Aigen's Alchemy trains farm robots on new crops in under a week

Aigen's approach pairs a simulation platform, Alchemy, with its solar-powered Element robots. Alchemy generates synthetic, pixel-level annotated training data, which Aigen says can be up to 100 percent synthetic, while grounding simulated conditions in data from its real-world agricultural operations. This lets Element learn a new crop before it reaches the field. The company says Element recently learned lettuce in simulation and was deployed in under a week, which Aigen says could reduce the time needed to adapt its robots to different crops.

Aigen has built more than 100 robots and accumulated over 15,000 autonomous operating hours across California, Minnesota and North Dakota, working with crops including soybeans, tomatoes, sugarbeets, lettuce and cotton. Its Element Gen 2 X4 robots are also being used by Fortune 500 companies for weed control, agricultural data collection and analytics. The robots are available with two, three or four robotic arms and operate on solar power with battery backup.

Aigen is taking orders for the 2027 growing season and plans to demonstrate Element at FIRA USA in Yakima, Washington, from October 20-22. Whether the under-a-week training claim holds across different crops and field conditions may be the test for growers weighing whether to adopt the system.