Hacking the Tractor: what the future of farming means for open science

I took a gamble when I bought the tractor mower. It was old, but well-cared for and ran well, but it was wearing out. Still, I couldn’t get a quarter as much machine for twice the price. Two months later, the mower deck cracked, rusted out from ten years of hard use. A replacement deck would run into the thousands of dollars, there were no options for just the shell, a cheaper, but still pricey option, and there were no off-brand decks to be found. Those three facts alone should tell you that we’re dealing with a John Deere, here.

John Deere made headlines earlier this year when Wired ran an impressive (and rare) expose on farm tech: New High Tech Farm Equipment is a Nightmare for Farmer. Deere followed up with a rebuttal letter, declaring that farmers did, in fact, own their equipment, while simultaneously outlining all the ways in which they didn’t: John Deere: of course you “own” your tractor, but only if you agree to let us rip you off.

Here’s the short version: Tractors are complicated and increasingly controlled by onboard computers. These computers use proprietary software, and that software is protected by the Digital Millennium Copyright Act (DMCA). The software itself is encrypted to prevent copying or modding. Those encryption are also protected by the DMCA. Breaking that encryption is illegal, regardless of the state of the software. If you have to decrypt the software to remove it, you’re breaking the law.

Here’s another example: I have published a few books online. I choose to publish them Digital Rights Management (DRM)-free, so that you can do what you want with your copy. Many authors don’t. If I had published a book with DRM and wanted you to read it, and sent you my downloaded copy, complete with DRM that you had to crack (which is trivial) you might think that’s fine. I am, after all the copyright holder or the book. You’d be wrong. I don’t own the encryption, it is, in this case, Amazon, and they could go after you for breaking their digital lock, despite the fact that I, as the author gave you permission. A book is not a tractor, but if you buy either, you should own it. Reading a book however you like and reading the engine diagnostics off your own equipment shouldn’t require breaking the law.

Especially when the law is hopelessly complicated.

Back to the mower deck. I’m not unskilled when it comes to building and fabricating. The deck was rotting, but it wasn’t all rotted out. So I did what any self-respecting Maker would do, I ripped it apart and put it back together in a configuration that would keep working. After a few hours of metal work, some heavy grinding, and a bit of cold welding, I had a functioning mower. If there were still a warranty attached to it, it would most certainly be void.

I’m lucky, the tractor mower was old enough to be largely free of software-driven control systems. So desirable are purely mechanical machines, that vintage John Deere equipment often fetches a higher price than brand new machines. Newer tractors, and even family riding mowers, have software that regulates and restricts how much they can be altered. Re-positioning pulleys and rerouting belts may cause these machines to lock the user out and prevent them from starting. Ostensibly this is about safety, but safety is a relative term, and is losing an entire growing season because your farm equipment decided to brick itself really “safer” than a skilled handyman making a useful modification?

The ABI 3730xl DNA Analyzer is an awesome piece of equipment. Sure, these old school sequencers are being phased out in favor of “Next Gen” machines which can churn out terabytes of data, but, more than 10 years after their introduction, these beasts are still plugging away, producing data by the well-full. As labs replace these and other Sanger-sequencing machines with newer, fancier equipment, the secondary market for older sequencers is booming. They’re cheap, reliable, and there’s a wealth of literature and methodologies already validated on the platform. The 3730 could be a workhorse for establishing molecular facilities in places where they were previously untenable. They could fuel a boom in conservation genetics in the developing world and other regions where capital equipment has been historically difficult to obtain.

Can you guess what the problem is?

I’ve visited numerous labs around the world that are operating second-hand sequencers. The machines themselves are robust, reliable, and reasonably affordable. The software is not. Crouched next to their prized sequencer is, in all cases, a dusty, withering computer, running the original software installation. They can’t update without a license costing almost as much as a new sequencer. They can’t transfer that software, even the same version, to a new computer. Like my farming friends and their John Deere tractors, they own the machine, but not the core software that runs it. Which raises the question, what do they actually own?

When the last ABI 3730xl goes offline, it will not be because the sequencer itself no longer functions. It will be because the software to run it is locked.

Look, I’m not picking on ABI, here. They aren’t the only biotech company going to extraordinary lengths to protect their digital property. And it’s not just digital. Look around your lab. Do your pipettors require service and calibration by an expensive technician? Do you genuinely think none of the PhDs in your lab are capable of calibrating a pipettor? What if your lab is in the middle of the Pacific, 6000 miles from the nearest technician? Should you have to choose between voiding a warranty or voiding your research?

I’m not saying that companies don’t have the right to protect their IP, they do. But the systems and locks they put in place to aggressively police the use of their tools is a barrier to open science. As older equipment falls out of favor, the very equipment which could go on to live a robust and vibrant afterlife in labs with funding and access woes, those machines may cease to function completely, not for want of service but for lack of a software license, with no recourse from the consumer. The more open our software is, the more people can do good science.