Dev Call March 30, 2026

CraigElder · March 30, 2026, 11:17pm

CI Work flows

Release Update

github.com/ArduPilot/ardupilot

ArduPilot 4.7 issues list

opened 12:30AM - 07 Mar 26 UTC

rmackay9

DevCallTopic ReleaseAdmin

This is a list of issues discovered during ArduPilot 4.7 testing. [Issues for 4….6 are here](https://github.com/ArduPilot/ardupilot/issues/28612) ### Reports requiring investigation - [ ] BRD_SAFETY_DFLT = 0 should force BRD_SAFETYOPTION = 0? should update param desc? ([discussion](https://discuss.ardupilot.org/t/rover-4-6-3-arming-issues-when-trying-to-arm-second-time/142747/12)) - [ ] Mount POI lock not included in default firmware ([discussion](https://discuss.ardupilot.org/t/copter-4-7-0-beta1-available-for-beta-testing/142738/43)) - [ ] Loiter feels slower, like looks inaccurate position hold. Likely due to https://github.com/ArduPilot/ardupilot/pull/31848 consider how we can make it feel closer to 4.6 automatically. Or provide instructions on parameter changes. Added by @IamPete1 on behalf of a partner company ### Confirmed issues - [ ] Altitude estimate offset when switching between baro and rangefinder source ([discussion](https://discuss.ardupilot.org/t/copter-4-7-0-beta1-available-for-beta-testing/142738/28), [issue](https://github.com/ArduPilot/ardupilot/issues/32612), [PR](https://github.com/ArduPilot/ardupilot/pull/32400)) - [ ] ArcWaypoints listed as WIP? (reported by lthall, [PR](https://github.com/mavlink/mavlink/pull/2444)) -- @hunt0r handling this - [ ] https://github.com/ArduPilot/ardupilot/issues/32491 - [ ] RC_OPTIONS = 1 (Ignore RC Receiver) is not being respected ([discussion](https://discuss.ardupilot.org/t/rc-options-to-1-make-copter-yaw-spinning/142948), [PR](https://github.com/ArduPilot/ardupilot/pull/32488)) - [ ] https://github.com/ArduPilot/ardupilot/pull/32382 ### Copter Specific Issues - [ ] https://github.com/ArduPilot/ardupilot/issues/31968 ### Plane Specific Issues - [ ] https://github.com/ArduPilot/ardupilot/issues/32606 - [ ] https://github.com/ArduPilot/ardupilot/issues/32518 -- @IamPete1 or @samuelctabor to resolve ### Rover Specific Issues - [ ] https://github.com/ArduPilot/ardupilot/issues/32444 ### Mission Planner Issues - [ ] https://github.com/ArduPilot/MissionPlanner/issues/3693 - [ ] Extended Tuning needs updating ([discussion](https://discuss.ardupilot.org/t/speedybee-f405-v4-parameter-not-shown-in-mission-planner/142797/3), [issue](https://github.com/ArduPilot/MissionPlanner/issues/3678), [PR](https://github.com/ArduPilot/MissionPlanner/pull/3685)) ### QGC Issues - None ### Wiki Issues - None ### Enhancements or Issues that may not be resolved before the stable release - [ ] Copter's SmartRTL always rises 2m before landing ([discussion](https://discuss.ardupilot.org/t/smartrtl-rises-to-2m-after-reaching-home-before-landing/143126/2)) -- present in 4.6 so not a blocker ### Resolved issues - [x] https://github.com/ArduPilot/ardupilot/issues/32511 -- could not reproduce, no response from user, created [security wiki page](https://ardupilot.org/dev/docs/security-landing-page.html) and improved [secure firmware instructions](https://github.com/ArduPilot/ardupilot/tree/master/Tools/scripts/signing) - [x] CUAV Pixhawk V3C not detecting IST8310 compass ([discussion](https://discuss.ardupilot.org/t/cuav-pixhawk-v3-ist8310-not-detected-with-4-7beta1/142785), [PR](https://github.com/ArduPilot/ardupilot/pull/32462)) -- resolved for 4.7.0-beta3 - [x] https://github.com/ArduPilot/ardupilot/issues/32506 ([PR](https://github.com/ArduPilot/ardupilot/pull/32508)) -- resolved ahead of 4.7.0-beta3 - [x] https://github.com/ArduPilot/ardupilot/pull/29087 missing from release notes ([discussion](https://discuss.ardupilot.org/t/plane-4-7-0-beta1-available-for-beta-testing/142742/9)) -- @IamPete1 updated the plane 4.7.0-beta1 release notes - [x] Plane Follow's FOLLP_SIM_TlF_FN should have upper case L ([PR](https://github.com/ArduPilot/ardupilot/pull/32456)) -- resolved for 4.7.0-beta2 - [x] UM982 GPS-for-yaw on CUAV V6X not working ([discussion](https://discuss.ardupilot.org/t/ardurover-4-7-0-new-issues-with-um982-moving-baseline/142916)) -- resolved by updating the UM982 firmware - [x] https://github.com/ArduPilot/ardupilot/issues/31584 -- resolved for 4.7.0-beta2 - [x] QGC altitude slider not working ([discussion](https://discuss.ardupilot.org/t/copter-4-7-0-beta1-available-for-beta-testing/142738/15)) -- fixed for daily build on 18Mar2026 - [x] QGC takeoff slider not working ([discussion](https://discuss.ardupilot.org/t/copter-4-7-0-beta1-available-for-beta-testing/142738/15)) -- fixed in QGC 14148 - [x] QGC Parameters need updating (@rmackay9 and @DonLakeFlyer discussed if dev-team channel, [PR](https://github.com/mavlink/qgroundcontrol/pull/14148)) -- fixed in QGC daily on 17-Mar-2026 - [x] EKF's use of rangefinder and baro never switches back to rangefinder after switching to baro ([discussion](https://discuss.ardupilot.org/t/copter-4-7-0-beta1-available-for-beta-testing/142738/15)) -- user error - [x] update PSC_ACCZ_I and PSC_ACCZ_P tuning advice https://ardupilot.org/copter/docs/initial-tuning-flight.html#test-althold ([PR](https://github.com/ArduPilot/ardupilot_wiki/pull/7482)) -- wiki updated ahead of 4.7.0-beta2 - [x] https://github.com/ArduPilot/ardupilot/pull/32381 -- resolved for 4.7.0-beta2 - [x] ATC_ACC conversion didn't work ([discussion](https://discuss.ardupilot.org/t/copter-4-7-0-beta1-available-for-beta-testing/142738/8)). Should we remove the "configured" checks ([code](https://github.com/ArduPilot/ardupilot/blob/master/libraries/AC_AttitudeControl/AC_AttitudeControl.cpp#L1161)). Caused by installing 4.7.0-dev prior to release of beta1 - [x] https://github.com/ArduPilot/ardupilot/pull/32428 -- resolved for 4.7.0-beta2

