--- title: AI agents create virtual playgrounds to help robots get... description: MIT CSAIL researchers developed SceneSmith, a system that uses three AI agents powered by GPT-5.2 to automatically generate realistic 3D indoor environments... canonical: https://daily.dev/posts/ai-agents-create-virtual-playgrounds-to-help-robots-get-crucial-training-data-yg0gnhko8 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: AI agents create virtual playgrounds to help robots get crucial training data | daily.dev og:description: MIT CSAIL researchers developed SceneSmith, a system that uses three AI agents powered by GPT-5.2 to automatically generate realistic 3D indoor environments... og:url: https://daily.dev/posts/ai-agents-create-virtual-playgrounds-to-help-robots-get-crucial-training-data-yg0gnhko8 og:image: https://api.daily.dev/og/posts/yg0gNHkO8.png og:image:alt: AI agents create virtual playgrounds to help robots get crucial training data og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # AI agents create virtual playgrounds to help robots get crucial training data **[MIT News](https://daily.dev/sources/mit)** · 7 min read · 0 upvotes · 0 comments ## Summary MIT CSAIL researchers developed SceneSmith, a system that uses three AI agents powered by GPT-5.2 to automatically generate realistic 3D indoor environments for robot training. A designer agent creates scene layouts, a critic evaluates realism, and an orchestrator manages the process. The system produces scenes with up to six times more objects than prior methods, including articulated items like cabinets. Over 1,300 diverse scenes were generated, and pretrained robot policies successfully operated in them without prior exposure. SceneSmith outperformed baselines like HSM and Holodeck in realism and prompt adherence, with 90%+ of 200 users preferring its visuals. The main trade-off is speed — generating a single scene can take hours — but the approach significantly reduces the need for costly real-world robot testing. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is SceneSmith and how does it generate 3D scenes for robot training? SceneSmith is a system from MIT CSAIL and Toyota Research Institute that uses three collaborating AI agents powered by the GPT-5.2 vision-language model to generate realistic 3D indoor scenes for robot simulation training. A designer agent creates scene elements, a critic evaluates realism, and an orchestrator manages their back-and-forth until the design is finished and ready to load into physics simulation software. _Roboticists tracking new approaches to simulation-based training data can follow developments like this on daily.dev._ ### How does SceneSmith compare to prior 3D scene generation baselines like HSM and Holodeck? SceneSmith produces environments with significantly more objects per scene, up to six times more than prior methods, and includes articulated items like cabinets that robots can open and close, which earlier baselines often lacked. Over 200 users rated its visuals as more realistic over 90 percent of the time and found it followed prompts more closely than competing systems such as HSM and Holodeck. _Engineers comparing simulation tools for robot training can keep up with benchmarks like these on daily.dev._ ### What is the main drawback of using SceneSmith to generate robot training environments? The main drawback is speed: producing a single detailed scene can take multiple hours because the agents create and closely scrutinize each object one by one. Researchers note that additional computing power could dramatically improve efficiency, and they hope to expand support to deformable objects like sponges as more 3D asset libraries become available. _Teams weighing simulation speed against realism for robot training can track updates like this on daily.dev._ ## Similar posts on daily.dev - [AI agents create virtual playgrounds to help robots get crucial training data](https://daily.dev/posts/ai-agents-create-virtual-playgrounds-to-help-robots-get-crucial-training-data-paryx7xxz) · Robohub · 0 upvotes · 0 comments - [Using generative AI to diversify virtual training grounds for robots](https://daily.dev/posts/using-generative-ai-to-diversify-virtual-training-grounds-for-robots-atuqskcr5) · Robohub · 0 upvotes · 0 comments - [Using generative AI to diversify virtual training grounds for robots](https://daily.dev/posts/using-generative-ai-to-diversify-virtual-training-grounds-for-robots-s3f7ogkoc) · MIT News · 0 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#robotics](https://daily.dev/tags/robotics) [View this post on daily.dev](https://daily.dev/posts/ai-agents-create-virtual-playgrounds-to-help-robots-get-crucial-training-data-yg0gnhko8) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"AI agents create virtual playgrounds to help robots get crucial training data","url":"https://daily.dev/posts/ai-agents-create-virtual-playgrounds-to-help-robots-get-crucial-training-data-yg0gnhko8","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/ai-agents-create-virtual-playgrounds-to-help-robots-get-crucial-training-data-yg0gnhko8"},"datePublished":"2026-07-13T18:57:11.898Z","dateModified":"2026-09-13T21:17:24.545Z","description":"MIT CSAIL researchers developed SceneSmith, a system that uses three AI agents powered by GPT-5.2 to automatically generate realistic 3D indoor environments...","image":"https://news.mit.edu/sites/default/files/images/202606/mit-csail-SceneSmith.gif","thumbnailUrl":"https://news.mit.edu/sites/default/files/images/202606/mit-csail-SceneSmith.gif","isAccessibleForFree":true,"articleSection":"MIT News","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"MIT News","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/881e20a40aa44b12ba4bd27f5c14289d","url":"https://daily.dev/sources/mit"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/ai-agents-create-virtual-playgrounds-to-help-robots-get-crucial-training-data-yg0gnhko8","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,robotics","timeRequired":"PT7M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"MIT News","item":"https://daily.dev/sources/mit"},{"@type":"ListItem","position":3,"name":"AI agents create virtual playgrounds to help robots get crucial training data"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/ai-agents-create-virtual-playgrounds-to-help-robots-get-crucial-training-data-yg0gnhko8#faq","mainEntity":[{"@type":"Question","name":"What is SceneSmith and how does it generate 3D scenes for robot training?","acceptedAnswer":{"@type":"Answer","text":"SceneSmith is a system from MIT CSAIL and Toyota Research Institute that uses three collaborating AI agents powered by the GPT-5.2 vision-language model to generate realistic 3D indoor scenes for robot simulation training. A designer agent creates scene elements, a critic evaluates realism, and an orchestrator manages their back-and-forth until the design is finished and ready to load into physics simulation software. Roboticists tracking new approaches to simulation-based training data can follow developments like this on daily.dev."}},{"@type":"Question","name":"How does SceneSmith compare to prior 3D scene generation baselines like HSM and Holodeck?","acceptedAnswer":{"@type":"Answer","text":"SceneSmith produces environments with significantly more objects per scene, up to six times more than prior methods, and includes articulated items like cabinets that robots can open and close, which earlier baselines often lacked. Over 200 users rated its visuals as more realistic over 90 percent of the time and found it followed prompts more closely than competing systems such as HSM and Holodeck. Engineers comparing simulation tools for robot training can keep up with benchmarks like these on daily.dev."}},{"@type":"Question","name":"What is the main drawback of using SceneSmith to generate robot training environments?","acceptedAnswer":{"@type":"Answer","text":"The main drawback is speed: producing a single detailed scene can take multiple hours because the agents create and closely scrutinize each object one by one. Researchers note that additional computing power could dramatically improve efficiency, and they hope to expand support to deformable objects like sponges as more 3D asset libraries become available. Teams weighing simulation speed against realism for robot training can track updates like this on daily.dev."}}]} ```