Optimizing a system’s resilience can be challenging, especially when it involves considering both the inherent resilience of a robust design and the active resilience of a health management system to a set of computationally-expensive hazard simulations. While prior work has developed specialized architectures to effectively and efficiently solve combined design and resilience optimization problems, the comparison of these architectures has been limited to a single case study. To further study resilience optimization formulations, this work develops a problem repository which includes previously-developed resilience optimization problems and additional problems presented in this work: a notional system resilience model, a pandemic response model, and a cooling tank hazard prevention model. This work then uses models in the repository at large to understand the characteristics of resilience optimization problems and study the applicability of optimization architectures and decomposition strategies. Based on the comparisons in the repository, applying an optimization architecture effectively requires understanding the alignment and coupling relationships between the design and resilience models, as well as the efficiency characteristics of the algorithms. While alignment determines the necessity of a surrogate of resilience cost in the upper-level design problem, coupling determines the overall applicability of a sequential, alternating, or bilevel structure. Additionally, the application of decomposition strategies is dependent on there being limited interactions between variable sets, which often does not hold when a resilience policy is parameterized in terms of actions to take in hazardous model states rather than specific given scenarios.

Modern digital manufacturing processes, such as additive manufacturing, are cyber-physical in nature and utilize complex, process-specific simulations for both design and manufacturing. Although computational simulations can be used to optimize these complex processes, they can take hours or days — an unreasonable cost for engineering teams leveraging iterative design processes. Hence, more rapid computational methods are necessary in areas where computation time presents a limiting factor. When existing data from historical examples is plentiful and reliable, supervised machine learning can be used to create surrogate models that can be evaluated orders of magnitude more rapidly than comparable finite element approaches. However, for applications that necessitate computationally-intensive simulations, even generating the training data necessary to train a supervised machine learning model can pose a significant barrier. Unsupervised methods, such as physics-informed neural networks, offer a shortcut in cases where training data is scarce or prohibitive. These novel neural networks are trained without the use of potentially expensive labels. Instead, physical principles are encoded directly into the loss function. This method substantially reduces the time required to develop a training dataset, while still achieving the evaluation speed that is typical of supervised machine learning surrogate models. We propose a new method for stochastically training and testing a convolutional physics-informed neural network using the transient 3D heat equation- to model temperature throughout a solid object over time. We demonstrate this approach by applying it to a transient thermal analysis model of the powder bed fusion manufacturing process.

Graph representation learning has revolutionized many artificial intelligence and machine learning tasks in recent years, ranging from combinatorial optimization, drug discovery, recommendation systems, image classification, social network analysis to natural language understanding. This paper shows their efficacy in modeling relationships between products and making predictions for unseen product networks. By representing products as nodes and their relationships as edges of a graph, we show how an inductive graph neural network approach, named GraphSAGE, can efficiently learn continuous representations for nodes and edges. These representations also capture product feature information such as price, brand, and engineering attributes. They are combined with a classification model for predicting the existence of a relationship between any two products. Using a case study of the Chinese car market, we find that our method yields double the F-1 score compared to an Exponential Random Graph Model-based method for predicting the co-consideration relationship between cars. While a vanilla Graph-SAGE requires a partial network to make predictions, we augment it with an ‘adjacency prediction model’ to circumvent this limitation. This enables us to predict product relationships when no neighborhood information is known. Finally, we demonstrate how a permutation-based interpretability analysis can provide insights on how design attributes impact the predictions of relationships between products. Overall, this work provides a systematic method to predict the relationships between products in a complex engineering system.

Coupled Engineered Systems can be characterized by the inherent interactions among design decisions. These interactions define the influence that one design decision exerts over another and require appropriate method to model such interactions. Robust design enables designers to design a product or process that is relatively insensitive to uncertainties. Hence, robust design of coupled engineered systems enables designers to, (i) design engineered systems while accounting for interaction among design decisions and (ii) identifying design decisions that are relatively insensitive to uncertainties. In this paper, an appropriate method to model interactions and identify robust solution is presented. The interacting decisions are categorized into concurrent and hierarchical decisions and are respectively modeled with horizontal and vertical coupling. Based on the strength of interaction between these decisions, two types of interactions are defined, weak and strong interactions. To enable robust decisions in a coupled engineered systems, robustness metrics are defined and included as goals/constraints. The metrics considered in this work are to explore the solution space and manage uncertainty by considering the design of robust systems. The method has been tested on three design examples, that are, (i) design of a fender, (ii) design of a gearbox and (iii) design of a composite structure.

