Computer-aided design (CAD) is vital to modern manufacturing, yet model creation remains labor-intensive and expertise-heavy. To enable non-experts to translate intuitive design intent into manufacturable artifacts, recent large language models-based text-to-CAD efforts focus on command sequences or script-based formats like CadQuery. However, these formats are kernel-dependent and lack universality for manufacturing. In contrast, the Standard for the Exchange of Product Data (STEP, ISO 10303) file is a widely adopted, neutral boundary representation (B-rep) format directly compatible with manufacturing, but its graph-structured, cross-referenced nature poses unique challenges for auto-regressive LLMs. To address this, we curate a dataset of 40K STEP-caption pairs and introduce novel preprocessing tailored for the graph-structured format of STEP, including a depth-first search-based reserialization that linearizes cross-references while preserving locality and chain-of-thought(CoT)-style structural annotations that guide global coherence. We integrate retrieval-augmented generation to ground predictions in relevant examples for supervised fine-tuning, and refine generation quality through reinforcement learning with a specific Chamfer Distance-based geometric reward. Experiments demonstrate consistent gains of our STEP-LLM in geometric fidelity over the Text2CAD baseline, with improvements arising from multiple stages of our framework: the RAG module substantially enhances completeness and renderability, the DFS-based reserialization strengthens overall accuracy, and the RL further reduces geometric discrepancy. Both metrics and visual comparisons confirm that STEP-LLM generates shapes with higher fidelity than Text2CAD. These results show the feasibility of LLM-driven STEP model generation from natural language, showing its potential to democratize CAD design for manufacturing.

Magneto-elastic Granular Architectures (MEGAs) consisting of 3D-printed star-shaped units and magnetic branch termini offer modularity and reversibility not seen in conventional monolithic materials. Tailoring magnetic strength and star topology leads to ordered structures with distinct mechanical responses. However, single-component MEGAs typically exhibit limited tensile extensibility. Here we show that binary superlattices, introduced by combining two-branch (Bar) and three-branch (Y) units, enable programmable tensile strength and ductility and yield architectures with extensibility far exceeding that of single-component systems. By tuning unit sizes, we uncover hierarchical dual networks in which subnetwork units provide hidden length released via sacrificial magnetic bond detachment during deformation. We construct a phase diagram, a performance map, and a predictive framework to reveal design rules linking unit geometry, deformation pathways, and macroscopic response. Our work opens opportunities for easy-to-assemble and reconfigurable architected metamaterials with controllable brittle–ductile transitions, with implications for adaptive load-bearing, impact mitigation, and robotic systems.

High-dimensional structure and composition spaces pose a fundamental challenge in materials discovery due to the lack of efficient approaches for navigating the vast and complex design space. Although machine learning (ML) has aided materials discovery, most existing ML models lack the ability to quantify epistemic uncertainty arising from limited data. Developing this capability is particularly challenging for tasks involving high-dimensional design representations, such as atomic structures. In this study, building on the Bayesian optimization (BO) framework, we propose an uncertainty-aware atomistic machine learning model, uncertainty-aware PointNet, which enables automated representation learning directly from high-dimensional design inputs, such as atomic structures, and achieves principled uncertainty quantification through the use of spectral-normalized neural Gaussian process. By utilizing a constrained expected improvement acquisition function, our BO framework simultaneously considers multiple design criteria. We demonstrate the effectiveness of our approach in two materials discovery case studies: (1) identifying catalysts for the carbon dioxide reduction reaction and (2) designing transparent conducting materials. The results show that our approach achieves high prediction accuracy, facilitates interpretable feature extraction, and enables multicriteria material design using constrained BO, leading to a significant reduction of computing power and time (a 10 & times; reduction in required simulation calculations). Beyond the demonstration examples, the developed method can accelerate materials discovery for various other applications with high-dimensional design inputs and expensive physics-based simulations.

