Large rover Not Following Track in Auto and Guided Mode

I’m facing a challenge with my large rover’s navigation. Even after putting the rover in auto mode and arming the motors, it consistently deviates from the intended path. Despite aligning the waypoints with both the “current heading” and “direct to current WP” lines, the rover fails to follow the designated route. Instead, it continually makes unnecessary turns and struggles to track the provided path.

I’ve made sure to remove failsafe and preflight checks for compass, GPS, and all other sensors from ARMING_CHECK. Despite these adjustments, the issue persists.

To clarify, this large rover weighs around 330lbs and is equipped with a CubeOrange+ and a NEO-7M GPS module for my Pixhawk rover setup. I’ll be sharing a video demonstration via the following link.ardupilot forum - Google Drive

I’m reaching out for assistance and guidance on how to address this navigation issue effectively. My goal is for the large rover to reliably follow its designated path in auto mode without manual intervention. Any insights or assistance would be immensely appreciated. Thank you for your support!

This is a bad idea. Re-enable all of them, forget the video and post a link to a .bin log file where this event happened.

Did you proceed thru all of the tuning steps?

2 Likes

Until you systematically and carefully go through the COMPLETE rover setup, you’re wasting your time. It just won’t work. Start from scratch on the tuning section , take your time and you’ll get it working. Good luck!

2 Likes

Fixed that for you…

Follow the above advice from Steve and Dave. Both have been through the wringer building and helping others build Rovers.

Here’s the section of documentation you should review closely:
https://ardupilot.org/rover/docs/rover-first-drive.html

1 Like

Thanks for the link. I’ll flash the rover firmware into the cube again and follow each step given in the link. After that, I’ll run a trial and upload a .bin log file link.

2024-04-29 14-34-57.bin (972.3 KB)

Here is the .bin file created after flashing the board and following the steps of setup for rover.
We have encountered another issue while setting up. We use to first disable BRD_SAFETY_DEFLT for the rover to run. After flashing the cube and setting up again, when we disable BRD_SAFETY_DEFLT parameter, motors start spinning without arming the cube.

Have you performed a motor test via Mission Planner?

Your log shows no intended throttle output above zero.

1 Like

I didn’t run the motor test via Mission Planner, but I uploaded a mission with one waypoint. After arming the motors in auto mode, there were no error messages or failsafe alerts in the message box. Additionally, the motors didn’t run.

If you are uploading missions prior to motor test and then driving in manual AND tuning in acro mode, you are skipping significant steps.

2 Likes

I will look over the tuning steps and perform them again. Thanks for the much needed help. I will upload a new bin file after I follow the steps, run motor test via mission planner.

1 Like

I can’t do motor test with BRD_SAFETY_DEFLT enabled and when I disable it, motor starts spinning.
I disabled it and tried motor test via mission planner and in the tuning graph I can see the pwm graph change and I can also see motor test armed the motor and then disarmed it but motors speed doesn’t change at all. Main issue right now is when i disabled the BRD_SAFETY_DEFLT parameter motor starts spinning when disarmed and the speed of the motor doesn’t change when armed and when pwm increases.

Here is the link of bin file:

An update. After selecting MOT_PWM_TYPE as ‘3’ which is brushed with relay motors we are able to stop the spinning motor while disarmed. Yet we have to force arm because of a pre-arm check which is “PreArm: relay function 5 unassigned”. Due to this pre-arm check we are unable to run motor test via mission planner. I am uploading the latest bin file here for our reference and we ran motor test once after force arming. I want to solve this prearm check and assign my relay function 5 and 6 as direction pin in the motor driver. I search different discuss pages but no one seemed to have such issue.

https://drive.google.com/drive/folders/1BzfzsniVqR6aI85cuVyg6-vOORK2eeu1?usp=sharing

Thanks for your support in advance !

Hi @Dhruv_thakkar,

I think it’s not clear what input your motors are expecting. Maybe you can provide a link to the motors you’re using and/or check what input they require? Up until this latest post the motors would have been receiving PWM from the autopilot. With the latest change of MOT_PWM_TYPE = 3 the autopilot is sending a “duty cycle” but in order to control the direction two pins on the autopilot should be setup as “relays”.

We have some instructions here on how to setup brushed motors.

It’s best to know what kinds of motors your using. AP has a lot of parameters to support a lot of possible setups so just changing them to see what works will take a lot of time.

1 Like

Thank you for your reply ! We were able to assign relays as direction pin in RELAY1_PIN set to “AUXOUT6” and RELAY2_PIN to “AUXOUT5” and pre-arm checks for relays are solved.
Now we are fine tuning the PID for speed and throttle. We are not facing any problem except every time we disarm, their is pre-arm check for “PreArm: Compasses inconsistent” even after following compass calibration multiple times and I solved this issue by calibrating it from “large vehicle MagCal” adding the direction degrees manually and then it use to run without any pre-arm error. But I still need to do that every time I arm/disarm motor.

main issue:- " prearm:- Compasses inconsistent " I am unaware on how to solve this issue after calibrating it multiple times.

Thanks for your support in advance !

1 Like

Hi @Dhruv_thakkar,

Just for reference we have a pre-arm check wiki page here which should list every pre-arm warning (please tell us if it doesn’t) including the compasses inconsistent one.

I think the advice on this page is accurate.

  1. Check compass orientations (e.g. COMPASS_ORIENT)
  2. Move compass away from metal in the frame
  3. Repeat compass calibration
  4. Disable internal compass.

The issue is probably not (1) and it sounds like you’ve already done (3) so I think just (2) and (4) are the next steps. Realistically many vehicles have the autopilot buried within the frame close to power wires and thus require the internal compass to be disabled.

2 Likes

I totally agree with you. Now that you have told me the possible reasons, I realise that I have cube pixhawk and gps both mounted on metal because my structure is metal. Thank you so much for your help. I will first mount it away from metal and see if it improves and if not then i will disable the internal compass.

Thank you for your support !

1 Like