Let‘s talk privacy in an open software platform for farmers
A lot of people have concerns about sharing farming data with others. Who could blame them? Data is always delicate and even more so for farmers.
If you are a professional farmer, you see your plant recipes as part of your intellectual property, maybe even your USP. Some have even described them as the Coca-Cola-recipes of farming. And even if that’s not the case, a particular batch might be an experiment, for the moment. Who would share data about experiments without validating their outcome first?
Another huge concern is safety. I read the following sentence about twice a week in posts on LinkedIn: “If you can control your farm over the internet, anybody can!”. I’m not here to tell you that this concern is not a valid one. I’m here to spark a conversation on how we can deal with these issues while still benefitting from digital services. And I truly believe we need to overcome them in order to build the open system we are dreaming about here at farmee.
Inventory
Let’s step back and have a look at what we’re actually talking about:
Farming data
What is farming data, to begin with? It’s a bunch of numbers describing values sensors are measuring in the farm. Temperature, humidity, CO2, PH, EC/ppm, etc. The combination of these numbers as well as linking them to a certain crop can be described as a plant recipe. It is a description of the climate a crop needs to successfully grow.
Control impulses
Another part of the message exchange is going into the opposite direction: control impulses used to tell the actuators what to do and when to do it. When should the lights go on and of? How much water is supposed to be pumped into the irrigation cycle?
Quality of data
I’ve used the term successfully grow before. This is an important distinction, because not every grow cycle is a successful one. Plus, what do we define as successful? Is it quantity or taste? Taste can be a very subjective criteria as well… So we need to implement some kind of quality control, either to be performed by a human being or some other measurement procedure, e.g. measuring weight or other pre-defined conditions.
Transmission
Messages are exchanged between computer systems, typically local installations and servers, e.g. used to offer cloud services. There also is a graphical user interface to involve the farmers, give them valuable insights or put them in a position of control over his or her farm, as well as notifications in case of emergency.
Data belongs to the user
Your farm, your data. Simple as that. Well, not quite. Let‘s have a look at different approaches to secure farming data:
Why not keep data on-premise?
If privacy is a major concern for your operation, you should do just that: keep your data on your site. Let’s assume you’ve just discovered the Coca-Cola-recipe of the century, say, you’ve just grown vanilla in a vertical farm and yes, they did blossom and you’re waiting for your first yield, DON’T SEND YOUR RECIPE OVER THE INTERNET. Thanks.
How to benefit from data
Let’s face it: Most of us are not working on the cutting edge of what’s possible in indoor farming. But we could all benefit from a little communication between farms. Let’s have a look at some value-adding services using aggregated data:
Compare yourself
Is your crop as energy efficient as other farmer’s? Do you need to spend more or less money producing it? How much could you charge for it in consideration of input and current market prices?
What if your software platform showed you aggregated values including mean, upper and lower limits? Not traceable to a specific farm, but still valuable for you as a farmer and entrepreneur. Is my nutrient consumption on-par with the industry standard? Should I use more energy efficient lights? This stuff helps, a lot! It eliminates a lot of legwork and guesswork, trying to figure out everything on your own.
Side note: Knowing your product
This has nothing to do with sharing data, but still it’s so amazing:
by knowing exactly what goes into your plant, the software could easily help you figure out pricing, find buyers or even certify the plant to a certain standard, if your parameters are within the limits of that standard, e.g. a sustainability certificate. I have to admit, I kind of stole that idea from a phone call with Mark Horler. My mind is racing with ideas surrounding this basic idea of plant input assessment and I‘m sure we‘ll dive into that a lot deeper with another story some time later.
More advanced data-driven services
So it’s pretty nice to compare yourself, but can’t data and algorithms do a whole lot more? Oh, you bet!
Let’s assume the farmer is suffering from a certain disease taking away from her yields every crop cycle. This shit is costing her hard money, but still is very hard to get by manually.
Let’s further assume an algorithm was able to prevent just that disease. To feed that algorithm different sensor values were necessary. This is not possible using aggregated data, rather opposite: the better and more precise input data, the better results would be.
No data, no algorithm
This is a bit of a trade-off: no data, no algorithm. The classic predicament of the modern homo deus (who are we?). But we‘re not here for dystopian warnings, but rather for an informed discussion and resulting decisions.
Let’s state this clear and simple: We need your data for certain tasks. And we need to hand it to other companies, if you want to use third party applications and more sophisticated algorithms.
The good news: this is standard operating practice for all software platforms out there. And if done right, user privacy rights can easily be protected. Your data belongs to you, so all that is granted is access to certain data, necessary to perform a specific task. Our technology of choice to implement such access is OAuth, if you‘re into such things: look it up!
Bells and whistles
Of course, we‘re also using https and security-certified servers to store your data. No computer system will ever be 100% secure, but the most probable source of data leaks will remain the user. In our case, this means there is an API key for every user. If some shady hacker got access to that key and your credentials, he could potentially access your data. But the good news is: only your data, no-one else‘s. Ok, that‘s not really good news for you, in that scenario, but for everyone else!
Cross-examination
What about these caveats I’ve been talking about before and what is our answer?
„If you can control your farm from your browser, anyone can!“
Yes, if you are careless with your credentials, that could happen. Everyone can control his or her bank account via the browser, so in general anyone can access your bank account if you are careless. But if you take good care, there‘s not too much to worry about.
I still don‘t believe you, you morons!
Alright, we get it. You did find out how to cultivate vanilla, right? We will enable two-factor authentication for people like you. We would also want that for anything bigger than a small home installation. The second factor would link your account to a phone number, for example. As long as you‘re the only one with access to that phone, your good.
Forget it, you will never get my data!
No problem, there will be an on-premise installation of farmee as well. But then you‘re really on your own, meaning: you have to install it on your own server and build a fence around it. Maybe you could even hire some muscle with big guns.
Just kidding, but local installation has some caveats, obviously. Yet still, as stated above, for sensitive operations, it‘s still the best way to go forward, despite being more costly and labor-intensive.
Wrap-up
We take privacy pretty seriously, as we understand it is crucial for professional farmers to remain in control of their farm, their plants and their intellectual property. But we do believe potential benefits heavily outweigh potential risks. If done right, there is nothing to fear, but a lot to gain!
If you have any questions, please ask them, we would be happy to intensify the discussion on such an important topic. Let‘s get to it!
