About

Josephine Voigt Carstensen is the Gilbert W. Winslow Career Development Associate Professor of Civil and Environmental Engineering at the Massachusetts Institute of Technology. She leads the Carstensen Group, contributing to the academic and research environment within the department. Professor Carstensen has been recognized with several prestigious awards, including the National Science Foundation (NSF) CAREER Award, the Maseeh Award for Excellence in Teaching, and the Ole Madsen Mentoring Award, highlighting her commitment to both research and mentorship. She holds a PhD and a Master of Science in Engineering (MSE) in Civil Engineering from Johns Hopkins University, as well as a Master of Science (MSc) and Bachelor of Science (BSc) in Building Structures from the Technical University of Denmark.

Research topics

• Computer Science
• Mathematical optimization
• Structural engineering
• Mathematics
• Engineering
• Applied mathematics
• Mathematical analysis
• Mechanical engineering
• Algorithm
• Electronic engineering

Selected publications

• Minimum Carbon Trusses: Constructible Multi-Component Designs with Mixed-Integer Linear Programming

  arXiv (Cornell University) · 2026-02-06

  articleOpen accessSenior author

  Truss optimization is a rich research field receiving renewed interest in limiting the carbon emissions of construction. However, a persistent challenge has been to construct highly optimized and often complex designs. This contribution formulates and solves new mixed-integer linear programs that enable consideration of the interplay between environmental impact and constructability. Specifically, the design engineer is enabled to design with multiple materials and/or structural components, apply separate minimum and maximum cross-sectional area bounds, and constrain the complexity of the structural connections. This is done while explicitly considering compatibility and constitutive laws. The results demonstrate that the lowest embodied carbon designs change significantly when constructability constraints are applied, highlighting the need for an integrated optimization approach. In one example, introducing a lower-carbon material option has almost no effect on the environmental performance, whereas another sees an improvement of nearly 29%. The extensibility of the formulation to design with three component types and additional constraints is demonstrated for a prestressed tensegrity example.

  PublisherOA PDF

• Topology Optimization for Materially Efficient Reinforced Concrete Design: Development, Fabrication, and Structural Evaluation

  ArXiv.org · 2026-04-23

  articleOpen accessSenior author

  The production of concrete generates roughly 8% of anthropogenic CO2 globally, largely because of the massive quantities that are manufactured. New design methods must be developed and deployed to improve the material efficiency of reinforced concrete structures, and reduce concrete's carbon impact. This research uses topology optimization, a free-form structural optimization method, for improved structural design. Two topology optimization frameworks are developed specifically for reinforced concrete design and construction. The automated design algorithms are used to generate geometries for materially-efficient reinforced concrete beams, which are fabricated and tested to compare performance to conventional design. The optimized results exhibit ductile failure and reach loads 36%-42% higher than the conventional design with the same material consumption. Through comparison to analytical models, the observed potential for material reduction while maintaining today's performance requirements without adding structural depth is around 33%, indicating a viable path forward in reaching carbon neutrality of reinforced concrete construction.

  PublisherOA PDF

• TopoCtrl: Post-Optimization Topology Editing Toward Target Structural Characteristics

  arXiv (Cornell University) · 2026-03-27

  preprintOpen access

  Topology optimization can generate high-performance structures, but designers often need to revise the resulting topology in ways that reflect fabrication preferences, structural intuition, or downstream design constraints. In particular, they may wish to explicitly control interpretable structural characteristics such as member thickness, characteristic member length, the number of joints, or the number of members connected to a joint. These quantities are often discrete, non-smooth, or only available through a forward evaluation procedure, making them difficult to impose within conventional optimization pipelines. We present TopoCtrl, a post-optimization control framework that repurposes the latent space of a pre-trained topology foundation model for explicit characteristic-guided editing. Given an optimized topology, TopoCtrl encodes it into the latent space of a latent diffusion model, applies partial noising to preserve instance similarity while creating room for modification, and then performs regression-guided denoising toward a prescribed target characteristic. The concept is to train a lightweight regression model on latent representations annotated with evaluated structural characteristics, and to use its gradient as a differentiable guidance signal during reverse diffusion. This avoids the need for characteristic-specific reformulations, hand-derived sensitivities, or iterative optimization. Because the method operates through partial noising of an existing topology latent, it preserves overall structural similarity while still enabling characteristic controls. Across representative control tasks involving both continuous and discrete structural characteristics, TopoCtrl produces target-aligned topology modifications while better preserving structural coherence and design intent than indirect parameter tuning or naive geometric post-processing.

