How We Grow

cut labor costs by having drivers manage fleets instead of just one tractor," said Monarch Tractor CEO Praveen Penmetsa. Penmetsa believes they are reimagining what a tractor can be. Large global farm equipment manufacturer CNH Industrial has bought into this vision by becoming an investor in Monarch. But many believe that the move to autonomous equipment shouldn't involve replacing an entire fleet. Companies such as Blue White Robotics, Kingman Ag, FieldIn and others are offering retrofit kits that give existing farm machinery autonomous capabilities. Blue White Robotics founder Ben Alfi believes that adaptability and accessibility are very important for growers who are utilizing autonomous capabilities for the first time. "We'll take any tractor that the grower has, no matter what type of color or size or design, and we'll retrofit it, connect it and provide on-the-job training and support," he said. These retrofit options allow the driver to switch back and forth from manual to autonomous depending on the task. They also provide a lower barrier to entry for growers interested in seeing if autonomy can provide sufficient return on investment for their operation. In some cases, it may make the most sense for autonomous technology to be offered as a service. This is especially true if growers won't need the equipment repeatedly throughout the year. As examples, Sabanto retrofits tractors with its technology and offers autonomous "farming as a service," and Rantizo is one example of spot-spraying as a service using swarms of autonomous drones. In specialty crops, companies like Farmwise offer robotic weeders as a service to vegetable growers. InsightTRAC is attempting a similar "robots as a service" model in almonds with its fully autonomous mummy removal rover. The unit is designed to roam through the orchards, spot mummies using computer vision, and remove them with a biodegradable pellet. The rovers are still being tested in California and Australia, but the company hopes to have a commercial service offering on a limited basis in 2023. 4 Future challenges There are still some technical barriers that need to be overcome to reach widespread adoption of autonomous technology, but economics and regulation also constitute significant factors. "A major challenge for autonomy is the cost. You have to make sure your autonomous system is cost-effective," said Dr. Stavros Vougioukas, professor of biological and agricultural engineering at UC Davis. Vougioukas said that in many on-farm use cases — for example, fruit harvesting — the majority of efficiencies and labor savings come from mechanization and automation, without necessarily replacing the equipment operator. "Whether you are replacing the driver or not doesn't always have a huge impact on the overall economics," he said. One additional charge to be aware of are software-as-a-service (SaaS) fees. These are generally not substantial, but it's a recurring charge to keep the software updated and current. A second significant challenge has to do with safety. Monarch Tractor recently had a petition denied by the California Division of Occupational Safety and Health that sought to clarify the use of driver-optional tractors without a human operator stationed on the vehicle. Despite this setback, Monarch continues to work with Cal/OSHA and sell its technology in hopes of regulatory clarity in the future. A recent challenge that the almond industry is all too familiar with are supply chain constraints. Some important components for these technologies have at times been difficult to obtain. Most in the industry seem to agree that this is a temporary problem that will get resolved over time, but it is certainly something to be aware of before investing in this technology. These challenges are real, but seemingly surmountable as the cost of technology trends lower, regulations get clarified and supply chains free up. The data flywheel Autonomous equipment has evolved as a result of the digitization of farming over the past couple of decades. Farm data has trained algorithms and artificial intelligence models to run the machines. Now with more autonomous machines finding their way into orchards, even more data will be generated through the computer vision used for guidance systems and the various sensors these vehicles are equipped with. These massive data sets will be used to create even more useful models and tools for growers in the future. 5 Dr. Mason Earles at the Plant AI and Biophysics Lab at UC Davis is preparing for this future by building models that can estimate yield based on video data. Earles' work focuses on full automation solutions from sensing and data pipelines to data analysis and grower insights. 3 "CNH Industrial, Monarch Tractor agree electrification technologies deal" https://reut.rs/3BmfiIp 4 "Cleaning Robots: Out of Your Home and Into Your Orchard" https://bit.ly/3Dn8Ylt Continue on page 11 Monarch Tractor's fully electric, driver-optional autonomous tractor. Courtesy: Monarch Tractor Almond Board of California 10

• Cover