Scientific and engineering problems often require an inexpensive surrogate model to aid understanding and the search for promising designs. While Gaussian processes (GP) stand out as easy-to-use and interpretable learners in surrogate modeling, they have difficulties in accommodating big datasets, qualitative inputs, and multi-type responses obtained from different simulators, which has become a common challenge for a growing number of data-driven design applications. In this paper, we propose a GP model that utilizes latent variables and functions obtained through variational inference to address the aforementioned challenges simultaneously. The method is built upon the latent variable Gaussian process (LVGP) model where qualitative factors are mapped into a continuous latent space to enable GP modeling of mixed-variable datasets. By extending variational inference to LVGP models, the large training dataset is replaced by a small set of inducing points to address the scalability issue. Output response vectors are represented by a linear combination of independent latent functions, forming a flexible kernel structure to handle multi-type responses. Comparative studies demonstrate that the proposed method scales well for large datasets with over 10 4 data points, while outperforming state-of-the-art machine learning methods without requiring much hyperparameter tuning. In addition, an interpretable latent space is obtained to draw insights into the effect of qualitative factors, such as those associated with “building blocks” of architectures and element choices in metamaterial and materials design. Our approach is demonstrated for machine learning of ternary oxide materials and topology optimization of a multiscale compliant mechanism with aperiodic microstructures and multiple materials.

The increasing number of product reviews posted online is a gold mine for designers to know better about the products they develop, by capturing the voice of customers, and to improve these products accordingly. In the meantime, product design and development have an essential role in creating a more sustainable future. With the recent advance of artificial intelligence techniques in the field of natural language processing, this research aims to develop an integrated machine learning solution to obtain sustainable design insights from online product reviews automatically. In this paper, the opportunities and challenges offered by existing frameworks — including Python libraries, packages, as well as state-of-the-art algorithms like BERT — are discussed, illustrated, and positioned along an ad hoc machine learning process. This contribution discusses the opportunities to reach and the challenges to address for building a machine learning pipeline, in order to get insights from product reviews to design more sustainable products, including the five following stages, from the identification of sustainability-related reviews to the interpretation of sustainable design leads: data collection, data formatting, model training, model evaluation, and model deployment. Examples of sustainable design insights that can be produced out of product review mining and processing are given. Finally, promising lines for future research in the field are provided, including case studies putting in parallel standard products with their sustainable alternatives, to compare the features valued by customers and to generate in fine relevant sustainable design leads.

Power systems are designed to meet power demands of the communities with high reliability. Distributed generators (DGs) could play an essential role in improving the power system reliability and resilience. To date, influence of the uncertainty of the DGs to power system reliability has not been well addressed. Consequently, placement of the DGs considering reliability constraints may not be optimally conducted. This paper proposes reliability analysis and design of power systems under time-dependent load uncertainty and wind power generation uncertainty using an efficient uncertainty quantification (UQ) method, i.e., the eigenvector dimension reduction (EDR) method. Furthermore, binary particle swarm optimization (B-PSO) is proposed to address the optimal placement of DGs considering the reliability constraint. Two case studies, including an IEEE 14-bus power system and an IEEE 57-bus power system, are used to demonstrate the effectiveness of the proposed methodology.

In this paper, we present “BIKED,” a dataset comprised of 4500 individually designed bicycle models sourced from hundreds of designers. We expect BIKED to enable a variety of data-driven design applications for bicycles and support the development of data-driven design methods. The dataset is comprised of a variety of design information including assembly images, component images, numerical design parameters, and class labels. In this paper, we first discuss the processing of the dataset, then highlight some prominent research questions that BIKED can help address. Of these questions, we further explore the following in detail: 1) Are there prominent gaps in the current bicycle market and design space? We explore the design space using unsupervised dimensionality reduction methods. 2) How does one identify the class of a bicycle and what factors play a key role in defining it? We address the bicycle classification task by training a multitude of classifiers using different forms of design data and identifying parameters of particular significance through permutation-based interpretability analysis. 3) How does one synthesize new bicycles using different representation methods? We consider numerous machine learning methods to generate new bicycle models as well as interpolate between and extrapolate from existing models using Variational Autoencoders. The dataset is available at http://decode.mit.edu/projects/biked/ along with referenced code.

