{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreihqd5gjcijazf4rvgysfzdccptmhr6iq4gbmgaazeacrwgwetupdu",
    "uri": "at://did:plc:mxzzpugn7bprjjrszwkbez3u/app.bsky.feed.post/3mgrzjn3e65e2"
  },
  "coverImage": {
    "$type": "blob",
    "ref": {
      "$link": "bafkreibdv7kiwnbzwszrtlquz6wz42j4jlfowcm6vie6mwutgny47vja44"
    },
    "mimeType": "image/jpeg",
    "size": 241094
  },
  "path": "/news/2026-03-hybrid-ai-planner-images-robot.html",
  "publishedAt": "2026-03-11T09:00:01.000Z",
  "site": "https://techxplore.com",
  "tags": [
    "Robotics"
  ],
  "textContent": "MIT researchers have developed a generative artificial intelligence-driven approach for planning long-term visual tasks, like robot navigation, that is about twice as effective as some existing techniques. Their method uses a specialized vision-language model to perceive the scenario in an image and simulate actions needed to reach a goal. Then a second model translates those simulations into a standard programming language for planning problems, and refines the solution.",
  "title": "Hybrid AI planner turns images into robot action plans"
}