Issues & Pull Requests

github.com/mavlink/mavlink

common.xml: change NAV_VTOL_LAND_OPTIONS enumeration values

master ← peterbarker:pr/vtol-land-options

opened 08:08AM - 30 Mar 26 UTC

peterbarker

+4 -6

Please do not merge until discussed at ArduPilot DevCall. I believe we just c…ompletely screwed up our synchronisation with mavlink/mavlink/master (patches were merged there, then modified before merging into ardupilot/mavlink/master. ArduPilot obviously uses the values being merged in here. I haven't found any other project using values from this enumeration (QGC, PX4-AutoPilot, MissionPlanner, betaflight, Paparazzi)

github.com/ArduPilot/ardupilot

autotest: make blimp manual flight test more robust

master ← tpwrules:pr/blimp-manual

opened 01:14AM - 14 Oct 25 UTC

tpwrules

+16 -6

The test would sometimes fail because the blimp would not coast to the next chec…kpoint or would fly in the wrong direction. * Set lower and constant `SIM_SPEEDUP` for more consistent RC inputs * Make north coast checkpoint (`ttl`) half as far * Come to a near-complete stop before heading east, as otherwise we will still have some north velocity that may push us off course, then start measurement from the stopped location * Again make east coast checkpoint (`ttr`) half as far * Allow more time for RTL from now-farther distance

