I want to use pixhawk with the apm rover firmware to operate an agricultural tractor. I need to set up my tractor as an rc vehicle. My plan is as follows: any feedback I can get is appreciated. I plan to order the parts in the next couple days but if there is something I've overlooked or my plan has no chance of success let me know.
Throttle control: Linear servo (https://www.servocity.com/html/25_lbs__thrust_linear_actuator.html#.VVOUCJNyP5U) I don't see too many problems here.
GPS: I plan to use the gps already installed in the tractor. I am planning on buying a pololu 23201a to convert the serial nmea to something I can input to the pixhawk.
Steering: This is what I'm most worried about. The tractor has an electric over hydraulic steering valve, so basically provide 12v to right side tractor turns right, provide 12v to left side tractor turns left. I am planning to put in a dual 12v relay controlled by pwm signals for the steering https://www.servocity.com/html/electronic_pwm_controlled_dual.html#.VVTFmZNyP5U Anybody have a gut feel on whether this will work or not? The problem is the steering output from pixhawk needs to reach a certain threshold before steering valve will operate this is quite a bit differnt than how it is set up on my rc car where each pwm value corresponds to a specific steering angle.
Clutch: Much testing has to be done with a person in the cab to operate the clutch and some sort of ignition failsafe installed before I get the tractor to be totally autonomous but I would like to start thinking about the clutch, and start testing some setups with a person in the cab. I could get a linear actuator that releases the clutch when throttle pwm reaches some predefined threshold? But I would really like a system that if the power is cut or signal is lost clutch is automatically depressed which will stop the vehicle. I could set the clutch up to be always depressed and use a linear actuator to engage it. Connect the actuator to the clutch with an electric solenoid so if power is lost actuator releases and clutch disengages. Anyone else have some good ideas?
Replies
So harvest is going really well. If the weather is as nice as the forecast says it is going to be next week I should get close to half done. The biggest problem with the whole automated tractor thing is keeping the tractor running. It sprung a fairly massive engine oil leak which I thankfully caught in time but was a pretty messy job to fix and it ran low on coolant and overheated on me. I don't think a driver would have caught the engine oil leak any sooner but the engine getting warm would've been noticed right away. So I created a very extensive checklist that now has to be filled out every morning before we start hopefully we can avoid blowing up the engine. I just need it to work this year, next year I will make it so I can monitor temperature and oil pressure, ect from the combine. Last night I finished in the dark on the far corner of a field and then I asked the tractor to go back to my yard (750m away) and I was sure it went the wrong direction but I followed it back and sure enough ended up right where I told it to go. I guess I would have been lost out there for a while without it.
Hello Matthew!
It is amazing work that you have made! Wonderful!
I understood that you would like to development it even further. I would like to share information and experience with you (and others of course). Would you like to make work together in this project and share information? Is it okay for you?
If yes, please continue reading. Otherwise this is too long message and lot of detail information needed from you.
I would like to know more details about devices and electrical connections that you have used. Like
Power supply
Steering Feedback sensor. What sensor you used? Type, model?
Steering, You said that you use tractor proportional valve for this and you shared link to the board where was two relays with PWM output.
Throttle control, what you use here? (in the video I saw it is actuator, could you please, share connections?)
Brake actuator
Gears, Do you have automatic transmission in you tractor?
I read that you use buttons or similar for avoiding collision with objects. Do you still use those or how you protect from collision example with trees?
GPS, What module you use today?
Could you please share screenshots from settings (if they are different than previous pictures that you shared)?
Could you please share your code, so maybe I can use it and modify it for my purpose?
What softwares you use in ground station? I mean example Mavlink, Droidplanner etc? Architecture?
I have planned to make similar project but to the peat bog environment (Please, see thelink for video from that environment ). That environment is much challenging than what is open field without ditches and canals. Peat bog is divided for tens or even hundreds different strips (? is it right word?) that goes to different directions depends of area. Both side of strip has ditch and very often also both ends has canals.
Strip width is about 12 meter (40ft) and width of the strip in the end of strip to the next strip is about 6m (20ft). That mean challenging turns. On the strip need to drive middle because of peat pile or/and soft ground. You need to also know that there are several ways to harvest the peat and now I’m talking about milled peat.
Also ground give some challenges. Like you know dry peat is easily flammable (need only spark) and hard to get flame out. Here and there are also ground where you can see stones and dust and dirt (mineral ground). Those are places where you should slow down and even lift up example turner (turn peat to dry, pull behind of the tractor) to avoid sparks that comes when metal hit to stones.
Also wetting soil and soft ground give some challenges because tractor can get stuck. Dust give challenging when you load out, for devices and visibility (can’t use reliably example lidar sensor for safety ). Please the link for video so you will get idea.
Here are two main tasks that I have thinking to use in first and second states. Turning (please, see the link for video), it means turn the peat that lying on the strip after milled (is it right word?). Milled peat need to turn about 3 times.
