object tracking for farm equipment

The above video shows a similar set up to my own.

What I would like to do with to have the spout track the opening in the wagon and adjust automatically to blow the chaff into the wagon behind.

The Neck of the spout on the chopper can swivel left and right and the flap at the top of the spout can move up and down already powered by 12v electric motors

Is it possible to connect some sort of tracking device, I do not really care what so that the spout will follow the opening in the wagon. whether 1 sensor is good or 4 sensors one in each corner of the opening in the wagon.

It would be nice for this to be arduino powered only, Possibly with a joystick over ride, on off tracking mode (as the wagon actually tips up and away to unload) and for it to only track with in a certain range both for the left/right and up/down maybe limit switches if they can handle this environment.

The motors are not servos and probably could not handle dozens of micro adjustments so if it could be set to be within such a parameter before adjustment.

I thought we could do something similar to a camera tracking a red ball that way the operator could also watch on a monitor (instead of constantly looking backwards while driving forwards and looking to the side to make sure your following the cut grass and not clogged) how ever that looks like you need a laptop as well. We would like to keep this as simple as possible.

I am new to Arduino, I have ordered an Uno R3, 2x12v motor driver boards for 5v logical input, two back up cameras and screens 150 degree visual range, breadboards, jumper wires, and a Arduino uno and kit (ir sender reciever and a ton of other bits and peices),2x Uno r3 Mega, 4 joystick break out modules, dc-dc converter to power the 5v arduino, 2x bluetooth Master and Slave Module for Compatible Xbee Arduino, and more

Some of this was to convert a diecast tractor into a RC for learning how to use arduino. but parts are still showing up and harvest season is getting close so to have everything needed would be nice and not wait till next year.

The_Woodsman:
What I would like to do with to have the spout track the opening in the wagon and adjust automatically to blow the chaff into the wagon behind.

IMHO it will be difficult to make an effective "computerized" solution to this problem.

One significant issue is that the grass will interfere with the sensors - whatever type you use.

Another issue is that you will actually need very accurate tracking. In the video when the grass misses the trailer the positioning error seems to be very small.

If it were not for the grass obscuring visibility I would suggest some sort of vision system - but that would need a PC, not an Arduino.

Have you considered a simple mechanical system - for example a "string" connecting the flap to the centre of the trailer.

Or, maybe, a bar (perhaps extensible) connecting the nozzle to the trailer with something (potentiometers ?) to detect changes of angle. That could be managed by an Arduino.

...R

See if a Pixy will do what you want, but be aware that many seemingly simple vision tasks fail outside of tightly-controlled lab lighting conditions.

The Pixy will do vision-tracking on an Arduino. Look for Ben Heck's video explaining it on YouTube.

But one fragment of grass over the lens and the tracker stops working.

I would try to put an angle encoder on the towbar linking the two machines. Then a little bit of maths can work out where to point the spout during turns. The output linkages will also require encoders. US Digital makes some great encoders which will survive the corrosion and vibration of the agricultural environment.

much simpler.

one transmitter on the wagon
one receiver on the chute.

almost like the sun follower, but with radio waves.

mechanically, you could have something on the hitch.
two end switches, heavy duty, dust proof obviously.
a power window motor from a car.

not sure how you can swivel, if there is a motor, or a gear, a belt or chaing.

a rod you place in blower that slips into the slot on the trailer, I am thinking a fiberglass one they use for marking parking lots for snow plowing. 3ft long, plenty of flex.

when the blower is aligned, the rod is inbetween the switches. wehn a switch is made, the motor spins the blower until the rod is centered and then just stops moving.

if you can live with a wide span, a couple degrees off on either side, it would run less often.

with this, you would not need an arduino, but with an Ardino, you could make it better.

other ideas is reflective stickers.
a light on the blower.
shines on one sticker, it signals to move the other way until it 'tracks' off.
don't like this as much as the area is so dusty.
but it could work well,

all of these require that the discharge is powered by a motor that you can control with relays or some other means.

couple if booms on the side of the chute, chains and bungees on the trailer.
it will keep it square to the trailer. no motors needed.

Thank you for all the replies.

This is the problem with setting anything up between the two of them like rods and so forth.

actually the trailer in this video is very close to ours and shows how large the opening is. it just needs to stay with in that. Most of the fields the grass it cut into straight lines so the majority of the problems come from hills bumps and just general uneven terrain that is our farm land. upwards of 7-9 lbs of feed a second every miss can add up.

Constantly looking back is very hard on the operators neck, it looks like currently the only feasible option is to just hook up a back up camera to the top of the spout and have a monitor in the cab with a joy stick or switches to control things which is what we already do minus the back up camera.

Pixy is what I saw last year and that is where the idea came from. just the dust could be a problem. Silage is usually cut wetter then normal hay. This year we had a very bad drought that followed over from the previous year so trying to get as much feed as possible could make or break the farm.

I would like to have started with a simpler task and learn up to this point but with the severe arthritis I have how ever I can help my family is very important to me.... maybe the arduino could help me drive the truck. lol. Thank you all. for your input and shared knowledge.

The_Woodsman:
This is the problem with setting anything up between the two of them like rods and so forth.

It would be nice to be able to solve ALL problems. But the search for perfection often means that reasonable options are ignored.

It seems to me it would be easy enough to design a linkage that could be disconnected before the tipping process starts and reconnected when it is finished. It would also probably be easy to use a cheap form of linkage that could be replaced easily and quickly if it got broken because someone forgot to disconnect it.

...R

does the chute have a motor now ?

since the wheels, the axle, the frame of the trailer remain firmly on the ground....
and we can expect that the dump section is always in a position that is fixed in relation to it, at least while harvesting,
then
any of the above might work.

the new information is that the target can move from side to side due to uneven terrain.
that means that the target is not always fixed in relation to the axial line of the tractor and trailer.

for the simple turns at the end of the row,
and with level ground
any of the above might work.

rods are not effected by the dump, but are a simple window type control. outside of the window, do something.
vision tracking is also a window type control, just a different type of sensor.

if you put on a reflective road sticker on the edges of the trailer, color should not matter.
then shine a light closer to the center, the sticker will not see the light and there will be no reflection.
when the trailer tilts from uneven terrain, the sticker will come into the light and reflect.
a light sensor located at the light beam source, mounted on the rotating chute,
would then see the light, a fault condition, and would want to signal the chute motor to spin to move the chute so the beam is no longer on reflective material.

you can paint the wagon white, put a black stripe on it,
put your vision camera on
watch for the black stripe
move the chute

same thing, different sensing, different costs of materials.

The rod idea was something you could install this weekend, but fails if the alignment of the trailer does not change on the axial, but only on the rotation around the axial.

The_Woodsman:
This year we had a very bad drought that followed over from the previous year so trying to get as much feed as possible could make or break the farm.

I would like to have started with a simpler task and learn up to this point but with the severe arthritis I have how ever I can help my family is very important to me.... maybe the arduino could help me drive the truck. lol. Thank you all. for your input and shared knowledge.

sorry to hear that such a small percentage of loss could mean so much.
as for a simpler task, do you have any experience with the Arduino ?
this project seems that it could be done with a toy car, and a trailer to test out.
also, the 'window of opportunity' seems to be directly related to the size of the opening on the trailer.
a larger funnel could direct more silage into the trailer by catching the overspray.

dave-in-nj:
a larger funnel could direct more silage into the trailer by catching the overspray.

I wish I had thought of that !
KISS

...R