I have been playing with the AOG user interface. This could all be setup but I basically just left the tractor implement set as default while I am playing with the user interface.
Is there a way to set the control point to be the center point of the implement instead of the center of the rear axle on the tractor?
I think the tractor drives to place the implement, that is why there is so many hitch styles in implement settings.
I have had this question too… does aog try to get the implement to the center of the line, that’s where it’s all about after all. But I could never see any influence of different tool settings to the steering of the tractor.
AOG experts should comment this but I don’t think AOG aims at the implement on the wayline but the tractor instead. Implement geometry is there for coverage painting. My understanding is that only AGCO (Fendt) has the option to aim the implement to the wayline. Then Trimble and some others have active implement steering options but they need another antenna on the implement.
My thought at first glance on this is it should be fairly simple to do. Right now the steering algorithm takes the front axle to be the front axle. The wheel angle is the angle of the wheel to the front axle. Then it keeps the center of the rear axle on the guidance line.
What if instead we looked at the tractor as a whole as being the front steering tire and the towed implement axle location is the rear axle and control point. Then the wheel angle would actually be the angle between the tongue of the implement and the drawbar of the tractor. This seems like it would work. What are others thoughts on this. The location of the gps antenna relative to the hitchpin in the drawbar would be critical to correct for the location change of the antenna when the tractor is turning.
What are the thoughts of the experts out there.
Driving around in the simulator I definitely get skips and overlaps when running contours. Unfortunately 90% of our farm is contours not AB lines.
Some of the geometry on sharp curves ends up being impossible regardless like turning a 90 degree corner catch my Steiger video on youtube for an example. Even the human cannot accomplish the task.
You can only control what you can measure, so in simplest terms what you describe would need a second gps unit on the implement as well. To notify the tractor that more correction is needed.
AOG is already being tested with implement steering in addition to the tractors if your equipment is equipped to do so.
But I would just try AOG out first. Its pretty solid at covering ground with minimal overlap. All Brands of GPS have there accuracy reliability rated at 95% of the time. Nothing is perfect.
The goal is the center of the implement. The linecis drawn ceter of the tractor. It is offset to put the implement on the real line.
Make an abline using the form. Make it somewhere away from you. Make it with no offset. After it is made, get on one of the lines and start autosteer. Once steering well and following the line stop the tractor. Now go set an offset of half your tool width. When you toggle autosteer the line will jump over putting your implement in line with where it was, and your tractor far enough over to drag the implement there. Now do a uturn and you can see it.
Hope this makes sense. The tool guidance is a separate feature being tested. It guides both tractor and implement. The implement has to have some sort of steering. There are basically two lines. A fixed system can’t follow two lines. One has to win. An adjustable ( tool steering) can follow two lines, until you are out of capacity.
I got PANDA to work last night, it felt like winning. Have not had a ride on any branded system that drives like that. Its much better on the small mower, so will be overkill for grain farming.
I did my potato planting and ridging with panda this year. It drives very nice but with heavy hitch mounted implements that couse draft there is still a few cm error sometimes. (has nothing to do with panda, that eliminates the steering wheel to go crazy on bumps and that works really nice!)
How fast are you running the simulator?
Keegan,
I am running it between 5 and 10 mph. typical planting and working speeds.
I mean I have had it running at 45mph but I don’t expect it to perform well at that speed. However it seems to make little difference.
I have parts ordered to get a real life system going. I will do some testing with it. I am sure it will be more than adequate just questions as I am starting with the system.
Thanks,
Jim
That’s my biggest issue with the small tractor, AOG is trying to correct but the front end gets light. Or on side hill the implement wants to start steering.
But it is way more stable on Panda, now if it wiggles its due to low DOP only.
I am like you, spent alot of time in the sim until my system gets here.
I made the false assumption that the simulator would not take into account the speed, but it very much does.
Doesn’t sound like that is your problem though.
I’ve been looking at keeping the implement closer on a contour line. Maybe this has been covered… I would argue that AOG knows very accurately where the implement is (it’s mapping coverage with it), and therefore the error to the contour line. I dont doubt that the code and stability may be vastly different. I guess my point is to ask why we need GPS on the implement to know where it is. Im ok with the theoretical geometric calculation accuracy.
I have in my mind somewhere that Topcon has an option to put tractor or implement as guide point.
Thanks
The GPS on the tractor can only guide and correct the tractor, if the trailing implement follows the tractor as bonus it all looks very straight. This is how almost all guidance systems work. This is also how the mapping of the implement is done modelling the hitch measurements and type in relation to the tractors gps position. The guidance does not really know exactly where the implement is, its just a really good geometric model based guesser.
For implement control you need to measure the implement position by itself, which means a second receiver. If the implement is not fixed 3pt to the tractor, the implement can be influenced by side draft forces, rocks, side hills and ruts into any random position. Also you would need some means of steering on a trailed implement for it to correct itself if it was offline. Or some tricky code to let the rear gps position, influence the front gps position and over correct the tractor.
Its impossible to control or guess the exact position of a random process, that is the reason for the second implement receiver.
*** in my previous comment I mean the weight of the implement is too great with gravity, and it “steers” or pendulums the tractors rear towards the bottom of the slope… not by choice. There is no GPS fix for this condition.
Can a anglesensor on the hitch be used insted of a GPS on the implement?
I’m not aware of any Topcon screen allowing the implement as the steering control point (but I’m not familiar with many Topcon screens). Fendt however has had that feature since the times they used the Topcon AGI-4 receiver (Fendt variant). Later on they added a setting to allow either the tractor or the implement being the control point.
What you are asking appears to be possible but probably not easy to reach accurate following of the wayline and steer stability. A receiver on the implement gives an exact position also when the predicted implement path fails because of unpredictable drift from soil variations or the slope of the field. I understand a second the second receiver would let the tractor guidance work as today except with an integral type of an offset added under the implement receiver control.
What sort of real life scenario are you actually after? Often it is enough to reach 100% coverage even if the implement does not follow the wayline. An example scenario I’ve seen would be a trailed seeder using tramlines but a mounted sprayer where the true tramlines do not follow the wayline on curves. Still almost perfect sprayer coverage but the sprayer runs at times a bit off tramlines.
We have older Raven system that can handle the seeder in the middle of line. Distance from rear axle to the seeding units is about 5-6m.
Our best real life example is field that is shaped like boomerang. Only sensible way to seed is running the long edge and make curved AB-line.
Raven drives over at the outside corner to keep the implement in line. Even that system without RTK goes quite nicely and makes really good results even the radius keeps changing pass by pass.
An angle sensor would give you an idea of how its tracking for sure, very hard to implement on the articulating hitch. But still would not be as accurate and easy as just knowing the position with gps.
Having a fixed 3pt implement is very different from trailed, or trained implements like an air seeder cart ahead of its drill.
Like everything you need to ask how much accuracy do you really need. Seeding on a 40’ seeder with 4” overlap on flat ground does not require implement guidance. Planting onions on a side hill at slow speed probably would benefit greatly from it.
Not really. AOG is only roughly estimating where the implement is. I’m not sure anyone has actually measured how it compares to real life. Under idea conditions it might be close.