  PublisherDOI

• Minimum Carbon Trusses: Constructible Multi-Component Designs with Mixed-Integer Linear Programming

  Open MIND · 2026-02-06

  preprintSenior author

  Truss optimization is a rich research field receiving renewed interest in limiting the carbon emissions of construction. However, a persistent challenge has been to construct highly optimized and often complex designs. This contribution formulates and solves new mixed-integer linear programs that enable consideration of the interplay between environmental impact and constructability. Specifically, the design engineer is enabled to design with multiple materials and/or structural components, apply separate minimum and maximum cross-sectional area bounds, and constrain the complexity of the structural connections. This is done while explicitly considering compatibility and constitutive laws. The results demonstrate that the lowest embodied carbon designs change significantly when constructability constraints are applied, highlighting the need for an integrated optimization approach. In one example, introducing a lower-carbon material option has almost no effect on the environmental performance, whereas another sees an improvement of nearly 29%. The extensibility of the formulation to design with three component types and additional constraints is demonstrated for a prestressed tensegrity example.

  DOI

• Topology Optimization for Materially Efficient Reinforced Concrete Design: Development, Fabrication, and Structural Evaluation

  arXiv (Cornell University) · 2026-04-23

  preprintOpen accessSenior author

  The production of concrete generates roughly 8% of anthropogenic CO2 globally, largely because of the massive quantities that are manufactured. New design methods must be developed and deployed to improve the material efficiency of reinforced concrete structures, and reduce concrete's carbon impact. This research uses topology optimization, a free-form structural optimization method, for improved structural design. Two topology optimization frameworks are developed specifically for reinforced concrete design and construction. The automated design algorithms are used to generate geometries for materially-efficient reinforced concrete beams, which are fabricated and tested to compare performance to conventional design. The optimized results exhibit ductile failure and reach loads 36%-42% higher than the conventional design with the same material consumption. Through comparison to analytical models, the observed potential for material reduction while maintaining today's performance requirements without adding structural depth is around 33%, indicating a viable path forward in reaching carbon neutrality of reinforced concrete construction.

  PublisherDOI

• Effect of fabrication restrictions on topology optimized 3D printed concrete structures

  SSRN Electronic Journal · 2026-01-01

  preprintOpen accessSenior author
  PublisherDOI

• TopoEdit: Fast Post-Optimization Editing of Topology Optimized Structures

  ArXiv.org · 2026-02-25

  articleOpen access

  Despite topology optimization producing high-performance structures, late-stage localized revisions remain brittle: direct density-space edits (e.g., warping pixels, inserting holes, swapping infill) can sever load paths and sharply degrade compliance, while re-running optimization is slow and may drift toward a qualitatively different design. We present TopoEdit, a fast post-optimization editor that demonstrates how structured latent embeddings from a pre-trained topology foundation model (OAT) can be repurposed as an interface for physics-aware engineering edits. Given an optimized topology, TopoEdit encodes it into OAT's spatial latent, applies partial noising to preserve instance identity while increasing editability, and injects user intent through an edit-then-denoise diffusion pipeline. We instantiate three edit operators: drag-based topology warping with boundary-condition-consistent conditioning updates, shell-infill lattice replacement using a lattice-anchored reference latent with updated volume-fraction conditioning, and late-stage no-design region enforcement via masked latent overwrite followed by diffusion-based recovery. A consistency-preserving guided DDIM procedure localizes changes while allowing global structural adaptation; multiple candidates can be sampled and selected using a compliance-aware criterion, with optional short SIMP refinement for warps. Across diverse case studies and large edit sweeps, TopoEdit produces intention-aligned modifications that better preserve mechanical performance and avoid catastrophic failure modes compared to direct density-space edits, while generating edited candidates in sub-second diffusion time per sample.