Digital Twin (DT) technologies are transforming manufacturing by enabling real-time prediction, monitoring, and control of complex processes. Yet, applying DT to deformation-based metal forming remains challenging because of the strongly coupled spatial-temporal behavior and the nonlinear relationship between toolpath and material response. For instance, sheet-metal forming by the English wheel, a highly flexible but artisan-dependent process, still lacks digital counterparts that can autonomously plan and adapt forming strategies. This study presents an adaptive DT framework that integrates Proper Orthogonal Decomposition (POD) for physics-aware dimensionality reduction with a Koopman operator for representing nonlinear system in a linear lifted space for the real-time decision-making via model predictive control (MPC). To accommodate evolving process conditions or material states, an online Recursive Least Squares (RLS) algorithm is introduced to update the operator coefficients in real time, enabling continuous adaptation of the DT model as new deformation data become available. The proposed framework is experimentally demonstrated on a robotic English Wheel sheet metal forming system, where deformation fields are measured and modeled under varying toolpaths. Results show that the adaptive DT is capable of controlling the forming process to achieve the given target shape by effectively capturing non-stationary process behaviors. Beyond this case study, the proposed framework establishes a generalizable approach for interpretable, adaptive, and computationally-efficient DT of nonlinear manufacturing systems, bridging reduced-order physics representations with data-driven adaptability to support autonomous process control and optimization.

Shear-thinning hydrogel composites are promising materials for transarterial embolization, a minimally invasive procedure used to occlude blood vessels, owing to their injectability, mechanical strength, and biocompatibility. However, designing these materials is challenging because their properties depend sensitively on both polymer type and composition, leading to a complex mixed-variable optimization problem. Here we show that a machine learning framework combining latent-variable Gaussian processes (LVGP) with Bayesian optimization (BO) can efficiently design hydrogel nanocomposites formulated using nanoclay and natural polymers such as chitosan, gelatin, and alginate, with targeted storage modulus. This approach enables simultaneous exploration of multiple polymer systems and reduces experimental effort by 67% compared to conventional Bayesian optimization methods that optimize multiple polymers one at a time. Furthermore, our model learns a latent representation that captures similarities between composition-property relationships of different polymers, providing insight into material behavior. These results demonstrate a scalable strategy for accelerating the design of multi-material systems and highlight the potential of machine learning to address complex formulation challenges in soft materials and biomedical engineering.

Digital twins, virtual replicas of physical systems that enable real-time monitoring, model updates, predictions, and decision-making, present novel avenues for proactive control strategies for autonomous systems. However, achieving real-time decision-making in digital twins considering uncertainty necessitates an efficient uncertainty quantification (UQ) approach and optimization driven by accurate predictions of system behaviors, which remains a challenge for learning-based methods. This article presents a simultaneous multistep robust model predictive control (MPC) framework that incorporates real-time decision-making with uncertainty awareness for digital twin systems. Leveraging a multistep-ahead predictor named time-series dense encoder (TiDE) as the surrogate model, this framework differs from conventional MPC models that provide only one-step-ahead predictions. In contrast, TiDE can predict future states within the prediction horizon in one shot, significantly accelerating MPC. Furthermore, quantile regression is employed with the training of TiDE to perform flexible and computationally efficient UQ on data uncertainty. Consequently, with the deep learning quantiles, the robust MPC problem is formulated into a deterministic optimization problem and provides a safety buffer that accommodates disturbances to enhance the constraint satisfaction rate. As a result, the proposed method outperforms existing robust MPC methods by providing less conservative UQ and has demonstrated efficacy in an engineering case study involving directed energy deposition (DED) additive manufacturing. This proactive, uncertainty-aware control capability positions the proposed method as a potent tool for future digital twin applications and real-time process control in engineering systems.

Bayesian optimization (BO) is a widely used framework for optimizing expensive black-box functions and has found applications across engineering, materials science, and machine learning. However, in many practical tasks, feasibility of design is not readily quantifiable. Conventional BO methods often ignore or poorly model design feasibility, leading to impractical or invalid solutions. This work addresses the challenge of integrating feasibility information directly into the BO process. Here we show that modeling feasibility using Gaussian process (GP) classification and treating it as an objective together with other performance objectives in a multi-objective BO setup significantly improves solution quality across different benchmark design tasks. We develop a latent variable Gaussian process classifier for modeling feasibility over categorical design spaces, and use a Dirichlet-based GP classifier for continuous spaces. Our approach provides quantification of feasibility, offering clear optimization guidance. Comparative studies on analytical and real-world test problems demonstrate enhanced performance in terms of both feasibility and optimality. This approach could be extended to a wide range of applications where feasibility is implicit or difficult to define, such as materials discovery, drug design, and chemical process optimization. By re-framing feasibility as a learnable objective, our work opens new avenues for constrained optimization under uncertainty.

