esc
Start typing to search the docs
Navigate Open

The Data Layer
for Physical AI

One layer from collection to training. Stream dataset mixes directly to GPUs. No preprocessing jobs, no stale copies, no idle GPUs.

Get started Try in browser

Open source SDKs.
A single toolchain to log, query, visualize, and train on multi-rate, multimodal data.

pip install rerun-sdk
rerun
Read the docs

@rerun-io/rerun 10.7k

Visualize everything

Review datasets, debug detail-level issues, and extend with custom visual tools across every stage of the pipeline.

Query and transform

Dataframe and SQL across raw, intermediate, and derived data. Add columns and slice mixes as easily as editing a table.

Log, store, and stream

Log multi-rate, multimodal data into the column-chunk format the entire data layer is built on.

Open Source Stewardship

Rerun Hub
Build intelligence, not infrastructure

The production backend for the Rerun data layer. Catalog, byte-range indexing, and retrieval that turns your object stores into a queryable, streamable foundation. Run transforms on the edge or close to the data.

Book a meeting

News

A new data layer for robot learning

With the 0.32 SDK release Rerun becomes a unified data layer for physical data — covering visualization, querying, transformation, and training of multi-rate, multimodal robotics data. This post walks through the architecture, tours the new capabilities, and introduces Rerun Hub, our commercial data catalog and storage engine.

Read post

The community loves building with Rerun

LeRobot

LeRobot is Huggingface's state-of-the-art AI project for real-world robotics. The team uses Rerun as an integrated part of their visualization tools to inspect and debug training runs.

Brush

Brush is a 3D reconstruction engine using Gaussian splatting, developed by Arthur Brussee at DeepMind. Written in Rust, it aims to be portable, flexible and fast. Rerun visualizes training.

PyCuVSLAM

PyCuVSLAM is the official Python wrapper for NVIDIA's cuVSLAM library, providing GPU-accelerated visual SLAM and camera tracking for real-time localization and mapping.

Ultra

Ultra is building intelligent warehouse robots that automate repetitive, dangerous, and variable tasks. They use Rerun Hub for their visualization and transformation pipeline.

Project Aria

Project Aria is a research platform from Meta Reality Labs Research, pushing the state of the art in egocentric AI. Rerun visualizes sequences in their Aria Dataset Explorer.