github.com/ArduPilot/ardupilot

AP_NavEKF3: add EK3_MAG_CAL=7 ground and in-flight mode

master ← andyp1per:pr-mag-cal-ground-inflight

opened 02:53PM - 14 Feb 26 UTC

andyp1per

+8 -5

## Summary Add EK3_MAG_CAL=7 (GROUND_AND_INFLIGHT) magnetometer calibration mod…e that combines ground learning with takeoff protection. ### Problem Existing mag cal modes have a gap for vehicles with significant motor-to-compass coupling: - **Mode 3 (AFTER_FIRST_CLIMB):** No ground learning — can't adapt to battery-swap mag changes - **Mode 4 (ALWAYS):** Learns on ground, but also learns during takeoff when motor interference corrupts offset estimates ### Solution Mode 7 uniquely combines: - **Ground learning** (like mode 4) — adapts to battery swaps and environmental changes before takeoff - **Takeoff protection** (like mode 3) — freezes mag cal during initial climb when motor interference is strongest - **In-flight learning** — resumes 3-axis fusion after the first in-air field and yaw reset completes ### Flight verification Tested on a small fast drone with compass mounted close to motors: - Yaw divergence reduced from 24° to 5.3° - Takeoff body offset change reduced from 56 mG to 0 mG ### Changes - `AP_NavEKF3_core.h`: Add `GROUND_AND_INFLIGHT = 7` to MagCal enum - `AP_NavEKF3_Control.cpp`: Add mode 7 logic to `magCalRequested` and `magCalDenied` - `AP_NavEKF3.cpp`: Update parameter description and @Values ## Test plan - [x] SITL build passes - [ ] Existing autotest suite passes (no behavioral change unless EK3_MAG_CAL=7 is explicitly set) - [ ] Flight test with EK3_MAG_CAL=7 on vehicle with motor-compass coupling

github.com/ArduPilot/ardupilot

AP_NavEKF3: use onGroundNotMoving for zero velocity fusion

master ← andyp1per:pr-ekf3-zervel

opened 08:55PM - 08 Mar 26 UTC

andyp1per

+243 -11

## Summary - Fuse synthetic zero velocity when stationary on the ground to cons…train gyro bias and Z-axis accel bias learning. XY accel biases remain unobservable on the ground and are inhibited by dvelBiasAxisInhibit - Applies in all aiding modes: AID_NONE uses zero velocity instead of the large _noaidHorizNoise fake velocity, and AID_RELATIVE/AID_ABSOLUTE modes fuse zero velocity when no real velocity sensor data is available (e.g. optical flow configured but stationary) - Gated on onGroundNotMoving to avoid injecting zero velocity when the vehicle is being moved - **Inhibit all accel bias learning when on ground and moving** (e.g. carried on a utility belt or on a boat). Previously only the gravity-alignment check was used, which allowed the gravity-aligned axis to learn biases from external motion. Adding onGroundNotMoving to dvelBiasAxisInhibit prevents incorrect biases that could cause velocity drift after takeoff - Broaden updateMovementCheck() to run unconditionally when on ground (previously gated on yaw source type) ## Test plan - [x] EK3_AccelBiasInhibitOnGroundMoving - verifies bias NOT learned during ground movement (SIM_SHIP), and IS learned when stationary. **Fails on master, passes with fix** - [x] EK3_AccelBiasZeroVelOptFlow - verifies Z-bias convergence with optical flow config (no GPS) via zero velocity fusion - [x] EK3AccelBias - existing test still passes (no regression) - [x] Verify no regression in normal GPS flight

github.com/ArduPilot/ardupilot

Clarify distance to destination & cross track error

master ← hunt0r:issue32295-clarify-distance-to-destination

opened 03:32PM - 16 Mar 26 UTC

hunt0r

+45 -43

## Summary No change to compiled output: This PR renames a field to add a uni…t suffix, and changes comments. The motivation for this PR is preparation for ARC waypoints (issue #32295 ). When ARC waypoints arrive, the crosstrack error will need significant improvement. There is nothing wrong with straight-line distance to destination, but we don't want anyone to accidentally confuse it with along-path distance to destination (which will be different during ARC navigation paths). ## Testing (more checks increases chance of being merged) - [x] Checked by a human programmer - [x] Tested in SITL (via CI autotest) - [ ] Tested on hardware - [ ] Logs attached - [ ] Logs available on request - [ ] Autotest included ## Description Add suffix so that the WPNav method is named `crosstrack_error_m()`. Comment fixes: - Clarify the sign of crosstrack error values. - Remind that distance to destination is "straight line" (not along-path, which will be different when ARC waypoints exist). (Originally this PR contained #32616 but that was spun out per reviewer request.)