Creep is a primary life-limiting mechanism for metallic components operating at high temperature, producing permanent deformation under sustained loads even when stresses remain below yield. The design of structures to minimize this deformation is critical to extending the service life of components. Incorporating creep into topology optimization (TO) remains open because the response is nonlinear, history-dependent, and thermomechanically coupled, and prior work often relies on linear viscoelastic models, which do not capture the behavior of metals at high temperatures. To bridge this gap, we introduce a differentiable thermo-structural TO framework. The approach considers creep deformation using the Norton model and leverages JAX's automatic differentiation to perform adjoint sensitivity analysis, enabling efficient gradient-based optimization. The transient material response is solved via a backward Euler scheme over a prescribed service life. Our objective is to minimize creep deformation subject to a volume constraint. We first demonstrate the framework on canonical two-dimensional benchmarks, showing that the proposed formulation significantly reduces permanent deformation compared to designs optimized solely for elastic stiffness. We then pose, as a challenge problem, the compositional design of a three-dimensional graded material turbine blade in which the local mixture of two candidate alloys is optimized. This challenge problem exercises the full capability of the framework, including transient nonlinear creep, coupled thermal loading, three-dimensional geometry, and gradient-based multi-material design, highlighting the need for creep-aware design in high-temperature applications.

Acoustic waves play a crucial role in various applications, including medical imaging, non-destructive testing, and sonar systems. One of the significant challenges in these applications is impedance matching, which is essential for minimizing reflections and maximizing the transfer of acoustic energy between different media. Acoustic metamaterials offer a promising solution to this challenge. In addition to impedance control, gradient stiffness can enhance structural efficiency and enable spatial control of wave propagation, making it a valuable feature in acoustic metamaterial design. In this paper, we present our developed computational method to design 3D stiffness gradient acoustic metamaterials for impedance matching. The key steps in our approach include generating initial designs using a periodic covariance function to provide unit cells that are both periodic on the boundaries and randomly formed inside the unit cell. Furthermore, we integrated manufacturing constraints into the design process, ensuring that the structures are interconnected for fabrication. We propose two computational optimization algorithms: GenUnit, based on a non-dominated sorting genetic algorithm (NSGA-II), and MLMatch, which leverages differentiable machine learning. The two approaches are not separate contributions but complementary components of a unified framework. GenUnit requires no training data and directly interfaces with physics-based simulations, making it highly accurate but slower for large-scale exploration. In contrast, MLMatch is data-hungry during training but, once trained, enables near-instantaneous inference and broad design-space coverage. Together, they form a hybrid strategy: MLMatch rapidly explores the global design space, and GenUnit provides local refinement with high-fidelity accuracy. This balance between training cost, inference time, and precision is the motivation for including both methods in the same study. We applied this dual-algorithm framework to generate two metallic-based metamaterial designs that match the acoustic impedance of water while exhibiting a controlled gradient in stiffness (from stiff to soft). The stiffness gradient is particularly advantageous in applications where one side of the structure must interface with soft or sensitive surfaces, such as human tissue or delicate components. This work paves the way for improved materials in various acoustic applications, particularly in ultrasound devices, by providing better impedance matching and thereby improving the efficiency of acoustic energy transfer.

Polymer nanocomposites, inherently tailorable materials, are potentially capable of providing higher strength to weight ratio than conventional hard metals. However, their disordered nature makes processing control and hence tailoring properties to desired target values a challenge. Additionally, the interfacial region, also called the interphase, is a critical material phase in these heterogeneous materials and its extent depends on variety of microstructure features like particle loading and dispersion or inter-particle distances. Understanding process-structure-property (PSP) relation can provide guidelines for process and constituents' design. Our work explores nuances of PSP relation for polymer nanocomposites with attractive pairing between particles and the bulk polymer. Past works have shown that particle functionalization can help tweak these interactions in attractive or repulsive type and can cause slow or fast decay of stiffness properties in polymer nanocomposites. In this work, we develop a material model that can represent decay for small strain elastoplastic (Young's modulus and yield strength) properties in interfacial regions and simulate representative or statistical volume element behavior. The interfacial elastoplastic material model is devised by combining local stiffness and glass transition measurements from atomic force microscopy and fluorescence microscopy. This model is combined with a microstructural design of experiments for agglomerated nanocomposite systems. Agglomerations are particle aggregations arising from processing artifacts. Twin screw extrusion process can reduce extent of aggregation in hot pressed samples via erosion or rupture depending on screw rpms and torque. We connect this process-structure relation to structure-property relation that emerges from our study. We discover that balancing between local stress concentration zones (SCZ) and interfacial property decay governs how fast yield stress can improve by breaking down agglomeration via erosion. Erosion is relatively more effective in helping improve nanocomposite yield strength. We also observe saturation in properties where incremental increase brought on by erosion is slowed due to increasing SCZ and saturation in interphase percolation.

