How Technology Best Fits With Food & Ag

With all the buzz right now about AI and Agriculture, I’ve been thinking a lot about the broader “Meta” perspective of emerging technology, decomposition my thinking from big ideas to the smallest ones to try and get some get some perspective around first principles (or “The first basis from which a thing is known.”) to get a better vantage point on how technology best fits into Agriculture. As I step into my new role at Agerpoint , I thought it was a good time to codify my thinking into a post and get other people’s feedback.

If we take a step back though, and think about where technology (defined as “the application of scientific knowledge for practical purposes”) fits into agriculture more broadly, I start with a legendary story from another time and place to help frame up my perspective.

Many of us have heard about the US building a space pen that cost several million dollars to develop during the early days of the space program, while the Russians used a pencil. It turns out, it’s mostly a myth (thought there are certainly parts that are true ). The real story behind that is that both the Russians and Americans used pencils but there was a concern that graphite would break off while the ship is in zero gravity which would then enter into the electrical system and short something given it’s conductive properties. Both sides I assume measured the risk, and went with a pen developed by the Fisher company (who did spend $1m to develop it) that could write in zero gravity which was then used throughout the space race by both sides for a couple of dollars per pen.

I think about farming, and the conversations I’ve had with farmers from which many feel like the state of AgTech is focused on developing high tech pens that write in zero gravity, and the pencil folks are using doesn’t work THAT bad so there isn’t a need to embrace something new when comparing the risks to benefits. The real cost lands on the AgTech company to do the expensive development, in hopes of providing a meaningful technology solution that replaces the cumbersome incumbent but doesn’t often get the same payday Fisher got when the space race adopted their pens after the $1m spent up front to develop it.?

I think this comes down to three things that need to be done better, if I break down my experience?

1. Making the technology adoption / transfer process as painless as possible

I talked to a farmer who sat on an AgTech advisory committee one time about technology adoption on a farm and his input was that companies can greatly overcomplicate what it takes to adopt a solution on a farm no matter how useful the solution is. The solution has to cross technology backgrounds, language barriers, education levels, and diverse climates / environments anticipating issues around situations such as poor connectivity, and adverse weather conditions. Though this should be considered a no-brainer for anyone working with agribusinesses, it’s an area that I hear often gets missed.?

One thing that got me excited about Agerpoint is that the software running on a smart phone makes it very straight-forward, using simple colors and no words, to properly capture images and video which is then sent off (at the next possible cell connection) where the complicated work happens behind the scenes.?

I like to think of this much like going to a butcher shop. When you walk in, you’re greeted with a nice clean straight forward display of meats and steaks, with none of the mess associated with preparing the meat present in the store. The butcher doesn’t talk about what knives he used to make the cuts, describe specifics on how the preparation occurred, or other extraneous information not related to the purchase and consumption of a steak. It’s usually a series of steaks in a well lit display cabinet with the relevant information around marbling, flavor, and price. What if AgTech solutions were presented in the same way, only discussing and showing the parts that really matter while the technical details and mechanisms are stored in the metaphorical back room? It relates to sales, training, and day to day use I think.

2. Using what the agribusiness already has on hand and improving what they can do with it?

In the example of the space pen, I wonder if it would have worked to make a mechanical pencil with a more effective pencil lead alternative vs resorting to zero gravity pens and ink with the cartridges to go with it. There may be a good deal of costly materials science involved that wouldn’t pencil out in the end (pardon the pun) but I think the parallel in an AgTech context is leveraging expensive hardware-based solutions to drive innovation (drones, ATVs, sensors, etc) vs improving on what’s already on the farm (tractors, smart phones, etc). The technology adoption cycle can increase in pain when new hardware is involved, which is sometimes unavoidable in the case of robotic applications, but maximizing the benefit of a previously owned smartphone for example, with all the sensors and compute power it already has within it is a great way to build a digital beachhead given just how many smartphones are already present at farms globally.?

Agerpoint can take a picture captured on any hardware, from drones to smartphones, with this principle in mind. There’s no hardware burden, complicated onboarding, or expensive training - just point, use the color coded guide on the device, and shoot. This not only improves the efficacy of rolling out the technology, but also makes the solution adaptable to use cases ranging from counting soy plants to checking corn crops overhead for damage with the impressive and complicated work happening in the cloud.?

3. Adapting the solution for the need

Sometimes all a farmer needs is a digital “clicker” to help count plants, which is more effective than having to count it themselves. In other cases, the farmer needs a set of 3D models and complex analysis over several crop years to better understand the changes in their yield. If the same technology can capture the data easily enough, but the use case can adapt to the need without the need to use multiple solutions in multiple use cases then stitch it all together on the backend, eliminating 15 point solutions for 15 different needs, you can be one of those rare solutions that can operate as a platform and a point solution at the same time. The backend of your platform can leverage concepts like micro-services, common and unified data models, complex ML/AI based analysis, and semi-structured data capture and storage while your front end is a simple app that points, focuses, and clicks images that can be used across the infinite number of use cases involving scouting, surveying, predicting, analyzing, etc.?

Starting from an architectural direction around adopting points 1 and 2 above, while also embracing a backend that can match the need as it’s presented without splintering the stack and creating excessive technical debt can give you the kind of solution that meets the need while anticipating what the long list of needs could be. There’s no silver bullet, but there are swiss army knives.?

Agerpoint had this philosophy from the get-go, which made me so excited to help build on the platform they’ve constructed to date. The list of customers and use cases is significant, and they have a product that can meet the need but be additive across use cases without breaking apart in the process.?

If I was to summarize all of this up, I think of a Farm CFO I knew that told me every investment he makes he wants to see a 3 year or less payback period. I called the category “Farm of the Near Future” investments vs a “Farm of the Far Future” investments that don’t have a known payback period, takes too long to cover the cost, or sits in the “Science Project” bucket that get a free five acre demo to kick the tires on the technology but never get past that phase. Agerpoint has done a great job navigating the AgTech arena to land in the category of a “Farm of the Near Future” solution which is why it’s gotten adopted to the extent it has.?

There’s no doubt more points that could be made, outside of the 3 I listed above, that aren’t technology or commodity specific, so I welcome feedback on what other foundational principles make for a successful AgTech solution.?

Disclaimer: I talk about AgTech, which is a broad category, and has lots of successful solutions that don’t fall in the category of being easy to implement, easy to use, or adaptable to dozens of use cases. I suppose I could use a phrase like “FarmTech” instead, vs software used in a back office or in a processing center, but I hope you all frame this up around the context of technologies used by growers / ranchers / farmers specific to interacting with their agricultural good throughout the year vs other niche use cases that still may fall under the category of “AgTech” but would no doubt have a different list of useful principles associated to it.