Dev Call Aug 15, 2022

Remote ID Update

Issues & Pull Requests



Attendee count (max): 20

UTC1100 - remote id update

  • Many new releases
    • Henry’s having some issues getting it running
  • Things are progressing a lot
  • Discord is active
  • Good out to 1km
    • Didn’t fail out that far
  • Working with BlueMark, Jeff and intel making sure we’re compliant
  • MAVProxy reference implementation updated several times in last week
  • Tested support using mobile phone as GPS for ground station
    • Must have GPS on GCS
    • Testing app to use phone as bluetooth serial port and act as a GPS
  • MissionPlanner support is coming along
    • See screenshot in OpenDronbeID channel
  • Other groups working on QGC groundstation support
  • BoardID and firmware checking merged into master
  • Next step is to prepare for supporting signing
    • Distribution keys etc etc
  • Will not deploy with signed firmware initially
    • Already required in India
    • Assuming it won’t be required under USA rules
  • GPLv3 compliance may be a problem for drones sold for “personal use”
  • D Sastre is working on QGC with Ben Harris’ support
  • Sid’s working on HereLink support
  • D Agar notionally can release to app store
    • Can he actually do a release, ‘though?
    • Maybe ask Don nicely if D Agar can’t release
    • Maybe D Sastre?
  • Probably back-port to 4.2.3
  • New comments in FAA re tamper resistance
    • Discussion had on ASTM workgroup with FAA
    • Declaration-of-conformity people
    • Risk assessment required and explanation of tamper resistance
    • Level we have now might be enough
  • Public keys
    • Three members would have keys
    • Vendors would use own keys and include upstream keys
    • So users can upgrade
    • Transmitter module would also need to be connected
    • PK would be needed on transmitter module too
  • A number of groups have had 1-1 with FAA
    • Substantial of a 6 month extension
    • For US companies there’s a letter-writing campaign spooling up
  • Will vendor PK stuff so Indian companies can use it
  • Option to set the DFU protection bits?
    • Many levels of fuse bits we could set
      • If we set them all it’s about as secure as you can make it
  • What we want to get into 4.2
  • Vendor selling vehicles have to assume they won’t be flown in Friya zones
  • If you are testing the esp32 transmitter module keep in touch as things are changing hour-to-hour
    • Discord!
  • Jeff W has a clean DevBoard implementation going to fab tonight
    • 1 to 2 weeks to get a board, couple of days to assemble.

UTC1126 - AP_NavEKF3: store and use lastKnownPositionD by peterbarker · Pull Request #20484 · ArduPilot/ardupilot · GitHub

  • Constant position mode return non-zero altitude
  • Still need Paul on this one

UTC1132 - AP_BattMonitor: Add FuelLevel_Analog support by TunaLobster · Pull Request #21008 · ArduPilot/ardupilot · GitHub

  • Analogue fuel support
  • Pixhawk1-1M won’t have fuel level by default any more

UTC1138 - Make NucleoH743 SimOnHW by default; break out an include file for SimOnHW by peterbarker · Pull Request #21079 · ArduPilot/ardupilot · GitHub

  • Simonhardware on NucleoH743
  • Currently using a crystal, if we stop that it can be a generic sitl-on-hardware build
  • Merged!

UTC1140 - AP_HAL_ChibiOS: don't make consumption of iomcu rssi dependent on HAL define by peterbarker · Pull Request #21084 · ArduPilot/ardupilot · GitHub

  • Needs to be flashed onto pixhawk1-1M and can then be merged

UTC1140 - Stop including GCS_MAVLink/GCS_MAVLink.h in headers by peterbarker · Pull Request #21145 · ArduPilot/ardupilot · GitHub

  • Need to discuss making things stabler
    • Randy thinks a week or two
    • Tridge thinks sooner
  • Merged
    • Mavlink header cleanups

UTC1140 -

  • Adding fields to mission_current
    • Need to fix the total field

UTC1140 - Take-off throttle slew time by lthall · Pull Request #21158 · ArduPilot/ardupilot · GitHub

  • Take-off throttle slew time
  • Need a variable name….
    • The primary benefit of this PR is making the take-off behaviour independent of the vertical position controller tune. It will also improve the consistency of take-off when different payloads are used.
  • Number based on hover throttle more useful for end user?
    • Leonard: user doesn’t care
      • Softer, double number,
      • Faster, halve number
  • Naming?
    • Takeoff-slew-time
  • Throttle isn’t really something we should be using
    • Should be thrust
  • Having _MAX in the name would be really good as it might not take this time
    • tkoff _slew_time

UTC0005 - Move Scheduler call to compass-cal-update up to AP_Vehicle by peterbarker · Pull Request #21190 · ArduPilot/ardupilot · GitHub

  • Move compass cal update up
  • Merged

UTC0007 - Custom Controller Support 2nd Version by esaldiran · Pull Request #21216 · ArduPilot/ardupilot · GitHub