github.com/ArduPilot/ardupilot

Tidy DPS280/DPS310 handling

master ← peterbarker:pr/dps310-for-realz

opened 02:06AM - 17 Mar 26 UTC

peterbarker

+24 -31

## Summary Stops using a state variable to record what sort of sensor we are;… use the class structure for that. ## Testing (more checks increases chance of being merged) - [x] Checked by a human programmer - [ ] Tested in SITL - [x] Tested on hardware - [ ] Logs attached - [x] Logs available on request - [ ] Autotest included ## Description we don't need to have a flag for which type we are as we are already an instance of DPS310 Tested on a YJUAV which claims to have a DPS310 Tested on a Qiotek ZealotF427 which claims to have a DPS280 Extracted from a larger PR which does a whole bunch of factoring which can take advantage of these changes. ``` Board AP_Periph antennatracker blimp bootloader copter heli iofirmware plane rover sub CubeOrange-periph-heavy -72 * Durandal -64 -72 * -72 -72 -72 -72 -72 Hitec-Airspeed * * KakuteH7-bdshot -72 -72 * -72 -64 -64 -64 -64 MatekF405 * * * * * * * * Pixhawk1-1M-bdshot * * * * * * * SITL_x86_64_linux_gnu * * * * * * * YJUAV_A6SE 112 112 * 112 112 112 112 112 f103-QiotekPeriph * * f303-MatekGPS 104 * f303-Universal -64 * iomcu * mindpx-v2 -72 -64 * -72 -72 -64 -72 -64 revo-mini * * * * * * * * skyviper-v2450 * speedybeef4 * * * * * * * * ```

github.com/ArduPilot/ardupilot

RC_Channel: understand ignore-receiver when reading rc inputs

master ← peterbarker:pr/fix-rc-input-processed

opened 01:50AM - 19 Mar 26 UTC

peterbarker

+31 -12

## Summary Corrects problem where we consume invalid channel values when thro…ttle-failsafe is ignored but we also have RC_OPTIONS set to 1 to ignore the receiver ## Testing (more checks increases chance of being merged) - [x] Checked by a human programmer - [x] Tested in SITL - [ ] Tested on hardware - [ ] Logs attached - [ ] Logs available on request - [ ] Autotest included (it is coming) ## Description previously this option only gated whether we read data from the HAL into our radio_in state. but that would mean read_input would still return true, even if we didn't actually read data from the receiver into the radio state. And that means everything in (e.g. Copter::read_radio thinks we have good channel data when we absolutely do not, meaning has_valid_input is true. Which is bad.

github.com/ArduPilot/ardupilot

Tradheli: rework of Rotor Speed Control library to improve code functionality and readability

master ← bnsgeyer:pr-rsc-rework

opened 03:24AM - 22 Mar 26 UTC

bnsgeyer

+267 -266

## Summary This PR removes RSC calls from the heli frames and puts as much of… it into the RSC library. ## Testing - [X] Checked by a human programmer - [x] Tested in SITL - [ ] Tested on hardware - [ ] Logs attached - [ ] Logs available on request - [ ] Autotest included ## Description This is a pretty big rework of the Rotor Speed Controller. It mainly affects how the heli frames interact with the RSC library. The calls to the library were reduced to the following initialize - initializes the servo configure - sets the parameters that could affect the RSC operation for both unarmed state and armed state update - updates the RSC parameters that allowed to be updated when the aircraft is armed. Conducts all calculations of the rotor state and control output output_to_motors - writes the control output to the motor The setpoint being set from the copter codebase was removed and is set internally to the RSC. Only the passthrough mode value is not passed The control mode was switched to desired spool state which is now passed to the RSC the update method passes back the spool state rather than spool state being determined in output logic. This removes the use of run up complete and spool down complete calls outside of RSC and keeps the spool state decision within the RSC.

