I’ve been thinking that it would be neat to have a yield monitor integrated into AOG. One for a combine would be great but I’ve been thinking of other types of harvesting like potatoes or sugarbeets. There are some commercial systems on the market but as always a little overpriced in my mind 
I think it would be fairly simple to put the hardware on the machine just load cells and shaft speed but the software would be tougher (for me at least). Then to have it map the yield by color on the display would be neat. Anyone done this already?
Instant yield is probably pretty simple, yield mapping not so much?
Yeah instant yield could be done with an lcd connected to a yield sensor, no code changes to AOG. Mapping, I don’t know. But doesn’t sound just easy
That makes sense. Other things that I’d like to do is be able to have counters able to add per load or field. That would be a great improvement over nothing!
I don’t know what it is, but I’d like to hear it when it’s done.
What sort of “yield sensors” are we talking about? I though the most common sensor is just an optical sensor that detects if there is any grain in front of each grain elevator flap. Quite a lot of intelligence and calibration needed to obtain yield data from that point onwards.
Any yield estimate would need to be combined with a past GPS position measurement, not the current one. Probably not too complicated, just a position memory to cover the combine harvester processing delay. Yield map would not be very different from the coverage maps variants we’ve already seen for other purposes?
This is something we use for weighing vegetable crops. It just uses 2 load cells mounted beneath belt carrying vegetables (potatoes in our case) and a speed sensor. We currently do not have it on our harvester but just where we unload trucks from the field. I’m not a programmer at this point but I think like has been said that using GPS position and coverage maps like other projects have used would work.
1 Like
New holland also uses load cells but they use a deflector plate. The grain comes off the clean grain chain and is deflected by a plate. And if you assume the grain chain runs at a constant velocity then the relationship between force on the sensor and the rate of mass harvested are proportional.
We would also have to account for how full the header is if you want to get useful data. AOG should be able to do that.
I always figured if you could just log position, time, velocity, how full the header is, and yield data, you should be able to go through with a post processing script to build a map.
If I remember right, Ceres 8000 does use header up/working switch but no means to detect how full the header is. FarmTRX on the other hand does not care about the header position.
I understand their approach is to stamp yield measurement samples with GPS position data and it does not matter if the combine was all the time working half width or full width. In the first case one would get yield samples at half the magnitude but twice for a given area, on the second case full samples but only one for the given area. Both reaching the same yield map (at the resolution that makes sense).