This paper presents a proposal for a density-adaptive design of bone screws using pedicle screws for spinal fixations as an example. The basis is the analysis and categorization of currently available variants of bone screws, which differ in principle in their thread design because of different application areas (cortical or cancellous bone). These screw variants are investigated in FEA simulations for pullout and bending with regard to occurring stresses. A corresponding simulation model is presented for this purpose. The precise design models for these screws are generated in a CAD system using a self-developed configuration tool. Based on the FEA evaluation, the proposal for a new pedicle screw design, consisting of several thread types merged into each other, is described in detail. By integrating different thread types over the shaft, the respective properties of the bone can thus be optimally utilized.

Additive manufacturing (AM) can produce complex geometrical shapes and multi-material parts that are not possible using typical manufacturing processes. The properties of multi-material AM parts are often unknown. For multi-material parts made using Fused Deposition Modeling (FDM), these properties are driven by the filament. Acquiring the properties of the products or the filament necessitates experiments that can be expensive and time-consuming. Thus, there is a need for simulation-based design tools that can determine the multi-material properties of the filament by exploring the complex process-structure-property (p-s-p) relationship. In this paper, we present a Goal-Oriented Inverse Design (GoID) method to produce feedstock filament for FDM process with specific property goals. Using this method, the designers connects the structure and property in the p-s-p relationship by identifying satisficing material composition for specific property goals. The filament properties identified in the problem are percentage elongation, tensile strength, and Young’s Modulus. The problem is formulated using a generic decision-based design framework, Concept Exploration Framework. The solution space exploration for satisficing solutions is performed using the compromise Decision Support Problem (cDSP). The forward information flow is first established to generate the necessary mathematical relationships between the composition and the property goals. Next, the target property goals of the filament are set. The cDSP is used for solution space exploration to identify satisficing solutions for material composition for the target property goals. While the results are interesting, the focus of our work is to demonstrate, and refine, the goal-oriented, inverse design method for the AM domain.

Heuristics are essential for addressing the complexities of engineering design processes. The goodness of heuristics is context-dependent. Appropriately tailored heuristics can enable designers to find good solutions efficiently, and inappropriate heuristics can result in cognitive biases and inferior design outcomes. While there have been several efforts at understanding which heuristics are used by designers, there is a lack of normative understanding about when different heuristics are suitable. Towards addressing this gap, this paper presents a reinforcement learning-based approach to evaluate the goodness of heuristics for three sub-problems commonly faced by designers: (1) learning the map between the design space and the performance space, (2) acquiring sequential information, and (3) stopping the information acquisition process. Using a multi-armed bandit formulation and simulation studies, we learn the suitable heuristics for these individual sub-problems under different resource constraints and problem complexities. Additionally, we learn the optimal heuristics for the combined problem (i.e., the one composing all three sub-problems), and we compare them to ones learned at the sub-problem level. The results of our simulation study indicate that the proposed reinforcement learning-based approach can be effective for determining the quality of heuristics for different problems, and how the effectiveness of the heuristics changes as a function of the designer’s preference (e.g., performance versus cost), the complexity of the problem, and the resources available.

As enablers of electric vehicles, lithium-ion batteries are drawing much attention for their high energy density and low self-discharging rate. However, “range anxiety” has remained a significant hindrance to its further development. Of the many design objectives, minimizing the charging time and maximizing the cycle life are conflicting design objectives. In the past, enormous efforts have been carried out to resolve the dispute between high charging rates and large capacity losses by either improving the battery design or optimizing the charging/discharging protocols. However, the battery design and the control are usually coupled that integration of the two discipline, or control co-design, may offer improved performances as compared with traditional sequential optimization approaches. In an previous study, we have shown that efficient control co-design is achievable for Lithium-ion batteries through surrogate modeling. In this work, a reliability-based design optimization framework is integrated to guarantee the performances under parametric uncertainties. The challenges, such as simultaneous model update for the dynamic system and excessive computation burden due to optimal control and reliability assessment, are resolved through coupling the first principle model and the empirical models by an adaptive surrogate modeling process. Such a combination captures the multi-scale nature of the battery and allows efficient numerical analysis for the reliability-based co-design (RBCD) problem. A nested co-design approach and a double-loop reliability assessment method were implemented. The results show that the algorithm can shorten the charging time while satisfying the probability constraint on the cycle-life performances under parametric uncertainties.