github.com/ArduPilot/ardupilot

waf: Drop support for end-of-life Python 3.8

master ← cclauss:drop-support-for-python3.8

opened 02:45PM - 23 Mar 26 UTC

cclauss

+13 -6

## Summary Blocked by: * #32562 Both Python 3.8 and 3.9 are [end-of-life](h…ttps://devguide.python.org/versions), but we need to continue to support Py3.9 until the 31 August 2026 EOL of Debian 11 Bullseye Standard Support, and we drop support for Cygwin or * cygwin/cygwin-install-action#55 and * #32520 are fixed and merged. Improvements: https://docs.python.org/3/whatsnew/3.9.html ## Testing (more checks increase the chance of being merged) - [x] Checked by a human programmer - [ ] Tested in SITL - [ ] Tested on hardware - [ ] Logs attached - [ ] Logs available on request - [ ] Autotest included ## Description

github.com/ArduPilot/ardupilot

AHRS: check backend return values when setting common origin

master ← garvit000:fix-ahrs-common-origin

opened 09:35AM - 25 Mar 26 UTC

garvit000

+15 -9

## Summary Fix handling of common origin across AHRS backends by checking ret…urn values when setting origin. ## Testing (more checks increases chance of being merged) - [ ] Checked by a human programmer - [X] Tested in SITL - [ ] Tested on hardware - [ ] Logs attached - [ ] Logs available on request - [ ] Autotest included ## Description Previously, done_common_origin was set to true even if one or more backends failed to set the origin. This change ensures that done_common_origin is only set when all available backends successfully set the origin. Addresses issue #32539

github.com/ArduPilot/ardupilot

.github: move non building job to ubuntu-slim

master ← khancyr:action_slim

opened 04:06PM - 25 Mar 26 UTC

khancyr

+2 -2

## Summary Unclog our CI by using ubuntu-slim instance for non building jobs …: https://docs.github.com/en/actions/reference/runners/github-hosted-runners . Those are 1cpu but should be enought for python docs/python checking and will free up machine for building. Those machine are already docker containers so we cannot use Ardupilot containers <img width="797" height="418" alt="Capture d’écran du 2026-03-25 17-20-36" src="https://github.com/user-attachments/assets/55b4bbb3-a8cd-41e0-be9b-d79ddc19669b" /> ^ show it works by launching the machine when all the other test are long on wait list ! ## Testing (more checks increases chance of being merged) - [ X] Checked by a human programmer - [ X] Tested in SITL - [ ] Tested on hardware - [ ] Logs attached - [ ] Logs available on request - [ ] Autotest included

github.com/ArduPilot/ardupilot

Unit tests for SITL::Battery

master ← hunt0r:issue10050-unit-tests-for-battery

opened 04:05AM - 26 Mar 26 UTC

hunt0r

+285 -6

## Summary Contributes unit tests for SITL::Battery to document & enforce the… behavior. One behavior change is made and clearly flagged for review: **This PR makes it impossible to set a battery's voltage higher than its max voltage.** ## Testing (more checks increases chance of being merged) - [x] Checked by a human programmer - [x] Unit tested (this PR creates the passing unit tests) - [ ] Tested in SITL - [ ] Tested on hardware - [ ] Logs attached - [ ] Logs available on request - [ ] Autotest included ## Description The first code-change is adding an optional overload of the battery's main time-stepping function (named `set_current()`) which accepts a specified "now" time instant. No existing calls to `set_current()` use this new option; Users who do not receive identical behavior as before: "now" is provided by `AP_HAL::micros64()`. The benefit of this new time-stepper is that unit tests can run (much) faster than wall-clock time. The second code-change is to prevent initializing the voltage to a value higher than the battery's maximum voltage. This looks like a simple bug, is fixed with an obvious one-liner. (All other unexpected behavior can be discussed via future PRs.) Finally, a strong set of unit tests is created which demonstrates the functionality. These tests will be especially valuable for upcoming discussions, changes, and refactors.

