v1.3

Never mind, here we are with a new version just a few days after the last one. This time I have mainly focused on making the mapping procedure easier to understand and more reliable with an improved visualization and cost map for the pathfinding algorithm...

Changelog:

• Removed debug prints.
• Switched to a circular collision box to reduce the required space for the robot to operate.
• Added a collision box indicator and the navigation path to the map visualization.
• Implemented a basic cost map to keep the robot away from obstacles.
• Some nice improvements to the README.

v1.2

After taking a break from development while on vacation, I'm back with a working and (mostly) reliable ICP pose estimate correction for the mapping screen and some nice bug fixes! The app will likely remain at this version for quite a while until I find the time and energy to dive into even more advanced algorithms...

Changelog:

• Heavily tweaked the ICP pose estimate correction to make it (mostly) reliable.
• Fixed a ClassNotFoundException while saving the map with the packaged deb.
• Cancel navigation thread when exiting screen or disabling mapping.
• Some minor tweaks to the README including a really good introduction video to SLAM.

v1.1

While working on the advanced mapping features using ICP, I decided to publish what I currently have working and do an interim release - enjoy it!

Changelog:

• Fixed crashes when going to the background (multi-tasking or leaving the app without closing it)
• Added some SLAM notes to the README for anyone interested.
• Implemented the mapping screen UI including the map visualization.
• Implemented a first and basic iteration of autonomous driving capabilities without using ICP yet.

v1.0

This is the very first public release of BotvacCenter for Android and Linux (.deb) after quite extensive personal testing.

Most bugs should be gone, but the overall debug interface of these Botvac robots is quite unstable, so don't expect a super smooth experierence and be prepared to run some operations multiple times before they succeed.

Unfortunately, cross-compilation isn't yet supported by Kotlin Multiplatform so I can't offer macOS, Windows or iOS builds. If you have a suitable device to run these builds on, feel free to reach out...