github.devotd.
ReleaseRobotics

rerun-io/rerun 0.40.0 — database and visualization updates

rerun-io/rerun 0.40.0 is a tool that visualizes, queries, and streams multimodal robotics data; this release updates the database and visualization tooling for multi-mode and time data.

Repositoryrerun-io/rerun ↗

rerun-io/rerun 0.40.0 is a tool that visualizes, queries, and streams multimodal robotics data; this release updates the database and visualization tooling for multi-mode and time data.

GitHub description: Visualize, query, and stream to train on multimodal robotics data.

stars
★ 11.6K
Forks
879
Language
Rust
License
Apache-2.0
Latest release
0.40.0
Read original source ↗View on GitHub ↗

rerun-io/rerun 0.40.0 was released on October 21, 2026. It is a tool that visualizes and queries the various kinds of data collected in robotics (multimodal), and handles streaming so the data can be used for training. This release updated the database and visualization tooling for multi-mode and time data. The details are collected in the repository's release notes (https://github.com/rerun-io/rerun/releases/tag/0.40.0).

The core of this tool is that visualization, querying, and streaming are handled in a single repository. You look at the data with your own eyes to retrace its flow (visualization), you find the records that match a condition (querying), and you keep delivering those records for training (streaming). The flow from collecting data, through inspecting it, to preparing it for training continues in one place. This release shows how to install with Python and how to install with Rust, and it also offers an online demo. The repository currently has 11617 stars.

However, neither the summary of the 0.40.0 release notes nor the provided research record details the individual changes in the updated database and visualization tooling. There are no performance figures, no comparison with the previous version, and the license terms are not carried in this record either. So it is too early to judge what difference this update makes in actual performance and ease of use. What remains can only be judged against the online demo and data collected directly.