github.com/ArduPilot/ardupilot

Upgrade Cygwin build to use Python 3.12

master ← tpwrules:pr/cygwin-upgrade

opened 09:27PM - 28 Mar 26 UTC

tpwrules

+3 -48

## Summary Migrate to 3.12 so we can move past 3.8 as a project. ## Testin…g (more checks increases chance of being merged) - [X] Checked by a human programmer - [X] Tested in SITL - [ ] Tested on hardware - [ ] Logs attached - [ ] Logs available on request - [ ] Autotest included ## Description Tested on Windows 10 that the install script works (after completely deleting Cygwin and other relevant packages) and the resulting environment can build and run SITL. Also confirmed that it can build an identical CubeOrange Copter binary to Linux. Further confirmed that GDB works in the terminal for debugging. In the possible event that Cygwin doesn't move past Python 3.12 when that goes EOL, we can probably use a native Windows Python install as it's only needed at build time and Cygwin can work with Windows executables just fine. But it may have funny implications for the build system and it's not necessary to get that complicated yet. It also may be possible to cross-compile from Linux but this may complicate debugging. There are more details in the commit messages too. @robertlong13 kindly check that everything is kosher with Mission Planner

github.com/ArduPilot/ardupilot

Tools: add MAVLink COMP_METADATA_TYPE_PARAMETER JSON emitter

master ← DonLakeFlyer:compinfo-parameter-metadata

opened 08:31PM - 29 Mar 26 UTC

DonLakeFlyer

+366 -3

## Summary Add a new parameter metadata emitter that outputs `compinfo-parame…ter.json` in the MAVLink component information parameter schema format. Fixes #32596 ## Testing (more checks increases chance of being merged) - [x] Checked by a human programmer - [ ] Tested in SITL - [ ] Tested on hardware - [ ] Logs attached - [ ] Logs available on request - [x] Autotest included ## Description This adds a new `--format mavlink_compinfo` emitter to `param_parse.py` that generates `compinfo-parameter.json` alongside the existing `apm.pdef.json`. The output follows the MAVLink `COMP_METADATA_TYPE_PARAMETER` JSON schema format, making ArduPilot parameter metadata consumable by GCS applications that support the MAVLink component information protocol. **This is phase 1.** The `type` field is hardcoded to `"Float"` and the `default` field is not yet emitted. Extracting actual parameter types and defaults from C++ source requires parsing `.h` file declarations which is a larger change. A follow-up PR will add type and default support. This is also a first step along the road to support COMPONENT_INFORMATION protocol. ### What is included - **New emitter** (`mavlink_compinfo_jsonemit.py`): Outputs flat `parameters` array with field mappings: DisplayName→shortDesc, Description→longDesc, User→category, Range→min/max, Values→array of {value, description}, Bitmask→array of {index, description}, plus units, increment, rebootRequired, readOnly, volatile, group. - **Local extended schema** (`compinfo-parameter.schema.json`): Based on the upstream MAVLink `parameter.schema.json` with an added `readOnly` boolean field (ArduPilot extension). Name pattern widened to 17 chars for 4 ArduPilot params that exceed the MAVLink 16-char spec limit. - **Schema validation tests** (`test_compinfo_parameter_schema.py`): 6 unit tests validating the schema and generated output using `jsonschema`. - **Build pipeline integration**: `build_parameters.sh` updated to copy `compinfo-parameter.json` and `.xz` to the firmware server output directory for all 7 vehicles. - **CI integration**: New `compinfo-parameter-validation` target in `build_ci.sh` and `test_scripts.yml` workflow matrix that generates and validates `compinfo-parameter.json` for all vehicles. ### Dependencies The schema validation tests require the `jsonschema` Python package. This has been added to `install-prereqs-ubuntu.sh`, but the CI Docker image (`ardupilot/ardupilot-dev-base:v0.1.3`) will need to be rebuilt to include it. In the interim, `build_ci.sh` does an inline `pip install jsonschema` so CI will work on the current image. ### AI Disclosure This contribution was AI-assisted (GitHub Copilot + Claude Opus 4.6). All code has been reviewed and tested by the human submitter (as well as Copilot and Opus 4.6)

