rroche
July 30, 2020, 3:00pm
1
Status update by component/project
Roadmap, and Release discussion
Community Q&A
In-Depth discussions
Meeting ID : 946 175 205Join using your mobile/desktop
Zoom is the leader in modern enterprise video communications, with an easy, reliable cloud platform for video and audio conferencing, chat, and webinars across mobile, desktop, and room systems. Zoom Rooms is the original software-based conference...
Call in using your phone, find your local number : Zoom International Dial-in Numbers - Zoom
UAVCAN workgroup met, see UAVCAN below.
uORB optimizations
uOrb top output sorted alphabetically
work queue status sorted alphabetically
PX4:px4_firmware_nuttx-8.2 ← PX4:px4_firmware_nuttx-8.2-pr-single_wire
opened 04:24PM - 04 Aug 20 UTC
Waiting on upstream for feed back on format changes. This is the code changes w… ithout formatting changes.
PX4:master ← PX4:frsky_telemetry_singlewire_inverted_f7
opened 07:01AM - 05 Aug 20 UTC
Requires https://github.com/PX4/NuttX/pull/112.
https://github.com/PX4/Firmware/pull/15415
https://github.com/PX4/Firmware/pull/15385
NuttX conference
(click for link)
PX4:master ← PX4:pr-paw3902_start
opened 03:25PM - 26 Mar 20 UTC
PX4:master ← PX4:commander_arm_disarm_reason
opened 09:02AM - 04 Aug 20 UTC
- I've come across multiple logs where the disarming reason was not clear and un… expected.
- There is an increasing number of components that can trigger arming/disarming (e.g. kill-switch or failure detector).
Adding the latest arming/disarming reason to vehicle_status makes things explicit.
PX4:master ← PX4:pr-ekfSymPyCovariancePrediction
opened 08:37AM - 23 Jul 20 UTC
This is the covariance prediction part of https://github.com/PX4/ecl/pull/868 wh… ich is being implemented in a series of smaller PR's to make review and testing easier.
See https://github.com/PX4/ecl/pull/868 for earlier discussion.
Difference testing agains the matlab output for 100 randomised input P matrices showed difference ratios in two tests over the nominal test Pass/Fail ratio of 5E-5 that had been set, eg:
Fail: Covariance Prediction max diff fraction = 6.775925e-05 , old = 4.104861e+23 , new = 4.105139e+23 , location index = 4,6
Fail: Covariance Prediction max diff fraction = 1.048742e-04 , old = 1.940318e+11 , new = 1.940521e+11 , location index = 3,16
However the results are acceptable given the unrealistically large values for those entries induced by those tests. For the other 98 randomised tests, not a single matrix entry exceeded the 5E-5 difference ratio.
SITL Test log for gazebo_plane: https://logs.px4.io/plot_app?log=49cd1e69-bdd8-492f-99a8-0f2f94c35105
PX4:master ← kamilritz:matrix_lib_usage
opened 07:20PM - 27 Jul 20 UTC
https://github.com/PX4/Firmware/pull/15425
PX4:master ← PX4:pr-FlightTaskAuto-fix-VTOL-transition-upstream
opened 04:18PM - 03 Aug 20 UTC
Fixes https://github.com/PX4/Firmware/issues/15409, which was introduced by http… s://github.com/PX4/Firmware/pull/15358.
It's certainly not the cleanest solution, but I guess it serves as a quick fix, as we already have the disable_weather_vane flag read in the FlightTask for a very similar purpose (to enable the aligning with the transition wp, weather vane needs to be switched off).
Maybe renaming the flag from "disable_weather_vane" to "enable_navigator_yaw_control" / "enforce_navigator_yaw_setpoint" makes it a bit cleaner. Or I guess even better it would be to have the alignment logic inside the flight task, and not Navigator.
