surface deformation. Using this method, called range control, a user modifies or changes the position of a characteristic control point of the control polyhedron for a Bspline surface, and regional surface deformation is then carried out through automatic displacements of a group of control points surrounding the characteristic point. The position changes of the group of control points are the interpolations of the maximum movement of the characteristic point related to a feature dimension and zero movements of some fixed points on the designated boundary. In this range control surface deformation method, three interpolating approaches are proposed to smooth out the changes between the characteristic vertex and fixed vertices. they are based on the linear interpolation under the consideration of index, distance and angle distributions. The ideas are implemented on a feature-based designing system for shoe lasts. Design examples are given to show that the range control method is simple, fast and practically applicable for shape deformation of freeform surfaces.     

DMS2015-33: Generative interface structure design for supporting existing objects

Increasing availability of high quality 3D printing devices and services now enable ordinary people to create, edit and repair products for their custom needs. However, an effective use of current 3D modeling and design software is still a challenge for most novice users. In this work, we introduce a new computational method to automatically generate an organic interface structure that allows existing objects to be statically supported within a prescribed physical environment. Taking the digital model of the environment and a set of points that the generated structure should touch as an input, our biologically inspired growth algorithm automatically produces a support structure that when physically fabricated helps keep the target object in the desired position and orientation. The proposed growth algorithm uses an attractor based form generation process based on the space colonization algorithm and introduces a novel target attractor concept. Moreover, obstacle avoidance, symmetrical growth, smoothing and sketch modification techniques have been developed to adapt the nature inspired growth algorithm into a design tool that is interactive with the design space. We present the details of our technique and illustrate its use on a collection of examples from different categories.     

Deep convolutional learning for general early design stage prediction models

Designers rely on performance predictions to direct the design toward appropriate requirements. Machine learning (ML) models exhibit the potential for rapid and accurate predictions. Developing conventional ML models that can be generalized well in unseen design cases requires an effective feature engineering and selection. Identifying generalizable features calls for good domain knowledge by the ML model developer. Therefore, developing ML models for all design performance parameters with conventional ML will be a time-consuming and expensive process. Automation in terms of feature engineering and selection will accelerate the use of ML models in design.Deep learning models extract features from data, which aid in model generalization. In this study, we (1) evaluate the deep learning model’s capability to predict the heating and cooling demand on unseen design cases and (2) obtain an understanding of extracted features. Results indicate that deep learning model generalization is similar to or better than that of a simple neural network with appropriate features. The reason for the satisfactory generalization using the deep learning model is its ability to identify similar design options within the data distribution. The results also indicate that deep learning models can filter out irrelevant features, reducing the need for feature selection.     

Behavioral,data-driven,agent-based evacuation simulation for building safety design using machine learning and discrete choice models

To improve occupant safety during building emergencies, evacuation simulations have been widely used for building safety design. Since occupant behavior is a determining factor for the outcome of building emergencies, accurately capturing how occupants make decisions and integrating occupants’ decision-making processes in evacuation simulations is important. In this study, based on the results of fire evacuation experiments in a virtual metro station, how different social (crowd flow) and environmental (visual access and vertical movement) factors would affect individuals’ wayfinding behavior was predicted using machine learning and discrete choice models. The trained models were further employed in agent-based evacuation simulations to examine crowd evacuation performance under different building design scenarios. Both the machine learning and discrete choice models could accurately predict individuals’ directional choices during emergency evacuations. Different building attributes could collectively influence occupant behavior, leading to distinct exit choices and evacuation times. While both the trained machine learning and discrete choice models generated similar results, the discrete choice model had better interpretability. Moreover, by comparing the trained models in this study with a model developed in a prior study, it was found that agents had significantly distinct responses to different building designs. Critical factors (e.g., type and size of buildings, occupants’ familiarity with the building) for the applicability of evacuation models were identified. Furthermore, recommendations were provided for future research that aims at employing evacuation simulations for building design evaluation and optimization.     

Interactive design of 3D models with geometric constraints

In this paper, an interactive graphical approach for the design of parameterized part-hierarchies is presented. Primitive solids can be grouped into compound objects, and multiple instances of a compound object can be used in further designs. Geometric relations between primitives and instances are specified by geometric constraints between their local coordinate systems. The user can specify and edit a model by direct manipulation on a perspective or parallel projection with a mouse, whereas a procedural model representations is automatically generated via visual programming. The obtained twoview approach offers two concurrent interface styles to the end-user and enables the combination of an intuitive direct manipulation interface with the expressiveness of a procedural modeling language.     

