Copter
- Intel Aero RTF update
Plane
- Tailsitter work
- Prop hanging
- Soaring patches
Sub
- progress on sub merge
Rover
Antenna Tracker
Hardware
- CAN update
Copter
Plane
Sub
Rover
Antenna Tracker
Hardware
10:04am - AeroFC and LdM (in lieu of TP)
Support is now actively being PR’d for the Intel Aero Flight Controller!
Need to get DF logging working over mavlink
Contains same stm32, slightly different sensors vs PixHawk
Flies on Aero for up to 8 months ago
Pins not exposed to fly with it
Redly drone has all pins to the flight controller
Can fly with aero itself alone
FPGA firmware is not public
Could possibly run ArduPilot on Linux side
Could pull data out over UART
“Sensor head”
Heaps of bandwidth
Latency could be an issue which is why you would run the UART at a high baud
Custom protocol wanted ,really
Header
Sequence number
CRC
May get a better flying behaviour
Lots more resources available on the Linux
F7 is coming
Double-precision!
Running a failsafe is harder than it sounds
We don’t poll sensors as fast as we could because we don’t want to annoy the flight code
Also UART
Could probably go IMU_MASK=1 and run a single EKF
10:23am - tridge and tailsitters
Tailsitter VTOL
Needed example for VTOL, chose Pogo
https://www.youtube.com/watch?v=Nh9dhBJY010
Tried a tailsitter on the same code
It just worked
Fixed wing that happens to be pointed up
AP_Motors_TailSitter backend
When you want to roll, put in aileron etc
Differential thrust tail-sitters also supported
Someone is sending tridge an example to test on
XCraft guys?
Would now work much, much better
Would need a small patch for differential pitch
No 90 degree problems
AP_AHRS_View class allows different parts of the code to view the attitude from a different angle
So plane sees itself as level when in level flight
So does Copter
AddictionX will be flown on Friday
Same as model used in SITL, so will be interesting comparison
Soaring support in Fixed Wing
Automatic thermalling
There’s a PR now, but it needs work
People have been running patches for a long time
Much cleaner split
10:33am - merging the sub stuff
Jacob thinks he’s addressed everything that was in the pull requests
Much squashing and generally making things compliant
Moving changes into Sub code rather than generally
11:12am - Rusty team membership
100% positive vote
Tridge will send email inviting him to the group
11:13am - Rover and Tracker
How many devs have a tracker?
CE has built 6 or 7
PR has access to one
And another half-built
How many have Rovers?
About 3 devs
Rover has been driving a slow-crawling skid-steering vehicle around his house
This is apparently to help getting a SabreTooth ESC working
Sonar obstacle avoidance
Update rate not-so-good
Grant will be looking at
Physics limitations
Vectored configuration of thrusters?
4 thrusters in a square at 45 degrees
Output mixer in rover needed to support that
Normal steering
Skid-steering
+Thruster
Add to SITL
Steerable thrusters?!
Going to have to add support for rover/sub/quadplane takeoff
Fixed-mounted antenna trackers discussion
11:28am - CAN update from PR and JC
Harrier is being sanded
BD has gotten Philip a vehicle with spinning motors on it
And JC has a vehicle running with his code on it
Hopefully something in the air by the end of this week
12 working CAN ESCs ATM
Hex will be building some CAN ESCs for Solo
Still using the PX4Firmware/NuttX CAN layer
Everybody really wants this in the core ArduPilot build
PR is making a real effort to get these things out there
Other efforts have been too one-off
11:33am - 1000 plastic insulator bits going out for PixHawk2
Short-term software solution will work nicely in the meantime
Older software may have issues with certain configurations (flaperons)
Kapton tape will work in the short term if you can’t wait for the insulator
Ballistic parachutes for the win
The screws were probably over-torqued
Philip is fully planning on being forthright and issuing Service Bulletins as required for PH2
In contrast to HK clones who will never tell you there is a problem with the HW
11:40am - CAN middleware
Tridge needs to review JC’s PR
PR will be funding work on CAN
11:41 - GSOC with OB
Google to decide whether we get accepted or not within 2 weeks
students will then be eligible for applying for it
http://ardupilot.org/dev/docs/gsoc-ideas-list.html
Beagleboard has been accepted and has had several successful applications
They send a board to successful applicants
Should we?
Dev-team funding could cover this
Buy-and-ship or give them money
Would need to tailor to which project gets accepted
GSOC student does get paid by Google
At half-way mark
Why does Google do this?
Recruitment exercise
Vet thousands of potential employees
Mentor reports
You can and should vet students
Helicopter autorotation added to list
RlF has a whitepaper which will help students to do this
11:53am - hardware announcement (Jeff Wurzbach, Wurzbach Electronics)
OLED thing
Gh connectors
Nice back and the like (plastics)
No price available as yet
Should be available in a couple of weeks
11:56am - Partners list issue
No Randy, so deferring to next week