  PublisherOA PDF

• TopoEdit: Fast Post-Optimization Editing of Topology Optimized Structures

  Open MIND · 2026-02-25

  preprint

  Despite topology optimization producing high-performance structures, late-stage localized revisions remain brittle: direct density-space edits (e.g., warping pixels, inserting holes, swapping infill) can sever load paths and sharply degrade compliance, while re-running optimization is slow and may drift toward a qualitatively different design. We present TopoEdit, a fast post-optimization editor that demonstrates how structured latent embeddings from a pre-trained topology foundation model (OAT) can be repurposed as an interface for physics-aware engineering edits. Given an optimized topology, TopoEdit encodes it into OAT's spatial latent, applies partial noising to preserve instance identity while increasing editability, and injects user intent through an edit-then-denoise diffusion pipeline. We instantiate three edit operators: drag-based topology warping with boundary-condition-consistent conditioning updates, shell-infill lattice replacement using a lattice-anchored reference latent with updated volume-fraction conditioning, and late-stage no-design region enforcement via masked latent overwrite followed by diffusion-based recovery. A consistency-preserving guided DDIM procedure localizes changes while allowing global structural adaptation; multiple candidates can be sampled and selected using a compliance-aware criterion, with optional short SIMP refinement for warps. Across diverse case studies and large edit sweeps, TopoEdit produces intention-aligned modifications that better preserve mechanical performance and avoid catastrophic failure modes compared to direct density-space edits, while generating edited candidates in sub-second diffusion time per sample.

  DOI

• AI-Guided Human-In-the-Loop Inverse Design of High Performance Engineering Structures

  ArXiv.org · 2026-01-15

  articleOpen accessSenior author

  Inverse design tools such as Topology Optimization (TO) can achieve new levels of improvement for high-performance engineered structures. However, widespread use is hindered by high computational times and a black-box nature that inhibits user interaction. Human-in-the-loop TO approaches are emerging that integrate human intuition into the design generation process. However, these rely on the time-consuming bottleneck of iterative region selection for design modifications. To reduce the number of iterative trials, this contribution presents an AI co-pilot that uses machine learning to predict the user's preferred regions. The prediction model is configured as an image segmentation task with a U-Net architecture. It is trained on synthetic datasets where human preferences either identify the longest topological member or the most complex structural connection. The model successfully predicts plausible regions for modification and presents them to the user as AI recommendations. The human preference model demonstrates generalization across diverse and non-standard TO problems and exhibits emergent behavior outside the single-region selection training data. Demonstration examples show that the new human-in-the-loop TO approach that integrates the AI co-pilot can improve manufacturability or improve the linear buckling load by 39% while only increasing the total design time by 15 sec compared to conventional simplistic TO.

  PublisherOA PDF

• TopoCtrl: Post-Optimization Topology Editing Toward Target Structural Characteristics

  arXiv (Cornell University) · 2026-03-27

  articleOpen access

  Topology optimization can generate high-performance structures, but designers often need to revise the resulting topology in ways that reflect fabrication preferences, structural intuition, or downstream design constraints. In particular, they may wish to explicitly control interpretable structural characteristics such as member thickness, characteristic member length, the number of joints, or the number of members connected to a joint. These quantities are often discrete, non-smooth, or only available through a forward evaluation procedure, making them difficult to impose within conventional optimization pipelines. We present TopoCtrl, a post-optimization control framework that repurposes the latent space of a pre-trained topology foundation model for explicit characteristic-guided editing. Given an optimized topology, TopoCtrl encodes it into the latent space of a latent diffusion model, applies partial noising to preserve instance similarity while creating room for modification, and then performs regression-guided denoising toward a prescribed target characteristic. The concept is to train a lightweight regression model on latent representations annotated with evaluated structural characteristics, and to use its gradient as a differentiable guidance signal during reverse diffusion. This avoids the need for characteristic-specific reformulations, hand-derived sensitivities, or iterative optimization. Because the method operates through partial noising of an existing topology latent, it preserves overall structural similarity while still enabling characteristic controls. Across representative control tasks involving both continuous and discrete structural characteristics, TopoCtrl produces target-aligned topology modifications while better preserving structural coherence and design intent than indirect parameter tuning or naive geometric post-processing.

  PublisherOA PDF

Recent grants

• CAREER: Integrated Design and Digital Fabrication using Topology Optimization and Material Extrusion 3D Printing

  NSF · $560k · 2021–2026

Frequent coauthors

• James K. Guest

  13 shared

• Serena Danti

  University of Pisa

  5 shared

• Reza Lotfi

  5 shared

• Jackson L. Jewett

  Massachusetts Institute of Technology

  5 shared

• Pankaj Pankaj

  4 shared

• Mario Milazzo

  National Interuniversity Consortium of Materials Science and Technology

  4 shared

• Grunde Jomaas

  University of Primorska

  4 shared

• Jan Schroers

  4 shared

Labs

• Carstensen GroupPI

Awards & honors

• NSF CAREER Award, 2021
• CEE Maseeh Excellence in Teaching Award, 2021
• Denmark-America Foundation Fellowship, 2012
• Ole Madsen Mentoring Award, 2024