Many real-world optimization problems in scientific discovery and engineering design optimization involve multiple design evaluation sources, such as simulators, experiments, or manufacturing sites, operate independently, and evaluate different functions of the same underlying objective (quantity of interest). In this work, we refer to these design evaluation sources as agents. Such agents often differ in their objective function mappings, evaluation budgets, and accessible optimization variables, which complicate coordination and information sharing. Bayesian optimization (BO) is a widely used framework for expensive blackbox optimization, yet its standard single-agent formulation assumes centralized control and full data sharing. Recent collaborative BO methods relax these assumptions but still rely on uniform resources, fully shared input spaces, and closely aligned tasks, and these requirements are seldom met in real applications. To address these limitations, we introduce adaptive resource-aware collaborative Bayesian optimization (ARCO-BO), a framework that explicitly accounts for heterogeneity in multi-agent optimization. ARCO-BO integrates three key components: a similarity- and optimal-location-aware consensus mechanism for adaptive information sharing, a budget-aware asynchronous sampling strategy for resource coordination, and a partial input-space sharing scheme for heterogeneous optimization variables. Experiments on synthetic benchmarks and high-dimensional engineering optimization problems demonstrate that ARCO-BO consistently outperforms independent BO and the existing consensus-based collaborative BO, achieving robust and efficient performance in complex heterogeneous multi-agent optimization settings.

Active suspension systems are critical for enhancing vehicle comfort, safety, and stability, yet their performance is often limited by fixed hardware designs and control strategies that cannot adapt to uncertain and dynamic operating conditions. Recent advances in Digital Twins (DTs) and Reinforcement Learning (RL) offer new opportunities for real-time, data-driven optimization across a vehicle's lifecycle. However, integrating these technologies into a unified framework for co-optimizing physical and control systems remains an open challenge. This work presents an RL-based Control Co-Design (CCD) framework for full-vehicle active suspensions using multi-generation design and DT concepts. Through integrating automatic differentiation into Deep Reinforcement Learning (DRL), we jointly optimize physical components of suspension systems and control policies under varying driver behaviors and environmental uncertainties. The DRL technique also addresses the challenge of partial observability, where only limited states can be sensed and fed back to the controller, by learning optimal control actions directly from available sensor information. The framework incorporates model updating with quantile learning to quantify data uncertainty, enabling real-time decision-making and adaptive learning from digital-physical interactions. The approach demonstrates personalized optimization of autonomous suspension systems under two distinct driving settings (mild and aggressive). The results show that the optimized systems achieve smoother trajectories and reduce control efforts by approximately 58% and 12% for mild and aggressive while improving ride comfort by approximately 17% and 28%, respectively. Contributions of this work include: (1) developing a DT-enabled CCD framework integrating DRL and uncertainty-aware model updating for full-vehicle active suspensions, (2) introducing a multi-generation design framework for self-improving systems across the whole lifecycle, and (3) demonstrating personalized optimization of active suspension systems for distinct types of drivers.

High-throughput materials characterization is essential for accelerating materials discovery. To enable high-throughput characterization, machine learning (ML) has been a powerful tool. However, the broader application of ML in experimental settings is limited by key challenges, i.e., the scarcity of labeled experimental data and the lack of uncertainty estimation in model predictions. In this work, we present Uncertainty-aware Simulation-to-Experiment Modeling (USEM), a novel approach that enables ML models trained on labeled simulation data to be adapted for analyzing unlabeled or sparsely labeled experimental data. The method uses adversarial domain adaptation in the latent space to reduce the distance between simulation and experiment data, while incorporating predictive uncertainty through spectral-normalized neural Gaussian processes (SNGP). We demonstrate the effectiveness of USEM on both synthetic and experimental X-ray diffraction (XRD) data, showing improved predictive accuracy and the ability to identify out-of-distribution samples. USEM offers a scalable and trustworthy solution for high-throughput characterization in ML-driven materials discovery.