When peat is dry enough, the lying peat will move to middle of strip into pile that goes along strip. Need to do this about 5 times so the pile is big enough for loading. That is similar work like you make. Tractor where you load peat just drive in the next strip (very often this way, but not always (please, see link for video from youtube).
I have planned that need to have two ground stations where you can control. In the loading the Loader will control tractor. When you drive tractor to the stack (It called stack where you unload all dry peat from strips. There can me more than one stack on the peat bog area) to unload tractor and back to the loading driver from another tractor will control it. This driver can take manual or remote control (RC) to unload it to the pile. Maybe this way it is possible control more than one tractor simultaneously (in future).
P.S Step motor for steering is not so good as gear motor because step motor can lose its home position example after power off. With gear motor you don’t need to take care of it. Feedback is enough. (For vehicles where is no proportional valve available for steering)
Please, see some drawing and links to the devices that I planned to use attached.
Pixhawk and tractor.pdf
Hey Jaakko I looked through your list. The connections on every tractor are going to be unique to the model. Essentially you have to hook up actuators to all the relavant controls in your tractor. How this happens exactly will depend a lot on the type of transmission your tractor has. Even the individual actuators have to be specd for speed, stroke length, and strength depending on how hard it is to move the controls. This really needs to be done on a case by case basis. I think I documented fairly well in these posts how to make the connections to different gps modules that were already on my tractor. The best way to begin on this type of project is just to get started. Buy a pixhawk and a cheap rc car and build up from there. Once you get the pixhawk to work in an rc car you will have a much better understanding of how to proceed to get it in your tractor. If you have more specific questions as you get going I will be happy to help.
I'm happy to help. I am right in the middle of harvest so the bulk of my answers will have to wait till I finish or have some rain days. One suggestion I have is to work on one tiny area of the project at at time and try and learn and figure out everything you can about that one aspect and then move on to the next part. No doubt you will change your plans as you go so don't buy everything on your list all at once.
I learned a lot about pixhawk by working with linear actuators but if you are putting this is semi modern equipment finding a way to hook it all into the electronics of the machine is the way to go. I plan to work on this area over my winter season so I don't have a lot of tips for this yet.
And always keep safety in mind. Big machines driving themselves around is no joke!
Ps. loved the videos you posted I've never seen that before looks really cool. What part of the world are you in I'd love to see that in person one day.
Thank you for your quick answer. That was the plan. Move step-by-step. Those lists just give me some kind of guideline.
I'm from Finland, Northern Europe. There should be also peat production somewhere in Northern America maybe also in Canada.
I will wait when you have "free" time to collect information, make drawings and files to share.
Please, check also this from John Deere project that has made with big money. Please, look at the link.
For sure! Got a full 14hr day in yesterday. If one of the components quits on me I will be really disappointed b/c it will take a while to get parts and its proving to be incredibly useful. I really should have ordered a second set of everything.
So here is a video of the tractor actually being useful. Up to this point it has been a giant time suck, but it just put in it's first full day of work. I see no reason why I wont use it all harvest without a driver. Testing went so well that I didn't even bother to have someone in the tractor at all yesterday, not even for one run. I was pretty sure I was wasting my time on this project so I am over the moon excited that it works as well as it does. I saved around $300 dollars in wages yesterday (more importantly I was able to redeploy the labor to other tasks), of the 80acres I harvested yesterday 20 of those was while I was actively unloading into the cart.
Thanks to everyone who has put time and energy into Pixhawk, Mavproxy, Mavlink, DroneAPI, MissionPlanner, APM Rover, and all the great documentation that goes along with those. I couldn't have done it without you! If you are ever in the area stop by for a drink! Sorry if I missed anything, or anyone.
In the video I mention that the tractor comes a little close to the combine and I attribute this to it being dusk (time of day when gps often does funny things) I am totally wrong about this being the cause of the problem, it is a thing though. I swath my fields with autosteer and I always set the direction either EW or NS, East West in this case. Then when I get to the field with the combine I can use the autosteer on the combine to follow the swath. Because gps can vary a bit the combine has a feature called "nudge" basically if you see that the combine is drifting a bit you can nudge the autosteer right or left. I didn't have my combine set to exactly the same width as my swather so over the course of the day I kept nudging it in one direction as I moved across the field. I clued in when I realized that I was having to move the grain cart the same amount away from the combine that the value of "nudge" had crept up too. I can't believe it but I checked and rechecked this behavior and it seems that when you use the nudge feature on the combine autosteer it adds some offset to the gps coordinates and then outputs the modified coordinates through the nmea serial connection which I feed into my laptop that ultimately determines where the cart will be relative to the combine. Now that I know this is how it works I can work around it but I think this has got to be a bug in the software in my combine, the nmea serial output should be the unadulterated gps position, at least in my opinion.
Congratulations Matthew!!!!! How envyyyy!!!!
How do you empty the cart to the truck? someone is waiting for it?
Yes the truck driver has to empty the cart manually. Id love to be able to automate that in the future as well, that would be seriously cool!
-
7
-
8
-
9
-
10
-
11
of 24 Next