July 1, 2020
Agenda
- Component update
- Release
- Community Q&A
- Any in-depth talks
Join Meeting
Meeting ID: 946 175 205
Join using your mobile/desktop
Call in using your phone, find your local number: https://zoom.us/u/aetSYKbiMF
Component update
System Architecture
No updates
OS / NuttX
No updates
Driver
Invensense cleanup and consistency changes according to datasheets.
Commander
Messageing cleanups:
Bigger low battery reaction changes. More main state manipulation. Needs review:
Estimation
Nothing bigger, small cleanups. PX4 is not using latest ecl at the moment and we should update when we think it’s safe.
Python derivation of the EKF is on its way. @Paul_Riseborough ported the Yaw estimator as well and starts testing now:
Integrate sensor data that is coming in at a too high rate. Such that replay logs don’t contain raw mag, baro data when they’re anyways just averaged shortly after by the estimator.
VTOL / Fixed-Wing
TECS maximum throttle instead of 100%. This was proposed before and @Paul_Riseborough had reasons against it so we should sync with him such that we understand the reasoning.
Multicopter
2 Orbit fixes:
MAVSDK
Nothing major, bug fixes and testing with MAVSDK.
MAVROS / DDS / ROS2
@TSC21is working on the MicroRTPS serial support. Such that it can be used simultaneously with MAVLink. There’ll also be some latency and bandwidth profiling such that it’s easier to configure.
Avoidance
No updates
Simulation
Not clear why SITL tests are failing more again.
A fix:
One idea:
We need to work towards having SITL closer to a real flight including default configuration, MAVLink rate configuration, UDP links.
Benjamin Perseghetti: Question about if rebooting SITL is on the radar with the main intent to restart EKF and not have preflight checks failing after a virtual movement of the vehicle. Example use cases a rover moves a drone and then the drone launches. Most of the time this doesn’t succeed. We’d like that use case to be working so it would be useful to have an issue and log. We should look at what is making a reboot necessary and make it more robust or resettable independently. He’d be happy to have a complete reset available as a workaround but we should make this unnecessary.
JFYI @Jaeyoung-Lim
New Java version works now:
UAVCAN
DS-015 document has now a UML diagram for e.g. hotplugging a device and some typical topologies. Focus went from message specs to necessary mechanisms (healthiness, heartbeats, bootloader states, …) such that it’s explained how to integrate.
Work on possibility to use PX4 as a stripped-down CAN node for optical flow.
Hardware call
Finalize FMUv6x specification. Until now there was not much negative feedback so it will be released to the public as a draft and then used if there are no problems reported. Similar to v5x but H7 chip instead of F7 and different pinning.
Payload bus connection with a wider connector (60 pin instead of 40) was decided to not be done.
Release
Like it was mentioned in the dev summit we aim to change the release cycle to be more often, have a forced branch off date and a stabilization phase.
Current release blockers:
- Land detector issue. We have feedback by @nrogelio that he had the vehicle land much slower.
https://github.com/PX4/Firmware/pull/15083 - Oneshot problems
Community Q&A
No additional question
Errata and Feedback
Let me know below if I failed to capture anything the right way, and if there are any updates, or you have feedback on the call format.