Less than a decade ago, the widespread adoption of robots on farms sounded like a futuristic fantasy. But today, the shift to farm automation is increasingly viewed as inevitable. At the recent World Ag Expo in Tulare, CA, autonomy-enabled robotic companies were on full display. From automated tractors to weeders, sprayers, and harvesters, it’s clear that autonomy in agriculture is reaching a tipping point of commercialization. In walking the show and talking to industry participants, we identified a few key factors that have driven adoption of farm robots, along with headwinds that still challenge the sector. We continue to see opportunities for growth in the space and are encouraged by the partnerships forming that enable a full-service solution for the farmer and scale in the long term.
Labor, Labor, and Labor
In asking what were their top 3 challenges, we heard from farmers and other agricultural partners at the Expo a recurring theme. The availability and rising cost of farm labor are typically the most significant concerns for the viability of farm operations today. 70 percent of produce growers and manufacturers struggled to recruit adequate staff for the 2021 season. According to a recent Western Growers survey, average labor cost exceeded 50 percent of total production costs in 2021. Around 50 percent of respondents also expect labor costs to increase by 10-30 percent in the next 3-5 years. Labor shortages can lead to less vigilant weed, disease, and pest management or food unharvested, resulting in food waste and economic loss for farmers.
Additionally, rising costs and scarcity of other resources, namely water, and the environmental impacts of overapplication are pushing farmers to be more precise with resource use. With rising labor and operational challenges, automation has become the proverbial “pain killer” for growers, moving beyond the “vitamin” of using cool technology. In the same Western Growers survey, 65 percent of participating growers indicated to have invested in automation over the past three years, with an average annual spend of $350,000-$400,000 per grower. Autonomy is beginning to provide a clearer value proposition for farmers.
Finding Early Adopters in a Niche Is Critical
Ag robotics companies stressed the importance of working with innovative growers. Farmers must be willing to learn a new system to adopt these technologies successfully. While companies were still figuring out the technology the past few years, certain early-adopting customers were patient and willing to work alongside developers, possibly even changing their operations to maximize gains from operating a robot. Getting true, honest, and rapid feedback from your customers during the iteration cycle is also critical. Companies were in agreement that good customers will grow with you.
Perception Has Reached Commercial Viability
Advancements in computer vision, machine learning, processing speed, and graphics processing units (GPUs) mean humanity has gotten very good at training computers to recognize something in an image quickly and accurately. This ability to “perceive” and distinguish between objects has empowered several farming applications to become autonomous to the point of commercial reliability - perhaps most notably in weed management. There were a handful of leaders in the autonomous weeding space at the show, attracting both funding and paying customers. “See and spray” robots have also seen adoption. Finally, the robustness of localization and navigation has driven wider-scale adoption of robots that can move around the field and serve as a platform to carry tools or take over a tractor as the driver. As Moore’s Law has shown with microchips, sensing hardware will continue to get cheaper, and the ability to store and process data will continue to improve, making better AI technology more affordable and widespread.
Field Operations Being Optimized for Robots
Genetic innovations and field designs that are better tailored to automation can help propel the sector forward. For example, new breeding variants can emphasize traits that facilitate the harvesting process for robots while delivering the same or better products to consumers. Trellising techniques can force fruits to grow in two rather than three dimensions, supporting robotic and human harvesting. At the Lincove Research and Extension Center in Exeter, CA, researchers are experimenting with high-density planting and 2D trellising to reduce labor needs and open the door for robotic harvesting.
While automation has been working its way into the agricultural industry for some time, the road to commercialization for true autonomous robots has not been without its challenges - some of which are listed below. Because of these obstacles, many companies have fallen into the “valley of death” before reaching scale and profitability.
It’s Expensive to Build Robots
Many take the “Apple approach” - own the core hardware and software. This is generally sound logic for when you need the robot to physically do something (most of the time if you are solving a labor challenge) so you can iterate quickly. However, this also requires two teams of advanced degree-holding professionals - mechanical and electrical engineers to develop the hardware as well as computer vision and machine learning experts to build autonomy. Not surprisingly, these individuals are rare and generally have other options for very high-paying jobs. The war on talent is real, often resulting in a high cash burn for startups in the space.
Ag Requires a Rugged Robot
Walking around at the World Ag Expo can leave a thick layer of dirt that cakes your face and clothes. The Central Valley is a challenging environment compared to a climate-controlled indoor warehouse with air filtration systems that keep out dust. These robots must not only perform in harsh environments and on uneven terrain but also in 110+ (°F) weather. It makes mechanical manipulation and soft robotics requiring dexterity that much more challenging.
Fundraising Market Has Tightened
It is well-documented that venture funding took a tumble in 2022 of over 35 percent from a peak in 2021 back to pre-COVID levels. According to Crunchbase data, robotics-related venture funding fared even worse, down 44 percent from $17.4 billion in 2021 to $9.8 billion in 2022. Funding, while challenging, is essential for such a capital-intensive sector. With funding opportunities tightening over the next year, companies must be as efficient as possible and focus on realizing ROI to justify commercial use.
Farmers Need Help Driving Actionable Insights From Data
Many robotics companies we spoke with at the World Ag Expo view data collection and actionable insights as the key to driving automation adoption and ROI for farmers. Crop data analysis requires an agronomic layer on top of the automation technology, which is still in the development phase for most companies. And even when these products can drive meaningful insights for crop management, they will only really solve the labor problem if they are then able to respond to those insights. The next evolution of robotics will be turning the collected data into value for farmers and having the robot act on that data in real-time.
So how can the industry move forward? While there are many paths to commercialization, we have seen focus and partnerships driving success for farm automation companies.
Companies Must Be Resource Efficient in Demonstrating Early Commercial Milestones
Frontier tech requires focus. It is hard to be good at all things when tackling very tough technical problems. The companies that have seen the largest deployments in the field have nailed a specific use case that demonstrates customer ROI. Investor patience has waned for investing in cool technology alone. Showing you have a commercially paying customer (e.g. beyond pilot trials) is the best way to attract funding in this market.
Partnerships Key to Bringing Complete Solutions to the Customer
If companies focus on one use case and perfect that technology layer, they can be resource-efficient and partner with other companies that have developed adjacent capabilities to provide a more comprehensive solution to farmers. Industry collaboration and partnerships will be essential for getting the technology to where it needs to be while staying capital efficient and offering a viable product to farmers. Several companies are retrofitting tractors with automation software and hardware or building autonomous implements that work with the machinery farmers already own. OEMs are increasingly seen as a path to scale, with some companies looking at dealership models to sell directly to OEMs. Other companies are collaborating to add additional capabilities to their products. For example, Burros support BloomX’s bio-mimicking pollination technology to manage the pollination process and drive increased output in blueberries. Burro is also working with Bitwise, an AI crop analysis technology that can be mounted on a Burro for scouting purposes
Ag Can Learn from Partnership Parallels Further Down Supply Chain
Earlier this month, we also attended the ProMat supply chain conference to understand automation happening across the value chain. One example is Locus Robotics which is quickly scaling in the warehouse space. Locus is squarely focused on movement and transportation of goods within the warehouse. They partner with other companies like RightHand Robotics to do the picking of parcels or Berkshire Grey for robotic shuttle put walls (think automated bin retrieval). Although indoor robotics and automation arguably face simpler and fewer challenges in the environment, successful commercial traction via partnerships can provide a framework for advancing agricultural automation.
At S2G, we look forward to continuing to explore this sector, support collaboration, and realize the benefits it can have for the agricultural industry.
To learn more how robotics are bringing new solutions to the food and agriculture sector, read our report, Trends Shaping the Future of Food in 2023.