Advances in manufacturing, stimuli-responsive materials, and actuation technologies have fueled the development of programmable material systems (PMS) that can change their functional states in response to external stimuli. This functional switch is instrumental for applications that demand adaptability and robustness under dynamic environments such as wearable electronics, haptic interfaces, and soft robots. In principle, achieving an interesting degree of programmability in material systems is a challenging design problem as it requires careful consideration of geometry, material, stimulus, and process amidst dynamic environments. This task becomes even more challenging when compounded by multifunctionality requirements, multiphysics modeling, coupling among the design entities, and resource-intensive performance evaluation. In practice, prevailing approaches often heuristically fix some entities or decouple their interactions, which significantly limits achievable performance and generalizability of the design approach. Motivated by this gap between promise and practice, we present a co-design formalism that highlights shared design considerations across prior works and categorizes design strategies for managing complexity. With the proposed lens, we reveal common threads in existing design approaches, identify opportunities and open research questions for co-design to achieve better performance and generalization. Through this perspective article, we aim to lay a design foundation for boosting progress in this emerging interdisciplinary field.

The evolution of artificial intelligence (AI) and machine learning (ML) is reshaping smart manufacturing by providing new capabilities for efficiency, adaptability, and autonomy across industrial value chains. However, the deployment of AI and ML in industrial settings still faces critical challenges, including the complexity of industrial big data, effective data management, integration with heterogeneous sensing and control systems, and the demand for trustworthy, explainable, and reliable operation in high-stakes industrial environments. In this roadmap, we present a comprehensive perspective on the foundations, applications, and emerging directions of AI and ML in smart manufacturing. It is structured in three parts. The first highlights the foundations and trends that frame the evolution of AI in smart manufacturing. The second focuses on key topics where AI is already enabling advances, including industrial big data analytics, advanced sensing and perception, autonomous systems, additive and laser-based manufacturing, digital twins, robotics, supply chain and logistics optimization, and sustainable manufacturing. The third section explores non-traditional ML approaches that are opening new frontiers, such as physics-informed AI, generative AI, semantic AI, advanced digital twins, explainable AI, RAMS, data-centric metrology, LLMs, and foundation models for highly connected and complex manufacturing systems. By identifying both opportunities and remaining barriers across these areas, this roadmap outlines the advances needed in methods, integration strategies, and industrial adoption. We hope this roadmap will serve as a guide for researchers, engineers, and practitioners to accelerate innovation, align academic and industrial priorities, and ensure that AI-driven smart manufacturing delivers reliable, sustainable, and scalable impact for the future of manufacturing ecosystems.

ABSTRACT Micellar gels formed by self‐assembling block copolymers have advanced applications spanning drug delivery, biofabrication, and soft robotics. However, their broader use as crosslinked hydrogels is limited by the challenge of simultaneously tuning their mechanical properties and stimuli‐responsive behaviors while maintaining the ability to process the pre‐hydrogel through techniques like extrusion. Here, we show that oligomerization of Pluronic triblock copolymers provides a single tunable parameter, the oligomer fraction, , for modulating the inter‐micellar connectivity and mechanical properties of Pluronic hydrogels. Increasing produces hydrogels spanning brittle to highly extensible responses. Coarse‐grained molecular dynamics simulations reveal that oligomerized Pluronic chains bridge multiple micelles, forming highly interconnected networks. Since oligomerization maintains the micellar network architecture of high‐ hydrogels, we demonstrate a cooling‐induced, reverse thermal shape‐memory behavior in which plastic deformation at room temperature is reversed upon cooling below Pluronic's lower critical solution temperature. Oligomerization also preserves the printability of Pluronic gels, enabling anisotropic hydrogels with spatially programmable mechanical properties via embedded 3D printing. Graph‐theoretic analyses of simulated networks establish intermicellar connectivity as a key structural parameter governing mechanical tunability. These results present oligomerization as a versatile processing strategy for tuning mechanical and responsive properties of micellar hydrogels.