Land detector PR has been merged
PX4:master ← PX4:pr-land_detector_mc_ground_contact
opened 06:32PM - 10 Jun 20 UTC
Opening this for discussion/testing of possible fixes for https://github.com/PX4… /Firmware/issues/13124, https://github.com/PX4/Firmware/issues/11683, https://github.com/PX4/Firmware/issues/13507.
https://logs.px4.io/plot_app?log=4b65cf44-80cb-4910-bc2a-151028afe9da
https://github.com/PX4/Firmware/pull/15370
No update this week. Things are working well.
PX4:master ← PX4:pr-standard-vtol-hover-throttle
opened 02:21PM - 04 Aug 20 UTC
I have no idea what I'm changing here but it brings the hover throttle from 0.85… to about 0.6 which is much more controllable but still realistic for that sort of airframe.
PX4:master ← PX4:pr-wind-updaterate
opened 08:51AM - 05 Aug 20 UTC
Realtime factor on the standard_vtol has been falling down to ~0.3 when run toge… ther with the wind plugin.
This is due to the fact that wind is being published at the full update of the simulation and being used in multiple components (e.g. `liftdrag_plugin`, `airspeed_plugin`, `gazebo_mavlink_interface`, `gazebo_motor_model`)
This PR constrains the wind topic publish rate by default to 2Hz, increasing the realtime factor even when wind is being published to ~0.97
PX4:master ← PX4:pr-fix-airspeed-sensor
opened 09:04AM - 04 Aug 20 UTC
While wind topic names have been fixed in https://github.com/PX4/sitl_gazebo/pul… l/552, airspeed sensor was still subscribing to the wrong topic name.
Fixes https://github.com/PX4/sitl_gazebo/issues/556
no update this week
The fix to the UART link of the microRTPS bridge was finally merged. microRTPS is now fully functional with serial:
PX4:master ← PX4:pr-micrortps_client_various_fixes
opened 04:12PM - 16 Jul 20 UTC
Currently the microRTPS bridge UART link is broken and it's resulting on continu… ous CRC errors on the agent side that after some seconds close the link. This is being investigated.
Also, it was verified that a stack increase was required to avoid the `micrortps_client` of causing and hardfault and rebooting the flight controller.
Agent protocol splitter
The protocol splitter allows propagating RTPS and MAVLink packets in the same link and demultiplexes them on the agent side
Finalizing details it’s scheduled to go live this week
After launch, it’s planned to be hosted next week as a standalone repository.
@PavelKirienko and @thirtytwobits @dagar to implement a “register interface” and “port ID configuration” on the PX4 side for the UAVCAN v1.0 client.
Catch up on all the developments below
@rroche
Join Zoom Meeting
One tap mobile
+14086380968,567710856# US (San Jose)
+16468769923,567710856# US (New York)
Dial by your location
+1 408 638 0968 US (San Jose)
+1 646 876 9923 US (New York)
+1 669 900 6833 US (San Jose)
Meeting ID: 567 710 856
Find your local number: https://zoom.us/u/aetSYKbiMF
The last remaining blocker (Land Detector Issue ) has been merged.
We are moving forward with v1.11-RC1
Calling for help from the community, we need tests!
After the v1.11 release, what’s next?
More regular release schedule going forward.
fixed periods for mater to beta, beta to stable
Shorter cycles
Dedicated test team (sponsored by Dronecode)
No discussions today.
Let me know below if we failed to capture anything the right way, and if there are any updates (not present here), or you have feedback you would like to share with the dev-team.
1 Like
NuttX online workshop had a ton of interesting talks. Thanks for the link!
Btw, how did NuttX workshop come up during the dev call? Just curious
rroche
August 26, 2020, 6:00pm
3
How did it come up?
I was following the event news and made sure the PX4 community was fully aware of the event, and decided to make an announcement during the dev call, I think I actually brought it up on a few dev-calls
1 Like
uORB optimizations
NuttX conference
support for paw3902 optical flow sensor
Added Arming/disarming reason for logging
Land detector PR has been merged
Agent protocol splitter
DS-015 work continues, team keeps meeting bi-weekly
The DS-015 team is now joined by @PavelKirienko and @thirtytwobits
On-going efforts by @dagar to implement a “register interface” and “port ID configuration” on the PX4 side for the UAVCAN v1.0 client.