github.com/ArduPilot/ardupilot

Add simulations for Topotek, Viewpro, AVT and Siyi camera/gimbals

master ← peterbarker:pr-claude/mount-camera-info-refactor-simulation

opened 07:18AM - 30 Mar 26 UTC

peterbarker

+2564 -15

## Summary Subset of the patches in https://github.com/ArduPilot/ardupilot/pu…ll/32515 (all patches in that PR are on top of this one) Adds simulations for several gimbals, adds autotests to add regressions for these gimbals. ## Testing (more checks increases chance of being merged) - [x] Checked by a human programmer - [x] Tested in SITL - [ ] Tested on hardware - [ ] Logs attached - [ ] Logs available on request - [x] Autotest included ## Description There are several bugs in these gimbals that are in need of fixing, but I think we can simplify that PR by merging some simulation changes first. The only changes to the real code here allows for a mavlink node to request `CAMERA_INFORAMTION` for a specific gimbal (rather than getting an instance for each!) I do intend to revisit the whole "param2 gives you a single instance" thing properly, but that's a large change involving resource allocation decisions. A test is added to show instantiation of two gimbals which will allow for straight-forward testing of [multi-gimbal support ](https://github.com/ArduPilot/ardupilot/pull/31983)from @hunt0r We can't test mavlink gimbal body stuff in here for mavlink gimbal (see patches on other PR which add that) These simulations based off an old branch of mine which added the Siyi simulation; it's probably closest to reflecting weird behaviour seen from the actual device.

github.com/ArduPilot/ardupilot

AP_NavEKF3: loop over states in FuseBodyVel Kalman gain computation

master ← tpwrules:pr/ek3-kalman-loop-FuseBodyVel

opened 04:49PM - 30 Mar 26 UTC

tpwrules

+46 -89

## Summary Completes https://github.com/ArduPilot/ardupilot/pull/32213 after …FuseBodyVel was made properly testable in https://github.com/ArduPilot/ardupilot/pull/32598 . ## Testing (more checks increases chance of being merged) - [X] Checked by a human programmer - [X] Tested in SITL - [ ] Tested on hardware - [ ] Logs attached - [ ] Logs available on request - [ ] Autotest included ## Description Same form and coefficients for most elements. Saves lots of flash, specifically some 4K. Tested that there is zero replay change with this as opposed to without it. Also tested that deliberately breaking it (slightly changing a coefficient in FuseBodyVel) causes a replay change, confirming the test is working.

hunt0r · March 31, 2026, 12:09am

Time-permitting, I plan to bring this “off the floor” at the end of the meeting.

Discussion-focus questions for the PR template changes:
1a. Is adding a PR classification section sufficiently valuable?
1b. Should classification instead be non-checkbox bullet points that an author is responsible to delete if they are not applicable?

Do we have consensus on the proposed checkbox changes?
Do we have consensus on adding a “Testing Details” section (distinct from the “Description” section)?

CraigElder · March 31, 2026, 9:45pm

Attendee count (max): 18

UTC2301 - CI

Colcon thing still in testing
Common origin issues
- Make the set origin call return false?
- We have a PR which doesn’t work

UTC2307 - 4.7beta

ArduPilot Dev Call 2026-03-30 - Google Docs
Randy was unable to reproduce the secure bootloader problem
- Closed the issue due to a lack of response to the user
IST8310 issue will be resolved in beta3
- Tested by the user
Board safety options and defaults acting weirdly?
- Working as Designed
Loiter performance
- Set the input to 0.1 from 0.15 and be happy, people!
- Input shaping
  - previously made the approach slow (jerk-limited approach to zero)
  - Now it does the same thing on departure (jerk-limited departure from zero)
  - Smooth transitions in both directions
  - Same time constant
  - For small adjustments you can see this
- We are planning on changing the default
- This was the biggest blocker
EKF alt offset increases when switching
- Need a fix
Questions around some parameter conversion
- Flight code / units to make GCS happy?
  - Randy and tridge think it would be worth the space
  - Plane: integer distance parameters prevent GCS unit conversion (m to ft) · Issue #32606 · ArduPilot/ardupilot · GitHub
Allow MissionPlanner the ability to force-load the same firmware
- Helps clear the crash dump off the vehicle
There’s some odd stuff with climbing 2m in land
- Ignores height in waypoint
- One ignores, one doesn’t
Beta releases every 2 weeks is the plan

UTC2329 - common.xml: update description of xtrack-loc in MAV_CMD_NAV_LOITER_* by peterbarker · Pull Request #495 · ArduPilot/mavlink · GitHub

Change description in xtrack fields in XML
Need to see if anybody is using these meanings, because they’re pretty out there

UTC2335 - common.xml: change NAV_VTOL_LAND_OPTIONS enumeration values by peterbarker · Pull Request #2450 · mavlink/mavlink · GitHub

