Manned autonomous boat rudder control

I’m in the process of building an autopilot for my old motorboat, It’s a slow/fat boat going about 5 knots with a diesel engine. I have experience building drones, electronics, and just started learing arduino. I promised my old grandpa to visit him with the boat but it’s a 40+ hour drive over a big lake and the problem right now is that as soon as you release the steering wheel the boat will deviate… And since i’m a student i don’t have the money to buy an autopilot system for the boat, and the idea of getting to build my own autopilot really sound like a fun challenge!
“if you cant afford it, build it”

The plan is to build my own and be available to control the rudder/steering wheel, the motor rpm will be set manually. This is only for open calm water and with constant supervision.

Right now i’m in the process of determining of what motor to use for the steering wheel, i’ll be driving it via belt drive, and it has to be frictionless while not engaged to not interfere which makes servos and stepper motors out of the question. So right now I’m looking at both Brushless and brushed motors which i have a lot at home. (the plan for the future is to have a hydraulic system with hydraulic pump)

I have some small questions that i hope i can get some help with here! :slight_smile:

  1. I’m guessing it would be good with some kind of rudder positioning potentiometer to not burn the motor, Is it possible to hook to up to a pixhawk 2.6.8 and have a brushless motor turning the wheel without any super expensive controller? Or is there really a need for that? Could i just program it to turn really slowly maybe?

  2. Should i go with a brushed motor instead? I’ve seen some simple/cheap boards where i can plug in an external potentiometer for rudder positioning.

  3. Is there any mode for boats which makes it hold it’s heading or is it just waypoints?

  4. How hard is the programming in mission planner? From what i can tell it looks to be the same as the old drone configurators.

I really appreciate all help i could get, been sinking my head down in to this for the past month but i cant find answers on these specific questions.

Thanks a lot in advance!

I am no expert in boats, but perhaps my thoughts on this may help.
How much force is needed to turn the steering wheel?
To get some numbers, I would meassure it while the boat is moving/different motor rpms/standstill. A simple spring based scale should be enough.
That way you can calculate what torque the drive will need.
Without more information it is hard to say, if any kind of motor without a gearbox will be able to turn the wheel. The problem with a brushless motor is, that it has problems to start from standstill against a load.
Freewheeling will be another issue. A belt drive needs tension depending on the torque it has to transmit. This means a belt drive is not as easy to turn as a chain drive for example.
Ideal would be some kind of a clutch to disengage the drive from the wheel.
Depending on the torque needed, I think of a big brushless motor with an encoder or hall sensors driven by a VESC or an Odrive. Those ESCs support position control, turning the brushless motor into a servo.
Good luck with your project!

Hey Sebastian!
Thank you for the quick reply, i had the same idea last night after i published this post. There should be a simple equation to measure how much power i need. I’ll get back to you when I’ve calculated this, and if anyone knows the equation i would love to hear it.

Do you think it’s important to have a rudder sensor?

And what do you think about the modes and is it possible to configure a motor in mission planner instead of a servo to control the rudder?

I am thinking about doing the same thing. Did you ever develope a autopilot for your boat?? BTW great stuff here for that. I plan on doing this myself. Going trolling by myself is a real pain.

Your biggest issue here will be the time constant of the yaw control. As the boat is fairly large and heavy, there will be a delay between moving the steering wheel and yaw response. So, you will have much less P and more I in your tuning.

During your long passage, you will not be making large course corrections, so you do not require large movement of your wheel. Therefore a linear servo connected to a ball joint on the wheel is the easiset option. Just pull it off if you need to steer manually. These can be picked up reasonably cheaply.