Chris Kroger put a great article together in Vision Magazine that should remind everyone in AgTech (yet again) of an important truth: the problem you are solving must be one the grower recognizes and agrees must be solved. If not, it is much harder to get them to do trials with you and even harder to get them to write a check for the solution. The article’s premise is that AI is influencing the buying process in multiple ways, including research on the space and competitive research, comparing similar products, and doing cost analysis (likely for a total cost of ownership style analysis). It’s worth noting that the use cases for AI for ag and AgTech practitioners have some high overlap because the research is useful for both groups.

The nuances of integrating AI into a supply chain process, an automation process, or an analysis process are often the real reason for failure. The tech can be fine, but it has to be delivered to the user in a way that is fine for them, and depending on how tops down the approach is and how well thought out the roll out was, things can go sideways. Lots of fancy AgTech had plenty of AI in it over the last several years (remember that AI for hardware and software R&D and development was a thing way before ChatGPT launched about three and a half years ago) sits dormant in equipment graveyards on farms all over the place. And in many cases, that’s after extensive testing was completed, lots (no – really, lots) of math was done, lots of time in spreadsheets doing analysis of before and after costs and then working with the operations team to get it all integrated.

The machine was evaluated, purchased, and rolled out. So there are some huge costs in those automation tools that sit idly by and are not getting used at all – not the least of which is that all future automation solution providers have to get growers past the concern that they’re the next piece of equipment in the equipment graveyard. In some cases, harvest crews have chosen not to use a Burro where ROI for the farm workers was clearly shown (i.e. you will make money for your harvesting work because the Burro will take the product back to the truck for you automatically). This was after extensive work by growers and Burro to confirm the math and the best way to integrate the machines into the operation. Where on the startup risk factors is “workers may not use the machine even after successful trial and ROI math is confirmed” – that is a bullet point that does not go on many Board Meeting decks. The best AI in the world doing the best job of helping the automation solutions move around and do their job thwarted by a human factors decision that was made at the last mile – the interface between the automation solution and the people doing the work (and that is almost always the right place for that decision to be made).

The first half of the article talks about retailers and tech, including Walmart’s roll out of RFID tags to track fresh product shipments. Traceability is a big deal, and retailers have invested plenty in helping to make it happen. How you integrate AI solutions that are embedded directly into solutions or AI tools that are used directly by the team internally will have an impact on how teams work and the results they can deliver. At this point, it’s almost strategic malpractice not to at least be exploring potential AI use cases ­– and maybe even worse strategy to not have a plan B if something goes wrong and something is going to go wrong or at the very least unexpected things will happen. It’s technology.

The second half of the article covers the automation space generally, including the commonly stated refrain that growers like to see a payback on automation investments of less than three years (and fewer than two AgTech solution moves up in interest level). The Western Growers Center for Innovation and Technology is now 11 years old, and the team has been focused on enabling startups and growers to work together more effectively since before the WGIT’s launch. The current focus for Ben Palone, Emily Lyons, and me is to help startups improve their understanding of grower challenges and reduce the cost and time they need to get to first product. A huge focus of the last two years has been the partnership with Danny Bernstein at Reservoir Farms and Sean Sundberg and Greg Christensen at John Deere. Together, we have built a partnership that all three organizations feel is vitally important and strategically built around the notion of having on-farm robotics incubators in different regions with different crops supported, shared R&D space, and shared John Deere equipment to keep both time and dollars lower for startups.

In addition, the technology validation space is heating up. Ben Palone has delivered Carbon Robotics, Stout Industrial, and Cal.NET WG Case Studies with detailed analysis that shows how the ROI is determined and then templates to help growers do their math (because the math varies depending on labor source, whether the labor is H-2A or not, and whether you lease or buy tractors, among other factors). Reservoir Farms is going to help startups with tech validation feedback as well, and we are working with other partners on helping. The reality is there are 700+ automation startups and all need several tests, so there is room to go around for testing work and write ups. As I say often, it takes a village to raise AgTech startups – and some capital.