I just had some electrical issues with a 4440 sprayer. Couldn’t get a case tech to come out and fix it and a local ag mechanic couldn’t find the issue.

I tried for days cleaning connections and probing pins with a multimeter until I finally had the brainwave to ask Chat GPT and I was amazed.

It walked me through where to probe, what the readings meant and eventually found a broken ground pin in a harmess on the boom. I fixed the broken pin and said a prayer and the sprayer has been running smoothly ever since.

So I definitely think there is potential in AI diagnostics for ag equipment.

 

This is good to hear, troubleshooting is exactly where I see the biggest potential.

Would it have been useful if you could save that ground pin fix and have it tied to your sprayer for future reference? That’s kind of where I see OTIS standing apart from ChatGPT, it uses the same reasoning power, but it also remembers and organizes fixes for each of your machines. Curious what your thoughts are?

 

Right now I just feed the reasoning model with the pdf service manual for that piece of equipment, describe the problem with as much context as possible and work through it as basically a sounding board/second opinion.

So yeah it's extremely useful, and the more context you give it the better it will be. I would start by collecting every single service/op manual you can get your hands on.

 

I agree, the more context the better. One way I’m thinking about solving that is by saving each conversation to the specific machine you’re working on, so over time it builds up a history of everything you’ve done with that piece of equipment. Of course, you’d still need to describe the issue in detail, but at least the system would already know the background.


Curious, do you upload the pdf manual every time you start a new convo? And do you save those convos somewhere to refer back to later?"