{
  "id": 28566,
  "title": "A better way to turn 2D designs into 3D models for rapid prototyping",
  "url": "https://urgent.news/2026/07/16/a-better-way-to-turn-2d-designs-into-3d-models-for-rapid-prototyping-28566",
  "topic": "ai",
  "section": "AI",
  "published": "2026-07-16T04:00:00.000Z",
  "source": {
    "name": "MIT News Research",
    "slug": "mit-news-research",
    "url": "https://news.mit.edu/2026/turning-2d-designs-into-3d-models-for-rapid-prototyping-0716"
  },
  "original_language": "en",
  "account": "Engineers frequently employ vision-language models to create 3D models for various components, such as airplane parts or auto elements. To assess the performance of these models, they generate 3D representations using conventional computer-aided design (CAD) software, which can then undergo virtual crash or durability tests. A team of researchers from MIT and other institutions has engineered a system that empowers vision-language models to automatically transmute 2D designs into CAD programs, enhancing both the precision and practicality of the outcomes while demanding fewer computational resources. This advancement has the potential to simplify rapid prototyping, lower expenses, and enable engineers to discover valuable design decisions that might otherwise go unnoticed. The system functions by generating novel data based on the model's performance when attempting to convert a 2D image into a CAD program. It identifies areas where the model falters, incorporates these findings into a dataset alongside successful solutions, and utilizes this data to instruct the model on how to rectify specific errors and tackle complex issues that would otherwise pose challenges. Lead author Giorgio Giannone, a research affiliate at MIT's Design Computation and Digital Engineering (DeCoDE) Lab and a principal research scientist at Red Hat, explains, \"We aim for engineers to be able to point our framework at an underperforming CAD model, set a compute budget, and let the system take over - transforming the model's own mistakes into superior training data.\" Co-senior authors Akash Srivastava from IBM and Faez Ahmed from MIT, along with other contributors, emphasize that this innovation brings trustworthy AI design tools closer to everyday engineering applications.",
  "summary": "Researchers developed an automated framework that helps AI models generate CAD programs more accurately and efficiently.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 2,
    "also_reported_by": [
      {
        "outlet": "MIT News AI",
        "title": "A better way to turn 2D designs into 3D models for rapid prototyping",
        "url": "https://urgent.news/2026/07/16/a-better-way-to-turn-2d-designs-into-3d-models-for-rapid-prototyping",
        "published": "2026-07-16T04:00:00.000Z"
      }
    ]
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}