  • Custom controller support
  • You can have more than one controller
    • Switched by parameter
  • Enabled by default on 2MB boards?!
    • Needs to change
  • Overwrites existing controller
    • Existing controller’s integrators are cleared
    • Can not overwrite things if you feel like it….
  • Allows people to experiment with new controllers
  • Some CI fixes to be done
  • AM] Peter Hall: Flight testing Custom Controller switching - YouTube
  • YouTube
  • MrIampete
  • Flight testing Custom Controller switching
  • Image
  • [10:09 AM] Peter Hall: Even has testing.
  • [10:11 AM] Peter Hall: Emre has code generated from Simulink working too.
  • [10:11 AM] Peter Hall: Needs a little manual copy and paste to get params and such.
  • [10:12 AM] Peter Hall: But its surprisingly easy.
  • [10:12 AM] Josh Henderson: That will be a major change to enable documentation of changes
  • [10:13 AM] Josh Henderson: @amilcar 's work to replicate the controllers in matlab could be tried out.
  • Change away from “CC” as we use that for companion computer

UTC0018 - AP_Mount: all backends send gimbal-device-attitude-status by rmackay9 · Pull Request #21243 · ArduPilot/ardupilot · GitHub

  • Send device_attitude_status from mavlink gimbals
  • Mount_status not being sent any more
  • ROS might break
  • Mav_mount_status was terminally broken
    • 2-axis gimbals worked differently to 3-axis
    • So broken it’s not worth keeping
    • Might break Solo….
  • Randy will merge

UTC0018 - Arming: add parameter and check for runaway missions by mentonin · Pull Request #21294 · ArduPilot/ardupilot · GitHub

  • Add parameter and check for runaway missions
  • Why not fence?
    • Because you get told very late
    • Tin-can fence
  • Km rather than metres?
  • Ship landing - landing somewhere else….
  • Should we default it non-zero
    • Would force people to clear their missions if they’re only flying in loiter
  • QGC’s wipe-after-mission is nasty
  • We can have this on by default

UTC0054 - Plane update

UTC0054 - Copter/Rover update

UTC0104 - autotest: Cache terrain data within autotest directory by peterbarker · Pull Request #21323 · ArduPilot/ardupilot · GitHub

  • Install tiles in repository for autotest
  • Merged

UTC0113 - Copter: logging disentangle and correct bugs and rename bitmasks by Hwurzburg · Pull Request #21372 · ArduPilot/ardupilot · GitHub

  • Correct bugs and rename bitmasks
  • Logging pids without attitude is strange?
    • I enabled PID logging but didn’t get PIDs?!
  • PH wants to see some logs with correct rates
  • Henry will test

UTC0118 - Plane: only trigger fence action for new breaches, don't allow mode change in breach unless fence option set. by IamPete1 · Pull Request #21371 · ArduPilot/ardupilot · GitHub

  • New fence parameter option to allow changing modes
  • Randy’s happy with having the fence option
  • Can be merged after the autotest changes are made

UTC0131 - AP_Scripting: save some flash, allow uint32 from negative numbers by IamPete1 · Pull Request #21391 · ArduPilot/ardupilot · GitHub

  • Save some flash, allow uint32_t from negative numbers
    • Pass in a negative number to get a large number
    • Shift in instead?
      • Maybe create a constructor to do the shift for you?
        • uint32(0xdead, 0xbeef)
          • Mavlink javascript bindings do this
    • Larger precision numbers in LUA would be nicer
    • Creating uint32_t on the fly in LUA is very, very costly
      • Limited control of a motor
      • Declare at top of script and then create once with magic numbers
        • Went very fast after that
    • Might actually save time….
      • Millis is very slow in lua?
        • Not as slow as generic as it uses a slightly different path
  • Saving 1kB of flash for other changes
  • PH will go back and to the two-argument version
    • Unless he really does want the negative one
  • He’ll also have a look at going double-precision in LUA

UTC0149 - Tools: added required fields for libraries by tridge · Pull Request #21411 · ArduPilot/ardupilot · GitHub

  • Param fix
  • Merged

UTC0150 - SRV_Channel: use enumeration for servo input type by peterbarker · Pull Request #21396 · ArduPilot/ardupilot · GitHub

  • Change to make range the default as zero

UTC0158 - Identical Waypoints lead to internal errors in SCurve · Issue #21409 · ArduPilot/ardupilot · GitHub

  • User input causes internal errors
  • 3D interface in UGCS is nice
    • Can’t download missions, ‘though?!
  • Peter will chase up with Leonard

UTC0158 - FilterTool: graph split and phase wrap option by IamPete1 · Pull Request #21419 · ArduPilot/ardupilot · GitHub

  • Filter tool changes
  • Splits to make things look more like bode plots
  • Merged

UTC0158 - ArduPlane: remove unused limit_length calculation in overshoot by mirkix · Pull Request #21447 · ArduPilot/ardupilot · GitHub

  • Remove unused variables
  • Merged!

UTC0205 - KakuteH7Mini-Nand rename by andyp1per · Pull Request #21451 · ArduPilot/ardupilot · GitHub

  • Rename KakuteH7Mini
  • Bumped to DevCallEU

UTC0207 - open votes

  • Everyone on call has already voted
    • Charlie would accept membership if it was offered
    • Which the vote says it should

UTC0210 - close