11-06, 11:25– (Australia/Hobart), Main Auditorium
An overview of the history of the Open Data Cube to date, the current status of ODC development and the improvements (and migration headaches) that users can look forward to in upcoming major datacube releases.
The first release of the Australian Geoscience Data Cube was in 2015. Early in 2017 it was renamed the Open Data Cube and the technical direction of the project was progressively delegated to an increasingly independent and multi-organisation Steering Council. This initial phase of rapid development and expansion for the ODC had tapered off by about the release of 1.8.4 in August 2021.
Some early design decisions proved to be innovative and trailblazing with many other open source projects subsequently adopting similar approaches. Other design decisions proved less foresighted, and as early as 2019 the Steering Council were discussing fundamental limitations that would require major architectural changes to properly address.
Discussions around these limitations and the steps required to address them largely stalled until 2022 when the first draft of an ODCv2 road map was published and subsequently refined through consultation with the Steering Council and the broader ODC user community.
The ODCv2 roadmap has three major areas:
1. Enhancements to the ODC index to allow faster search and more efficient indexing and easier index management;
2. Reunifying/re-aligning datacube-core with the odc-geo and odc-stac utilities
3. The ability of the Open Data Cube architecture to scale to significantly larger volumes and rates of data
With a new generation of satellites on the horizon, offering faster return times, higher spatial resolution and higher spectrality (not just open access Landsat Next, but also the hyperspectral satellites) the third area quickly became an important focus of these discussions.
Since the publication of the ODCv2 roadmap, work to implement it has been progressively picking up pace, at both Geoscience Australia and CSIRO.
This talk will provide an overview of the history of the Open Data Cube to date, the current status of ODC development and the improvements (and migration headaches) that users can look forward to in upcoming major datacube releases.
Paul is a Canberra based software engineer. He has been working on and with the Open Data Cube since 2017, originally while working at NICTA, then CSIRO, and since 2021 with Geoscience Australia. He has been a member of the Open Data Cube Steering Council since 2019. In his spare time he brews beer and is working on a science fiction rock opera.