Generative Modeling of Autonomous Robots and their Environments using Reservoir Computing

Autonomous mobile robots form an important research topic in the field of robotics due to their near-term applicability in the real world as domestic service robots. These robots must be designed in an efficient way using training sequences. They need to be aware of their position in the environment and also need to create models of it for deliberative planning. These tasks have to be performed using a limited number of sensors with low accuracy, as well as with a restricted amount of computational power. In this contribution we show that the recently emerged paradigm of Reservoir Computing (RC) is very well suited to solve all of the above mentioned problems, namely learning by example, robot localization, map and path generation. Reservoir Computing is a technique which enables a system to learn any time-invariant filter of the input by training a simple linear regressor that acts on the states of a high-dimensional but random dynamic system excited by the inputs. In addition, RC is a simple technique featuring ease of training, and low computational and memory demands.     

Interactive design of fair hull surfaces using B-splines

Two related techniques for the interactive computer aided design of lines and molded surfaces are presented. The first technique is aimed at lines generation from hull form parameters. The second is aimed at fairing a mathematical surface that is based upon the lines. Both schemes employ B-splines to represent waterlines and stations in the lines drawing. The lines of a modern ship which have been generated by the first method are shown. A new indicator of surface fairness called Euler's net is illustrated. Indications are given of the expected future developments in these continuiing efforts.     

Data-driven generative design for mass customization: A case study

Generative design provides a promising algorithmic solution for mass customization of products, improving both product variety and design efficiency. However, the current designer-driven generative design formulates the automated program in a manual manner and has insufficient ability to satisfy the diverse needs of individuals. In this work, we propose a data-driven generative design framework by integrating multiple types of data to improve the automation level and performance of detail design to boost design efficiency and improve user satisfaction. A computational workflow including automated shape synthesis and structure design methods is established. More specifically, existing designs selected based on user preferences are utilized in the shape synthesis for creating generative models. For structural design, user-product interaction data gathered by sensors are used as inputs for controlling the spatial distributions of heterogeneous lattice structures. Finally, the proposed concept and workflow are demonstrated with a bike saddle design with a personalized shape and inner structures to be manufactured with additive manufacturing.     

Automated defect inspection system for metal surfaces based on deep learning and data augmentation

Recent efforts to create a smart factory have inspired research that analyzes process data collected from Internet of Things (IOT) sensors, to predict product quality in real time. This requires an automatic defect inspection system that quantifies product quality data by detecting and classifying defects in real time. In this study, we propose a vision-based defect inspection system to inspect metal surface defects. In recent years, deep convolutional neural networks (DCNNs) have been used in many manufacturing industries and have demonstrated the excellent performance as a defect classification method. A sufficient amount of training data must be acquired, to ensure high performance using a DCNN. However, owing to the nature of the metal manufacturing industry, it is difficult to obtain enough data because some defects occur rarely. Owing to this imbalanced data problem, the generalization performance of the DCNN-based classification algorithm is lowered. In this study, we propose a new convolutional variational autoencoder (CVAE) and deep CNN-based defect classification algorithm to solve this problem. The CVAE-based data generation technology generates sufficient defect data to train the classification model. A conditional CVAE (CCVAE) is proposed to generate images for each defect type in a single CVAE model. We also propose a classifier based on a DCNN with high generalization performance using data generated from the CCVAE. In order to verify the performance of the proposed method, we performed experiments using defect images obtained from an actual metal production line. The results showed that the proposed method exhibited an excellent performance.     

A deep learning approach to the classification of 3D CAD models简

Model classification is essential to the management and reuse of 3D CAD models. Manual model classification is laborious and error prone. At the same time, the automatic classification methods are scarce due to the intrinsic complexity of 3D CAD models. In this paper, we propose an automatic 3D CAD model classification approach based on deep neural networks. According to prior knowledge of the CAD domain, features are selected and extracted from 3D CAD models first, and then pre-processed as high dimensional input vectors for category recognition. By analogy with the thinking process of engineers, a deep neural network classifier for 3D CAD models is constructed with the aid of deep learning techniques. To obtain an optimal solution, multiple strategies are appropriately chosen and applied in the training phase, which makes our classifier achieve better per-formance. We demonstrate the efficiency and effectiveness of our approach through experiments on 3D CAD model datasets.