3D printing enables the production of personalized designs that are desirable in the medical industry for applications including orthopedics, tissue engineering, and personalized nutrition. Currently, the design process relies on trial-and-error approaches, especially for biomaterial development, and there is a need for methodologies to streamline the design process to facilitate automation. Here, we investigate a design methodology for printing foods by mixing novel biomaterial combinations informed by rheological measurements that indicate printability. The process consists of first printing basic designs with chocolate, marzipan, and potato biomaterials known to print consistently. Rheological measurements are collected for these materials and compared to a novel pumpkin biomaterial. The pumpkin had a higher complex modulus and lower mechanical loss tangent than all other biomaterials, therefore motivating the addition of rheological agents to reach more favorable properties. Varied concentrations of corn starch and guar gum were added to the pumpkin to improve printability while altering the nutrient distribution. A 4% inclusion of guar gum provided the most consistent pumpkin prints. A complex 3D object was fabricated with the 4% guar gum pumpkin material, therefore demonstrating the merits in using rheological properties to inform printability for use in design automation routines. The design approach enabled comparisons of relative nutrition and printability trade-offs to demonstrate a proof-of-concept user interface for design automation to facilitate customized food production. Further research to develop a complete design methodology for linking rheological properties to printability would promote consistent prediction of print quality for novel formulations to support design automation, with potential generalizability for diverse biomaterials.

This paper discusses the design of axisymmetric structures with self-supporting features that can be additively manufactured without requiring internal support structures. This is motivated by wire-fed additive manufacturing processes, many of which can fabricate designs with enclosed pores that inherently exist in many axisymmetric structures, such as double walled pressure vessels. Although enclosed pores are possible, features that rise at shallow angles from the build plate typically cannot be fabricated without the use of support structures, which require removal and thus are unfavorable in such applications. In this paper, an overhang constraint is applied to ensure that all designed features rise at a designer-prescribed self-supporting angle to eliminate the need for such support structures. This is achieved by coupling the projection-based overhang constraint approach with topology optimization and axisymmetric finite elements whose stiffness is interpolated using Solid Isotropic Material with Penalization (SIMP). Gradients are computed with the adjoint method and the Method of Moving Asymptotes (MMA) is employed as the gradient-based optimizer. Two numerical examples related to a canonical pressure vessel and an optical mirror support structure are used to demonstrate the approach. Solutions are shown to satisfy minimum feature size requirements and designer-prescribed (process dependent) overhang constraint angles, while providing clear and crisp representations of topology. As observed in past works on overhang constraints, a clear trade-off is illustrated between the magnitude of the overhang constraint angle and the structural performance (mass or stiffness), with more strict requirements producing designs with lower performance.

This paper examines decision making under radical uncertainty in engineering design, that is, engineering decision making in those situations where it is not possible to know the outcomes and/or construct the utility functions and probabilities needed to support rational-human decision making. In these situations, despite being faced with radical uncertainty, engineers do (and must) proceed forward in a linear, clear, and predictable manner. Yet, they may not proceed in a manner that is well described by current engineering design frameworks. Examining the role of decision making in business and other social enterprises, Tuckett and Nikolic [1] have proposed conviction narrative theory (CNT) to describe how rational decision-makers confronted with situations in which insufficient information is available to support traditional decision-making tools use narrative and intuition to reach convincing and actionable decisions. This paper proposes that, in a manner similar to what is described in CNT, narrative and engineering judgment play a critical role in engineering design situations dominated by radical uncertainty. To that end, this paper integrates the traditional rational-human view of decision making as expressed by Hazelrigg in the well-known Decision-Based Design (DBD) framework and CNT as proposed by Tuckett and Nikolic. In the resulting rational, narrative-based design framework, narrative structures are used to describe and develop design alternatives and provide the ideas, beliefs, and preferences needed by the DBD framework. The resulting preferred design is expressed as a narrative and tested using engineering judgement. Specifically, the goal of the design process is expressed as a high-level guiding narrative that fosters the development of design narratives (design alternatives), and ultimately results in a convincing narrative that describes the preferred design. The high-level guiding narrative outlines the event(s), entity(s), preferences, and beliefs needed to support the design. The design narratives are narrative fragments that are nested within the high-level narrative and include the proposed action (idea), the specific challenges that the design faces, and the possible (but not yet verified) outcomes. The convincing narrative is the validated, preferred option that results from the DBD analysis and optimization process and is reviewed using engineering judgement. Following development of the rational, narrative-based design framework, the value of the framework is discussed within the context of practical engineering design.

