monolithic 3d printing

This work addresses the tight coupling between design intent and printer-specific representations in heterogeneous manufacturing, which hinders cross-platform reuse. The authors propose a novel compiler architecture that models fabrication-aware design as a staged, type-directed lowering process, decoupling source design, attribute translation, and backend compilation to enable manufacturing-agnostic expression. Introducing compiler paradigms to heterogeneous manufacturing for the first time, the approach unifies volumetric information—such as material composition, hardness, and color—through implicit geometry and typed spatial attribute fields, automatically generating voxel stacks, G-code, or slicer projects. Experiments demonstrate successful fabrication of complex objects embedding CT data, Shore hardness fields, and full-color fields on both material jetting and extrusion platforms, validating cross-process reusability. The accompanying Python toolkit is publicly released.

Operation sequence planning for hybrid additive-subtractive manufacturing (HADD) of arbitrarily shaped parts remains challenging due to geometric complexity and intermediate shape instability. Method: We propose a unified inverse-planning framework that starts from the target model and iteratively applies reversible additive “retraction” and subtractive “undo” operations until reaching an empty shape. The approach employs voxel-based modeling and a scalable search algorithm, with theoretical guarantees of solution existence for any manufacturable geometry. It inherently supports automatic tool switching and scales to large models. Contribution/Results: We validate the algorithm on multiple complex digital models and demonstrate high-precision physical fabrication on a hybrid manufacturing platform. Experimental results confirm both process feasibility and robust structural stability throughout the entire machining sequence—addressing critical challenges in HADD planning and execution.

High computational cost of conventional additive manufacturing (AM) simulations impedes real-time process optimization for medium-scale, complex structures. This paper proposes a digital twin framework tailored for extrusion-based printing: it employs G-code-driven adaptive octree voxelization to represent part geometry, and—uniquely—leverages the same parallel adaptive octree mesh simultaneously for high-resolution geometric modeling and transient thermo-phase-change coupled simulation, enabling tight integration of geometry generation and physics solving. The method supports real-time prediction of thermal field evolution for complex topologies—including variable-density infill—while preserving high voxel resolution and significantly enhancing simulation scalability. Experimental results demonstrate a 10–100× speedup over traditional approaches. The framework provides a deployable virtual process exploration platform that facilitates print quality improvement, waste reduction, and intelligent parameter optimization.

Current 3D printing defect detection methods rely heavily on expert intervention or task-specific models requiring extensive labeled data, exhibiting poor generalization across printers and firmware versions. To address this, we propose the first large language model (LLM)-based real-time monitoring and closed-loop control system for additive manufacturing. Our approach integrates multimodal perception of inter-layer images and leverages an LLM for zero-shot fault attribution, policy reasoning, and repair instruction generation—without domain-specific fine-tuning or annotated data. The system interfaces directly with printer APIs to autonomously execute corrective actions. It generalizes across heterogeneous hardware and firmware, accurately identifying common defects—including inconsistent extrusion, stringing, warping, and interlayer adhesion failure—localizing root-cause parameters (e.g., nozzle temperature, print speed, bed leveling), and dynamically adjusting them in real time. Fully automated and human-in-the-loop-free, it achieves diagnostic and corrective performance comparable to that of experienced AM engineers.

This work addresses the challenge of achieving complex shape transformations and multifunctional integration within a unified technical framework, which has been difficult for conventional shape-changing interfaces. The authors propose a co-design methodology that synergistically combines fused deposition modeling (FDM) 3D/4D printing with heat-sealed pneumatic actuation. By employing an integrated fabrication process, programmable structures and pneumatic components are seamlessly embedded to enable dynamically coupled actuation and constraint. The approach is grounded in a design space defined by four fundamental interaction primitives and supported by a dedicated authoring tool that facilitates end-to-end manufacturing. Multiple application examples demonstrate the method’s effectiveness and feasibility in realizing intricate, controllable deformations alongside functional integration.

This work addresses the persistent challenges in fused deposition modeling (FDM) printing—such as poor printability, insufficient mechanical strength, and complex post-processing caused by geometric defects like steep overhangs—by introducing the first end-to-end multi-agent system capable of automatically repairing original CAD models. The proposed framework integrates B-Rep parsing, graph neural network–based semantic recognition, and multimodal large language model reasoning to detect manufacturability issues and generate optimized STEP files along with detailed modification reports. By constructing face adjacency topological graphs, applying GraphSAGE for semantic labeling, leveraging Claude Sonnet for design suggestions, and validating modifications via GPT-4o’s visual reasoning, the system automates the entire pipeline from geometric analysis to natural-language design recommendations. Evaluated on a birdhouse model, it accurately identified overhang regions and effectively proposed corrective strategies such as chamfering, filleting, or part reorientation, substantially overcoming the reliance on manual intervention inherent in traditional design-for-manufacturing approaches.

This work addresses a critical gap in functionally graded additive manufacturing, where existing approaches focus predominantly on geometric or material distribution while neglecting the coordinated control of slicing and process parameters, thereby forcing users to manually configure slicer region settings—an inefficient and error-prone practice. To overcome this limitation, the authors propose a novel slicer project compilation pipeline that, for the first time, automatically maps implicit heterogeneous designs to the slicing layer. By leveraging spatial attribute partitioning, submesh extraction, and slicer-dialect serialization, the method generates syntactically compliant .3MF project files embedding submeshes, parameter recipes, and process states. It supports calibration from high-order properties—such as those of thermally responsive foaming materials—to process parameters, integrating three parameter systems: setting meshes, virtual extrusion, and color/material halftoning. Experimental validation demonstrates successful fabrication of specimens featuring gradient toolpaths, performance tuning, texture-process co-control, and color blending, replacing over 2,500 manual operations, with an open-source framework enabling extensible functionally graded manufacturing.

This study addresses the challenges of low accuracy, difficult control, and unreliable error prediction in collaborative robots used for 3D printing, which stem from dynamic complexity. The authors propose a full-process integrated parameter identification framework tailored for collaborative robots, comprising five sequential steps: geometric and inertial analysis, friction modeling, controller parameter identification, and system residual parameter estimation. This approach yields a physically consistent dynamic model and, for the first time, enables unified identification of robot body, actuator, and controller parameters—making it suitable for real-world scenarios with limited sensing and programming capabilities. Experimental validation on a six-degree-of-freedom collaborative robot performing thermoplastic extrusion demonstrates excellent agreement between the identified model and empirical data, significantly enhancing printing accuracy, control performance, and error prediction fidelity.