{"version":1,"type":"story","url":"https://digestai.news/story/nasa-and-ibm-release-open-source-lunar-foundation-model-for-moon-scien","json":"https://digestai.news/story/nasa-and-ibm-release-open-source-lunar-foundation-model-for-moon-scien.json","markdown":"https://digestai.news/story/nasa-and-ibm-release-open-source-lunar-foundation-model-for-moon-scien.md","slug":"nasa-and-ibm-release-open-source-lunar-foundation-model-for-moon-scien","headline":"NASA and IBM release open-source Lunar Foundation Model for moon science","summary":"NASA and IBM Research, together with academic partners, have unveiled the NASA‑IBM Lunar Foundation Model, an open‑source AI system trained on nearly 2 million tile bundles collected over 17 years of lunar orbiter missions.\n\nThe model, built on IBM’s TerraMind architecture and trained from scratch, ingests explicit lighting geometry for each tile and uses a FlexiViT technique to handle varying image patch sizes. In benchmark tests it cut polar‑ice prediction error by up to 22 % and improved coarse‑scale crater detection by nearly 19 % compared with the SwinV2‑B baseline, while requiring only half the labeled data. IBM and the researchers note that even a randomly initialized control model performed competitively on ice prediction, highlighting the value of the data handling approach.\n\nAll code, pretrained weights and the pretraining datasets are publicly available on Hugging Face and GitHub, and the model is integrated into the open‑source TerraTorch toolkit. Kevin Murphy, NASA’s chief science data officer, emphasized that the release makes decades of lunar observations easier for scientists to use, and Juan Bernabé‑Moreno of IBM Research Europe called the model a reusable foundation for downstream lunar research, not a substitute for physical measurements.","keyPoints":["NASA and IBM open-source Lunar Foundation Model trained on nearly 2 million multimodal tile bundles from 17 years of lunar data","Model reduces polar‑ice prediction error by up to 22 % and improves coarse‑scale crater detection by about 19 % versus SwinV2‑B baseline","Model, datasets and code are released on Hugging Face, GitHub and integrated into TerraTorch for lunar research"],"whyItMatters":"The model provides a reusable AI foundation that speeds lunar science, enabling better resource mapping and crater analysis with limited labeled data, and showcases open‑source AI for planetary research.","category":{"slug":"research","name":"Research","url":"https://digestai.news/category/research"},"entities":{"companies":["NASA","IBM"],"models":["NASA‑IBM Lunar Foundation Model"],"people":["Kevin Murphy","Juan Bernabé‑Moreno"]},"firstPublishedAt":"2026-10-04T10:10:12Z","updatedAt":"2026-10-04T10:10:12Z","sourceCount":1,"hasPrimarySource":false,"sources":[{"outlet":"The Decoder","title":"NASA and IBM's open source lunar model turns 17 years of orbiter data into a foundation for lunar science","url":"https://the-decoder.com/nasa-and-ibms-open-source-lunar-model-turns-17-years-of-orbiter-data-into-a-foundation-for-lunar-science","publishedAt":"2026-10-04T10:10:12Z","type":"press","primary":false,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"NASA and IBM release open-source Lunar Foundation Model for moon science\", 4 October 2026, https://digestai.news/story/nasa-and-ibm-release-open-source-lunar-foundation-model-for-moon-scien","publisher":"Digest AI","title":"NASA and IBM release open-source Lunar Foundation Model for moon science","datePublished":"2026-10-04T10:10:12Z","url":"https://digestai.news/story/nasa-and-ibm-release-open-source-lunar-foundation-model-for-moon-scien"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}