Human subject experiments are often used in research efforts to understand human behavior in design. However, such research is often time-consuming, expensive, and limited in scope due to the need to experimentally control specific variables. This work develops an initial digital simulation of team-based multidisciplinary design, where the actions of individual team members are simulated using deep learning models trained on historical human design trends. The main benefit of this work is to simulate design session events and interactions without human participants, developing a complimentary method to rapidly perform digital team-based experiments. This research merges the benefits of purely data-driven modeling with minimal assumptions about process, along with the strengths of agent-based modeling in which it is possible to tailor agent behavior. Initial results show that the simulated design team sessions are able to replicate trends and distributions compared to human-based team sessions, but run approximately 21 times faster than equivalent human subject studies. The multi-disciplinary design problem currently simulated is loosely coupled, in the sense that agent behaviors can be modeled in isolation of other agents and yet replicate the behavior of the ensemble. Future work will extend the agents to sense and respond behaviors that can be used to model tightly coupled problems, and truly evaluate team formulations.

This paper focuses on the development of a small scale agricultural robotic platform with advantages over the current agricultural phenotyping platforms that lack the size-scale and sensor resolution needed to study hard to reach under-canopy row crops. The AgBug utilizes a sensor suite consisting of a LiDAR and RGB camera for crop monitoring on a 12″ × 9″ footprint platform. The main challenge for this platform design is not only its compact size and portability, but its ability to navigate and obtain geo-referenced and time-tagged data in the GNSS-denied environment that exists under the crop canopy. LiDAR and RGB sensors typically rely on inputs from GNSS data. Therefore, a new approach was developed here fusing direct feedback from the RGB camera and a visual-inertial tracking camera supplying robot odometry data. Indoor and outdoor tests were conducted to demonstrate the AgBug’s ability for in-row and under canopy crop monitoring along with the efficacy of sensor fusion approach.

Origami-inspired structures and materials have shown remarkable properties and performances originating from the intricate geometries of folding. Origami folding could be a dynamic process and origami structures could possess rich dynamic characteristics under external excitations. However, the current state of dynamics of origami has mostly focused on the dynamics of a single cell. This research has performed numerical simulations on multi-stable dual-cell series Miura-Ori structures with different types of inter-cell connections based on a dynamic model that does not neglect in-plane mass. We introduce a concept of equivalent constraint stiffness k* to distinguish different types of inter-cell connections. Results of numerical simulations reveal the multi-stable dual-cell structure will exhibit a variety of complex nonlinear dynamic responses with the increasing of connection stiffness because of the deeper energy well it has. The connection stiffness has a strong effect on the steady-state dynamic responses under different excitation amplitudes and a variety of initial conditions. This effect makes us able to adjust the dynamic behaviors of dual-cell series Miura-Ori structure to our needs in a complex environment. Furthermore, the results of this research could provide us a theoretical basis for the dynamics of origami folding and serve as guidelines for designing dynamic applications of origami metastructures and metamaterials.

Auxetic material using corner-connected kinematic tiles has been applied to different kinematic designs. However, existing works rely on the connectivity of regular polygonal tilings because of the overconstraining nature of kinematic tiling. This study proposes a new family of auxetic structures based on non-regular and aperiodic rhombic tiling inspired by the Tokyo 2020 Emblems. We convert emblem-like patterns on rhombic tilings into kinematic structures by regarding the rectangular figure as voids and the region between rectangles as rigid bodies. Due to the geometric properties of rhombic tiling, the structure forms a one-degree-of-freedom planar mechanism with a constant Poisson’s ratio of −1. The large combinatorial family of rhombic tilings provides design variations of kinematic structures with non-regular topology. Furthermore, we show a kirigami-based method for fabricating the structure as a compliant mechanism. This connection between math and art potentially broadens the range of architected materials based on folding, kirigami, and tessellation.

This article presents a new hand architecture with three under-actuated fingers. Each finger performs spatial movements to achieve more complex and varied grasping than the existing planar-movement fingers. The purpose of this hand is to grasp complex-shaped workpieces as they leave the machining centres. Among the taxonomy of grips, cylindrical and spherical grips are often used to grasp heavy objects. A combination of these two modes makes it possible to capture most of the workpieces machined with 5-axis machines. However, the change in grasping mode requires the fingers to reconfigure themselves to perform spatial movements. This solution requires the addition of two or three actuators to change the position of the fingers and requires sensors to recognize the shape of the workpiece and determine the type of grasp to be used. This article proposes to extend the notion of under-actuated fingers to spatial movements. After a presentation of the kinematics of the fingers, the problem of stability is discussed as well as the transmission of forces in this mechanism. The complete approach for calculating the stability conditions is presented from the study of Jacobian force transmission matrices. CAD representations of the hand and its behavior in spherical and cylindrical grips are presented.