With the 1.12 release of Autoware.AI just about out the door (), it is time to start thinking about what you want to see in Autoware.AI 1.13.
Although the dates for the release and how long the development period are not yet set, we plan to do as we did for 1.12 and decide on planned scope and features upfront to prevent the scope exploding as the release draws near. So let us know if you are planning to develop something, or if there is some feature you urgently need (we will try to balance requirements in our planning).
Remember that asking for a huge feature in a single chunk is less likely to be feasible than asking for incremental development of bits of that feature over a series of releases. The 1.12 release suffered a bit from big monolithic features, so this time around we will either need to narrow the scope or accept a longer development cycle.
I’ve been slowly working on ROS2 support - amentifying packages. Within the 1.12 it was difficult due to repo restructuring work, but I think for the 1.13 it is feasible to start useing ament and (probably) ROS2 in general.
Geoff
I think Autoware.auto and Autoware.ai should be working together as soon as possible.
then we/user can get a workable stack with autoware.auto inside. this will promote evolution.
new features and refactoring works can start from ROS2 side.
The first release of Autoware.Auto will be published soon after Autoware.AI. However, we needed some features to be stabilized in Autoware.AI before we could combine the two projects (most importantly, the move to Gitlab and support for ROS 1 Melodic so the bridge can be used).
I’ll make a proper announcement once we’ve finalized the documentation for Autoware.Auto, this should happen in about a week.
We welcome contributions from the community in the form or code reviews or bugfixes to make our first release of Autoware.Auto as solid as we can. After the first release we will be accepting new features from the community.
), it is time to start thinking about what you want to see in Autoware.AI 1.13.