Designing physical artifacts that serve a purpose---such as tools and other functional structures---is central to engineering as well as everyday human behavior. Though automating design using machine learning has tremendous promise, existing methods are often limited by the task-dependent distributions they were exposed to during training.

The widespread use of neural surrogates in automotive aerodynamics, enabled by datasets such as DrivAerML and DrivAerNet++, has primarily focused on bluff-body flows with large wakes. Extending these methods to aerospace, particularly in the transonic regime, remains challenging due to the high level of non-linearity of compressible flows and 3D effects such as wingtip vortices. Existing aerospace datasets predominantly focus on 2D airfoils, neglecting these critical 3D phenomena. To address this gap, we present a new dataset of CFD simulations for 3D wings in the transonic regime. The dataset comprises volumetric and surface-level fields for around $30,000$ samples with unique geometry and inflow conditions. This allows computation of lift and drag coefficients, providing a foundation for data-driven aerodynamic optimization of the drag-lift Pareto front. We evaluate several state-of-the-art neural surrogates on our dataset, including Transolver and AB-UPT, focusing on their out-of-distribution (OOD) generalization over geometry and inflow variations. AB-UPT demonstrates strong performance for transonic flowfields and reproduces physically consistent drag-lift Pareto fronts even for unseen wing configurations. Our results demonstrate that AB-UPT can approximate drag-lift Pareto fronts for unseen geometries, highlighting its potential as an efficient and effective tool for rapid aerodynamic design exploration. To facilitate future research, we open-source our dataset at https://huggingface.co/datasets/EmmiAI/Emmi-Wing.

Tendon-driven mechanisms are useful from the perspectives of variable stiffness, redundant actuation, and lightweight design, and they are widely used, particularly in hands, wrists, and waists of robots. The design of these wire arrangements has traditionally been done empirically, but it becomes extremely challenging when dealing with complex structures. Various studies have attempted to optimize wire arrangement, but many of them have oversimplified the problem by imposing conditions such as restricting movements to a 2D plane, keeping the moment arm constant, or neglecting wire crossings. Therefore, this study proposes a three-dimensional wire arrangement optimization that takes wire crossings into account. We explore wire arrangements through a multi-objective black-box optimization method that ensures wires do not cross while providing sufficient joint torque along a defined target trajectory. For a 3D link structure, we optimize the wire arrangement under various conditions, demonstrate its effectiveness, and discuss the obtained design solutions.

Designing physical artifacts that serve a purpose---such as tools and other functional structures---is central to engineering as well as everyday human behavior. Though automating design using machine learning has tremendous promise, existing methods are often limited by the task-dependent distributions they were exposed to during training. This constitutes a simple, fast, and reusable approach that solves high-dimensional problems with complex physical dynamics, including designing surfaces and tools to manipulate fluid flows and optimizing the shape of an airfoil to minimize drag. This framework produces high-quality designs by propagating gradients through trajectories of hundreds of steps, even when using models that were pre-trained for single-step predictions on data substantially different from the design tasks. In our fluid manipulation tasks, the resulting designs outperformed those found by sampling-based optimization techniques.

Multi-fidelity Reinforcement Learning (RL) frameworks efficiently utilize computational resources by integrating analysis models of varying accuracy and costs. The prevailing methodologies, characterized by transfer learning, human-inspired strategies, control variate techniques, and adaptive sampling, predominantly depend on a structured hierarchy of models. However, this reliance on a model hierarchy can exacerbate variance in policy learning when the underlying models exhibit heterogeneous error distributions across the design space. To address this challenge, this work proposes a novel adaptive multi-fidelity RL framework, in which multiple heterogeneous, non-hierarchical low-fidelity models are dynamically leveraged alongside a high-fidelity model to efficiently learn a high-fidelity policy. Specifically, low-fidelity policies and their experience data are adaptively used for efficient targeted learning, guided by their alignment with the high-fidelity policy. The effectiveness of the approach is demonstrated in an octocopter design optimization problem, utilizing two low-fidelity models alongside a high-fidelity simulator. The results demonstrate that the proposed approach substantially reduces variance in policy learning, leading to improved convergence and consistent high-quality solutions relative to traditional hierarchical multi-fidelity RL methods. Moreover, the framework eliminates the need for manually tuning model usage schedules, which can otherwise introduce significant computational overhead. This positions the framework as an effective variance-reduction strategy for multi-fidelity RL, while also mitigating the computational and operational burden of manual fidelity scheduling.

