I´m really impressed for what you´re doing here, either with arduplane, ardurover..
You seem to have solved really difficult challenges..
I´m a farmer with some high knowledge about hydraulic, electrical systems, c programming..
I´m looking for a gps guidance over some known patterns for tractors and harvester.
I know one of the main assumptions for this group, or biggest one, is to keep system autonomous but you seem to have reached further than my hopes.
So, as 1 first approach, do you think ardurover would fit my needs? do you know similar proyects?
Thanks in advance , and keept it up!
Hi Carlos, I'm new and was about to ask similar questions. I'm interested in amateur geophysical survey which needs GPS guidance to have lane discipline as required in harvesting together with constant speed over ground.
This needs a 'lane width' variable to reset next waypoint alongside Lat and Long adjustments for calculation so you achieve 100% field coverage.
I'm not a programmer but is this a difficult sub-routine to include at some point in the mission planner to make this calculation automatic? or has someone already done it?
If anyone can help or offer advice it's gratefully received..
Depending on row width and row to row spacing, the present GPS (uBlox LEA-6) is probably not accurate enough to keep an Agrarover in the correct row and for accurately moving to the adjacent rows. There are more expensive GPSs that are accurate down to a foot, but it depends on how much money is in your budget for the GPS.
Just a thought.
Hi Derek/Thomas. Good to hear from you.
I think, correct me if i´m wrong, the key here is if Arduroves´s capable of heading straightf within a margin of 4-5 inc .
AS Thomas says, GPS precision is really important.I Have a Raven Cruizer with DGPS and 10 Hz , so there should´nt be a problem to get an accuracy of +-10 cm.
I think the hardest part once you have a a trusted GPS receiver is to calculate based on accelerometers, Gyros(..) where is the rover really heading to, so that we may get the error from the ideal path calculated by the GPS.
Will Ardurover do this quite good? Actually it must be doing so now, to reach propodes point.
If true, Derek, your problem and mine would be the same. We could replace 3 proposed points for UAV, for a thousand of points in parallel lines in your case, and in a known patter in mine.
I´m really interested in this solution.
We need as much accuracy as we can squeeze out from a system. preferably less than 12 inches. If the need is for a more accurate GPS then it will have to be paid for. The LEA-6 spec is quite impressive. GPS accuracy may not be the only way of looking at the problem, consistency may be a key.
I'm also a farmer working on something similar to yourself. I think the ArduRover is a great starting point for this kind of project. The main problem that I'm having is understanding the ArduRover code. Currently I'm using a second Arduino to adjust the output from an APM 2.5
I would be interested hear how your project progresses.
I suspect that we are all going to have the same problem, so please let's keep in touch about this. My initial thoughts are that the Mission Planning software will require code that automatically pre-loads the waypoints dependent on parameters we give the macro/sub-routine that has to be written.
One neat solution I am exploring at the moment in use on a marine survey package is giving the software the number of lanes required between two top corner GPS positions of the survey grid. This could be easily worked out on a calculator using your lane width, he said before he went to lie down in a darkened room.
I agree, if the Mission Planner can be automatically pre-loaded this should give us a solution.
There were some comments earlier regarding the accuracy of the GPS. I'm using the uBlox and I'm not overly worried about its precision at this stage, I'd rather sort out the code first and come back to that problem at a later stage if necessary.
What would be really cool would be if we could link back into some simulation software like Farming Simulator. I've already figured out how to get my own farm map into Farming Simulator with accurate boundaries and contours but haven't looked at how the GPS coordinates could be feed in. I do believe that Unity 3D is able to use GPS data but again i haven't checked it out.
The software on this site has taken thousands of hours to write. The investment in hardware and firmware has also been colossal and I am truly amazed at the effort and dedication that has, and is going into these projects.
Software mods have to be thought through properly. Hardware accuracy must be suitable and available and the project considered as whole entity before committing resources. Even then we will probably get it wrong.
Winston Churchill quotation :- 'Success is the ability to go from failure to failure without losing enthusiasm.'
What is the most efficient way of crop spraying or harvesting a field. Is it in lanes or starting with an outside perimeter and reducing the perimeter lanes towards the center in a maze like effect?
If you have a 50 foot harvesting array or a lawnmower the answers may be different. You can't write a program until you have the method and know the hardware.
Anyone with any thoughts? All contributions gratefully accepted with a smile and a sense of humour.
I´ve come across another guy trying to do the same. I´ve told him to join us
I never heard about farming simulator, but i´m downloading right now.. Thanks Niall. i´ll try to chech how to feed it with gps data.
I feel the same about the code. Before i spend hours reading thousands of c lines, is there a ny manual, high level arq desing i may satrt with?
I mean, is Ardurover based on a Pid controller to minimize errors ( for example)
I definitely will try with arduino..
I am keeping an eye on this one. Keep us up to date with your progress.
I think If you guys just want to do straight swaths It doesn't matter how long the swath is we don't need to deal with an endless of points, we only 2 points, our work width and the course in degree where our work is pointing, a little bit of Pythagoras, juts a few lines of code and then after some milliseconds we'll always have the closest and easiest waypoint to reach by our vehicle.
One x y coordinates in any swath, doesn't matter wich one, and the point where the rover is right now, from this info we can know the closest swath we should get, from this point and the orientation of our swaths we'll obtain our new waypoint, It is not important whether we will able to reach the point or nor, because after 200 milliseconds we'll have another one better and easier to reach, besides this information if we have each point forming the perimeter, ( A square has just 4 ) we can configure our vehicle to do the whole work by itself.