Correct mavlink/mavlink/master version of enum only used by AP
Approved

UTC2340 - autotest: make blimp manual flight test more robust by tpwrules · Pull Request #31280 · ArduPilot/ardupilot · GitHub

Fix blimp manual test
Long-standing PR which fixes a problem
Not fair leaving a CI job failing all the time
- Need to go one way or the other
Randy will ping Michelle directly
It was always close to the edge

UTC2341 - AP_NavEKF3: add EK3_MAG_CAL=7 ground and in-flight mode by andyp1per · Pull Request #32200 · ArduPilot/ardupilot · GitHub

Add EK3_MAG_CAL=7
Want testing evidence but good to merge past that

UTC2350 - ArduPlane: Fix bound check in mode_switch_changed by sayoojkkarun · Pull Request #32222 · ArduPilot/ardupilot · GitHub

Fix bound check in mode_switch_changed
Merged
Can’t actually cross this path

UTC2352 - AP_NavEKF3: use onGroundNotMoving for zero velocity fusion by andyp1per · Pull Request #32396 · ArduPilot/ardupilot · GitHub

AP_NavEKF3: use onGroundNotMoving for zero velocity fusion#32396
Are we really not moving when that state is true?
- Boat mode?
We want to know what’s behind this PR
- What’s the real-world thing that this fixes?
- And a log?

UTC2358 - Clarify distance to destination & cross track error by hunt0r · Pull Request #32454 · ArduPilot/ardupilot · GitHub

Fix some method and attribute names
Will merge when confirmed as no compiler output change
The rover breakout Randy will look at it

UTC0001 - Tidy DPS280/DPS310 handling by peterbarker · Pull Request #32459 · ArduPilot/ardupilot · GitHub

Clean up DPS280 vs DP310 state
Merged

UTC0007 - RC_Channel: understand ignore-receiver when reading rc inputs by peterbarker · Pull Request #32488 · ArduPilot/ardupilot · GitHub

Fix RC input in face of combination of RC input and failsafe values
Merged!

UTC0011 - Tradheli: rework of Rotor Speed Control library to improve code functionality and readability by bnsgeyer · Pull Request #32516 · ArduPilot/ardupilot · GitHub

Rework rotor speed control library
Streamline it
Removes extra code which is now in the motor frames up
Decisions are being moved up into the RSC
Questions around desired vs actual spool states.
- Whether outputs should be based on one vs the other….
It’s ongoing

UTC0022 - waf: Drop support for end-of-life Python 3.8 by cclauss · Pull Request #32526 · ArduPilot/ardupilot · GitHub

Drop support for focal in Vagrantfile and environment install script
Marked for merge
Along with a companion PR

UTC0035 - AHRS: check backend return values when setting common origin by garvit000 · Pull Request #32547 · ArduPilot/ardupilot · GitHub

Check backend return values when setting common origin
Doesn’t work at the moment because of the set-once thing
Suggestion is that the backends return true if you are trying to set the same location to same spot to avoid the “already set” problem
Will come back to DevCallEU

UTC0041 - SITL: Unit tests for Battery by hunt0r · Pull Request #32561 · ArduPilot/ardupilot · GitHub

Unit tests for SITL battery
Merged!

UTC0043 - Add simulations for Topotek, Viewpro, AVT and Siyi camera/gimbals by peterbarker · Pull Request #32613 · ArduPilot/ardupilot · GitHub

Simulations for topotek, viewpro AVT and Siyi gimbals
Merged!

UTC0051 - AP_NavEKF3: loop over states in FuseBodyVel Kalman gain computation by tpwrules · Pull Request #32618 · ArduPilot/ardupilot · GitHub

Loop over states in FuseBodyVel Kalman gain
Saves ~4kB on 2MB board (boards with velocity feature)
MergeOnCIPass

UTC0054 - off the floor

PR template changes
Classes of PR?
- No compiler output change
- Random code changes
- Behaviour change
Keep the checkboxes
Peter will merge reasonable changes

UTC0009 - partner call tomorrow

UTC0009 - close

StrikeEagle · April 1, 2026, 4:47pm

Is this referring to ATC_INPUT_TC?

CraigElder · April 2, 2026, 2:59am

Yes it is referring to ATC_INPUT_TC