In conceptual aircraft design, the problem of shaping a typical wing is usually decomposed into two parts: planform design and airfoil design. The latter, which is the focus of this work, is a multidisciplinary design problem that requires consideration of a variety of aerodynamic, structural, and manufacturing objectives and constraints. A non-exhaustive list of major considerations could include: Profile drag across the expected operating range of the airfoil (spanning lift coefficients, Reynolds numbers, and Mach numbers), including adequate off-design performance [1]; Pitching moment and aft-camber coefficients, which can drive tail sizing (modifying trim drag), affect divergence speed; Hinge moments and control effectiveness of any control surfaces, which drive actuator design and weight; Stall behavior, which can affect handling qualities and safety; Thickness at various points, in order to accommodate fuel volume and required structural members to resist failure (e.g., by bending, buckling, divergence, flutter, or control reversal);[2] Sensitivity to boundary layer performance, freestream turbulence, and trips, all of which impose constraints on surface finish, cleanliness, and manufacturing tolerances [3-5]; Peak suction pressures, which affect the critical Mach number in transonic applications or cavitation in hydrodynamic applications; Shock stability and buffet considerations in transonic applications; Manufacturability, which might include flat-bottom airfoil sections, strictly-convex airfoil shapes (e.g., to

Musculoskeletal humanoids have various biomimetic advantages, and the redundant muscle arrangement allowing for variable stiffness control is one of the most important. In this study, we focus on one feature of the redundancy, which enables the humanoid to keep moving even if one of its muscles breaks, an advantage that has not been dealt with in many studies. In order to make the most of this advantage, the design of muscle arrangement is optimized by considering the maximization of minimum available torque that can be exerted when one muscle breaks. This method is applied to the elbow of a musculoskeletal humanoid Musashi with simulations, the design policy is extracted from the optimization results, and its effectiveness is confirmed with the actual robot.

The need to explore and/or optimize expensive simulators with many qualitative factors arises in broad scientific and engineering problems. Our motivating application lies in path planning - the exploration of feasible paths for navigation, which plays an important role in robotics, surgical planning and assembly planning. Here, the feasibility of a path is evaluated via expensive virtual experiments, and its parameter space is typically discrete and high-dimensional. A carefully selected experimental design is thus essential for timely decision-making. We propose here a novel framework, called QuIP, for experimental design of Qualitative factors via Integer Programming under a Gaussian process surrogate model with an exchangeable covariance function. For initial design, we show that its asymptotic D-optimal design can be formulated as a variant of the well-known assignment problem in operations research, which can be efficiently solved to global optimality using state-of-the-art integer programming solvers. For sequential design (specifically, for active learning or black-box optimization), we show that its design criterion can similarly be formulated as an assignment problem, thus enabling efficient and reliable optimization with existing solvers. We then demonstrate the effectiveness of QuIP over existing methods in a suite of path planning experiments and an application to rover trajectory optimization.

The design of soft robots is still commonly driven by manual trial-and-error approaches, requiring the manufacturing of multiple physical prototypes, which in the end, is time-consuming and requires significant expertise. To reduce the number of manual interventions in this process, topology optimization can be used to assist the design process. The design is then guided by simulations and numerous prototypes can be tested in simulation rather than being evaluated through laborious experiments. To implement this simulation-driven design process, the possible design space of a slender soft pneumatic actuator is generalized to the design of the circular cross-section. We perform a black-box topology optimization using genetic algorithms to obtain a cross-sectional design of a soft pneumatic actuator that is capable of reaching a target workspace defined by the end-effector positions at different pressure values. This design method is evaluated for three different case studies and target workspaces, which were either randomly generated or specified by the operator of the design assistant. The black-box topology optimization based on genetic algorithms proves to be capable of finding good designs under given plausible target workspaces. We considered a simplified simulation model to verify the efficacy of the employed method. An experimental validation has not yet been performed. It can be concluded that the employed black-box topology optimization can assist in the design process for slender soft pneumatic actuators. It supports at searching for possible design prototypes that reach points specified by corresponding actuation pressures. This helps reduce the trial-and-error driven iterative manual design process and enables the operator to focus on prototypes that already offer a good viable solution.

Designing physical artifacts that serve a purpose---such as tools and other functional structures---is central to engineering as well as everyday human behavior. Though automating design using machine learning has tremendous promise, existing methods are often limited by the task-dependent distributions they were exposed to during training. This constitutes a simple, fast, and reusable approach that solves high-dimensional problems with complex physical dynamics, including designing surfaces and tools to manipulate fluid flows and optimizing the shape of an airfoil to minimize drag. This framework produces high-quality designs by propagating gradients through trajectories of hundreds of steps, even when using models that were pre-trained for single-step predictions on data substantially different from the design tasks. In our fluid manipulation tasks, the resulting designs outperformed those found by sampling-based optimization techniques.