# The standard data layer for physical AI The data primitives to build, understand, and improve your data loop. Designed for multi-rate, multimodal data, from the first recording to massive scale. For humans and agents. - [Get started](/docs/getting-started/quick-start) - [Try in browser](/viewer) ## A single toolchain to work with multi-rate, multimodal data A flexible framework to build with rather than a platform you can't change. Childus meshum wasmus schemae, dataframae sdkum childus childus wasmus quaternionae keypointum containerae axisus. ### Visualize everything Review datasets, debug detail-level issues, and extend with your own views and tools for every stage of the pipeline. ### Query and transform Full dataframe or SQL queries over any robotics data. Extend with post-processing or annotations as easily as adding columns to a table. ### Log, store and convert Store multi-rate, multimodal data as column-chunks in rrd files. Either log directly or easily convert from any other format. ## Ready for agents Give agents a visual workspace for physical data. From loading a dataset to arranging views, keep every step grounded in data you can inspect. ![A colored point cloud of the Colosseum and surrounding streets, not a recorded GLOMAP reconstruction.](/scenes/colosseum/poster.webp) [Colosseum, Rome, Italy](https://sketchfab.com/3d-models/colosseum-rome-italy-535dc96e586f40bd956ea3cbff810055) by Brian Trepanier. [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). Adapted. Example agent prompt: Open this .rrd in Rerun with camera images beside the 3D point cloud. Overlay the camera trajectory and sync the views on the same timeline. Save the layout as a blueprint I can reuse for the next recording. Model: GPT-6 Astra **Let agents use the Viewer** through MCP to inspect data and control a running session. Give them a visual workspace for investigating recordings alongside you. **Give agents structure** with entities, components, and timelines. Dataframes and SQL let them query the data they need instead of sorting through files. **Keep every action inspectable** in the Viewer. Review what an agent selected or generated alongside the source data, so you can check its work in context. ## Rerun Hub. Infrastructure that powers your data loop. The production backend for the Rerun data layer. Catalog, byte-range indexing, and retrieval turn your object stores into a queryable, streamable foundation. Run transforms on the edge or close to the data. **Query inside every recording** with SQL or dataframes across columns, time ranges, frames, and values. **Evolve data without copies** by adding derived columns and schemas beside the raw recordings. **Stream dataset queries to GPUs** with a column-aware, video-aware dataloader and no export pipeline. **Give the team one shared view** to explore, annotate, and trace failures back to the source. - [Explore Rerun Hub](/hub) ## Used to ship physical AI at scale **Dexory** turns daily robot scans into a live view of warehouse inventory, helping teams see what is on every shelf. ![](https://static.rerun.io/home-feature/4d9d8370b8cbba5363cb6bae785b1181165e4276/full.jpg) **Ultra** builds warehouse robots that pack, sort, and kit orders, adapting to changing items and workflows on the floor. **Maven Robotics** combines dexterous manipulation and mobility in autonomous robots built to work alongside people. ![](https://static.rerun.io/home-feature/173677ef007a3b9d590436b484308e6628c4193b/full.jpg) **Sereact** builds AI that lets robots pick, place, and sort varied products, adapting to new items in busy warehouses. ## Loved by the community - [View post](https://x.com/jaiselsingh/status/2057969651896684865) - [View post](https://x.com/_ruma_v/status/2046989657620013303) - [View post](https://x.com/jaiselsingh/status/2047936537531142509) - [View post](https://x.com/rustaceans_rs/status/2021214835036315920) - [View post](https://x.com/kd3d_ai/status/2020979008494596329) - [View post](https://x.com/YutongLiang_/status/2011121848231301455) - [View post](https://x.com/stash_pomichter/status/2010640032764452890) - [View post](https://x.com/jueewo/status/2003144960380735681) - [View post](https://x.com/nudelbrot/status/1987481301285532022) - [View post](https://x.com/chris_j_paxton/status/1982822586623934866) - [View post](https://x.com/zshobbs/status/1981390444396417046) - [View post](https://x.com/novarobotics_eu/status/1970237860365132282) - [View post](https://x.com/johnbrandess/status/1975051916448235623) - [View post](https://x.com/PeterMitrano/status/1848515622184821166) - [View post](https://x.com/ZizuResearch/status/1846978669744451744) - [View post](https://x.com/vidstige81/status/1837030038899667256) - [View post](https://x.com/rauchg/status/1834763888640114983) - [View post](https://x.com/ConveLab/status/1829319248348401898) - [View post](https://x.com/YutongLiang_/status/1824702706675749349) - [View post](https://x.com/raghavaupp/status/1823910917781745716) - [View post](https://x.com/PeterMitrano/status/1823744492215632022) - [View post](https://x.com/Parvkpr/status/1815507533542592695) - [View post](https://x.com/heyellieday/status/1808936004105220390) - [View post](https://x.com/YongkieWiyogo/status/1770334611127648271) - [View post](https://x.com/raghavaupp/status/1769201746016911455) - [View post](https://x.com/changh95/status/1767940833666564376) - [View post](https://x.com/LvZhaoyang/status/1760702082800287917) - [View post](https://x.com/goodguyguybrush/status/1758277525904449725) - [View post](https://x.com/degentr0py/status/1728075963445977269) - [View post](https://x.com/rauchg/status/1710085538836025818) - [View post](https://x.com/steventey/status/1710316611851497879) ## Start building today Build, understand, and improve your data loop on one set of primitives, from your first recording to massive scale. Start with the SDK, or explore Rerun Hub. - [Install the SDK](/docs/getting-started/quick-start) - [Explore Rerun Hub](/hub)