What problem it solves

Complex robotics and computer-vision processes produce multiple synchronized data streams whose relationships over time are difficult to understand from text logs or conventional debugging tools. Rerun addresses this by recording, visualizing, querying, and streaming multimodal data from sources such as robots, human-data rigs, simulation, and web video.

Key features

• Multimodal ingestion and visualization — Ingest multi-rate data such as images, point clouds, transforms, time series, joint states, and video, then inspect synchronized streams in the built-in viewer.

  Source: README

• Queryable recorded data — Query raw, intermediate, and derived recordings with dataframes or SQL, and extract clean datasets for model training and evaluation.

  Source: README

• Direct training streaming — Stream dataset mixes directly into training without export jobs or stale copies, using the same recorded data available for visualization and querying.

  Source: README

• Multiple language SDKs — Develop with Python, Rust, or C++ SDKs and use the viewer through the supported installation paths.

  Source: README

• Purpose-built multimodal storage — Store multi-rate physical data in a Rust-based column-chunk storage model and preserve supported data types end to end.

  Source: README

Quick start

Copied verbatim from the project's README.

import rerun as rr  # pip install rerun-sdk

rr.init("rerun_example_app")

rr.spawn()  # Spawn a child process with a viewer and connect
# rr.save("recording.rrd")  # Stream all logs to disk
# rr.connect_grpc()  # Connect to a remote viewer

# Associate subsequent data with 42 on the “frame” timeline
rr.set_time("frame", sequence=42)

# Log colored 3D points to the entity at `path/to/points`
rr.log("path/to/points", rr.Points3D(positions, colors=colors))
…

Who it may suit

Inference: may suit robotics, simulation, and computer-vision developers who need to inspect temporal multimodal data, debug sensor-driven processes, or prepare recorded datasets for training and evaluation.

Topics to discuss

• Which ingestion and query workflows should receive the most community attention?

• How should multimodal recordings be organized for reproducible training-data curation?

• What examples would best demonstrate debugging and dataset-extraction workflows?

Written by ReadmeX AI from the project's README, source and public GitHub data; it may contain mistakes.

Read the original README https://github.com/rerun-io/rerun#readme

ReadmeX opened this community for a trending project; it is not run by the project. Maintainers can claim it from the community page.