Well, by the white-hat route, you would talk to a farmer for a long time, figure out the exact dimensions of the black box, and build a white box to replace it.
By the grey-hat route, you analyze the black box and patch out or bypass the security features in it. You allow the owner to jailbreak the hardware, while still largely using the original manufacturer software and firmware.
By the black-hat route, you illegally access manufacturer proprietary information and publish it, possibly along with statements indicating the exact extent to which the company profits by arbitrarily limiting the self-repair abilities of its customers. This would make it easier for the grey-hats to write jailbreak exploits and discourage farmers from buying in to the abusive business practices.
If you ever want to be paid for it, white-hat is the way to go. You have to look at how farmers currently use the technology and then figure out how to replicate those use cases without the use of closed technologies. Think about how Tomato or OpenWRT help people use their wireless routers in the way they desire. You would essentially be writing new firmware for a piece of hardware, albeit hardware that can cost the owner tens of thousands of dollars if used incorrectly.
Without knowing anything about tractors, I would guess that the modern ones include a digital hardware controller that can monitor and control every aspect of a diesel cycle engine. They may include GPS receivers and mapping and routing software. They would also include modules to control the interchangeable tools that may be mounted on the tractor, which would require sensor feedback.
By this time, they should be largely self-driving, and maybe even have their own Twitter accounts. I'd save that crap for last.
The gray hat and black hat approaches are actually fairly profitable in the automotive industry and could probably sell well in the farm industry too.
Gray-hat reverse-engineered tools like Durametric for Porsche and VAG-COM for VW sell with moderate degrees of success into the enthusiast and independent service shop market (where $800 for a Durametric with 40% of the capability is a lot more appealing than $14,000 + service plan for a PIWIS).
The black-hat approach seemingly can be profitable (like all black-hat pursuits, depending on your moral beliefs) as well: various Chinese manufacturers make enough money selling knockoff Mercedes diagnostic "multiplexers" on eBay to make it worth their while to produce them.
There are a ton of value-add products you can sell once you've reverse engineered a control system, too, like custom tuning and add-ons unavailable from the manufacturer.
...with the associated legal risk. If you start making too much money, you need to carefully structure your company to be raid-proof and judgment-proof. Make daily off-site, offshore backups. Don't ever do sales and manufacturing in the same country. Put your supply chain in a different business structure than your customer management, which is likewise separate from your reverse engineering operation. Use strawman purchasers to acquire your OEM hardware for analysis.
All that represents overhead costs. If you don't pay them, your company could potentially be eaten by lawyers.
Only white-hats can sleep soundly at night. (They would be naive to do so, but they do have that option.) They are also the only ones able to accept an over-the-table buyout offer.
That said, the grey-hat route is probably the most fun, and likely to attract people from the "because its there" crowd. Judging from the jailbreakers in the console and mobile markets, they might just do it just to kick a multinational in the nuts and run away, laughing.
By the grey-hat route, you analyze the black box and patch out or bypass the security features in it. You allow the owner to jailbreak the hardware, while still largely using the original manufacturer software and firmware.
By the black-hat route, you illegally access manufacturer proprietary information and publish it, possibly along with statements indicating the exact extent to which the company profits by arbitrarily limiting the self-repair abilities of its customers. This would make it easier for the grey-hats to write jailbreak exploits and discourage farmers from buying in to the abusive business practices.
If you ever want to be paid for it, white-hat is the way to go. You have to look at how farmers currently use the technology and then figure out how to replicate those use cases without the use of closed technologies. Think about how Tomato or OpenWRT help people use their wireless routers in the way they desire. You would essentially be writing new firmware for a piece of hardware, albeit hardware that can cost the owner tens of thousands of dollars if used incorrectly.
Without knowing anything about tractors, I would guess that the modern ones include a digital hardware controller that can monitor and control every aspect of a diesel cycle engine. They may include GPS receivers and mapping and routing software. They would also include modules to control the interchangeable tools that may be mounted on the tractor, which would require sensor feedback.
By this time, they should be largely self-driving, and maybe even have their own Twitter accounts. I'd save that crap for last.