Innovation in agriculture has been soaring to new heights, quite literally, with drones mapping fields and satellites beaming down real-time data. Software solutions are guiding decisions at an unprecedented scale, revolutionizing the way we approach farming. However, as Reservoir CEO Danny Bernstein points out in a recent CropLife article, there’s a layer of complexity that these advancements have yet to fully address. This complexity lies closer to the ground, in the rugged terrain, fragile crops, and unpredictable labor that characterize open-air farms. It’s a realm that demands something more tactile, more hands-on. Enter physical artificial intelligence (AI), technology built to operate in the dirt, alongside the crops, in real time.
Bernstein, whose career was shaped in the fast-paced, iterative world of Silicon Valley software culture, acknowledges that agriculture plays by a different clock. The unpredictability of open fields doesn’t adhere to roadmaps or release schedules. The stakes are high, with grower balance sheets, rural economics, ag communities, and food systems hanging in the balance. Yet, there’s a valuable lesson from Silicon Valley that’s worth transplanting to the agricultural sector: the habit of questioning assumptions.
What if startups worked in sync with growers from the outset? What if tools were designed to adapt to the timing of the crop, rather than forcing the crop to fit a technological schedule? What if the people designing the tech were present in the field when it was being used? These questions are the driving force behind Reservoir, an innovation community for agtech that Bernstein describes as an “Olympic Village.” Starting in California, Reservoir brings together startups, growers, and retailers to work side-by-side, pressure-testing ideas in the field rather than in isolation.
This approach is a stark contrast to the often-isolated development of agtech, where teams removed from the realities of the farm design solutions for ideal conditions that don’t exist. Agriculture is shaped by unpredictable variables—weather, labor, regulation, crop biology—that don’t respond to roadmaps or pitch decks. The 2024 CropLife/Purdue University Precision Agriculture Dealership Survey underscores this point, projecting that the use of drones for crop input application will grow from 35% today to over 50% by 2027. Similarly, AI-driven weed identification is set to become a standard offering for a quarter of dealers in the same timeframe.
Yet, despite these advancements, the reality remains sobering. Less than 2% of specialty crop production is currently automated, according to the Western Growers Association. This gap presents a massive opportunity for innovation, but it also highlights the need for a more grounded, collaborative approach. The Reservoir model, with its emphasis on hands-on, immersive development, could be a significant step in this direction. By fostering a community where technologists are deeply embedded in the day-to-day realities of farming, we might see agtech innovations that are not just disruptive, but truly transformative.