Adaptive learning implementations for materials design are challenged by the complex, nonlinear relationships between composition and properties, particularly in high-performance applications such as high-temperature compositionally complex refractory alloys. Traditional Bayesian optimization (BO) methods, which typically rely on a single Gaussian Process (GP) surrogate, often struggle to model heterogenous behaviors across the design domain. To address this limitation, we introduce collaborative BO as a multi-agent framework for materials discovery. In the context of optimizing compositions for desired properties, each agent models a specific subregion of the design space, where subregions share similar property trends, and exchanges information with the other agents to expedite exploration and design optimization. Comparative evaluations demonstrate that, when compared to single-agent BO and other approaches discussed in this article multi-agent BO allows flexible information-sharing protocols and effectively reduces iterations of adaptive learning while reliably delivering designs that meet the targeted mechanical properties. These findings provide novel insights into the behavior of refractory multi-component alloys, using the Hf-Ti-Ta-Nb system as a case study, and illustrate the potential of adaptive multi-agent learning in efficiently screening extensive materials libraries. Moreover, the framework is broadly applicable to other problems characterized by diverse data sources, where advanced optimization strategies are essential for accelerated materials discovery.

Gel-elastomer composites, comprising an active swellable hydrogel and a passive elastomer, are a compelling class of programmable material systems (PMS) capable of shape morphing under multiphysics actuation. The precise design of the topology and material distribution unlocks complex programmability instrumental in wearable electronics, soft robots, and drug delivery; however, the structure-function relationship is highly non-intuitive, rendering both trial-and-error and conventional design approaches largely intractable. To address this, we present a topology optimization (TO) framework for the automated design of such structures, enabling systematic exploration of the design space for target functionalities realized via programmable shape morphing. In particular, we propose a multi-material TO framework that concurrently optimizes the structural topology and the spatial distribution of the gel-elastomer phases. The design is represented via a coordinate-based neural network, and the mechanical response of both phases is described within a unified constitutive framework based on the Flory-Rehner theory. Furthermore, we present an end-to-end differentiable design framework with implicit differentiation that accommodates various objective functions, constraints, and discretizations. We demonstrate the framework on shape-programming structures and soft actuators. The framework is further validated through the design of organogel-hydrogel composites for multi-stimuli responsiveness across chemically distinct solvent environments, and of anisotropic hydrogels wherein the local fiber orientation is optimized concurrently with the topology. The codebase implemented in JAX is publicly shared to support benchmarking and reproducibility.

Control co-design (CCD) integrates physical and control system design to improve the performance of dynamic and autonomous systems. Despite advances in uncertainty-aware CCD methods, real-world uncertainties remain highly unpredictable. Multigeneration design addresses this challenge by considering the full lifecycle of a product: data collected from each generation informs the design of subsequent generations, enabling progressive improvements in robustness and efficiency. Digital twin (DT) technology further strengthens this paradigm by creating virtual representations that evolve over the lifecycle through real-time sensing, model updating, and adaptive re-optimization. This article presents a DT-enabled CCD framework that integrates deep reinforcement learning (DRL) to jointly optimize physical design and controller. DRL accelerates real-time decision-making by allowing controllers to continuously learn from data and adapt to uncertain environments. Extending this approach, the framework employs a multigeneration paradigm, where each cycle of deployment, operation, and redesign uses collected data to refine DT models, improve uncertainty quantification through quantile regression, and inform next-generation designs of both physical components and controllers. The framework is demonstrated on an active suspension system, where DT-enabled learning from road conditions and driving behaviors yields smoother and more stable control trajectories. Results show that the method significantly enhances dynamic performance, robustness, and efficiency. Contributions of this work include: (1) extending CCD into a lifecycle-oriented multigeneration framework, (2) leveraging DTs for continuous model updating and informed design, and (3) employing DRL to accelerate adaptive real-time decision-making.

Compositionally Graded Alloys (CGAs) offer unprecedented design flexibility by enabling spatial variations in composition; tailoring material properties to local loading conditions. This flexibility leads to components that are stronger, lighter, and more cost-effective than traditional monolithic counterparts. The fabrication of CGAs have become increasingly feasible owing to recent advancements in additive manufacturing (AM), particularly in multi-material printing and improved precision in material deposition. However, AM of CGAs requires imposition of manufacturing constraints; in particular limits on the maximum spatial gradation of composition. This paper introduces a topology optimization (TO) based framework for designing optimized CGA components with controlled compositional gradation. In particular, we represent the constrained composition distribution using a band-limited coordinate neural network. By regulating the network's bandwidth, we ensure implicit compliance with gradation limits, eliminating the need for explicit constraints. The proposed approach also benefits from the inherent advantages of TO using coordinate networks, including mesh independence, high-resolution design extraction, and end-to-end differentiability. The effectiveness of our framework is demonstrated through various elastic and thermo-elastic TO examples.