TerraScout brings robot agronomic scouting to the farm

"It can drive through the field and look at every square inch. We really honed our craft on model performance and detection with rocks, and now we've transitioned that into weeds. The robot uses a boom-mounted camera system to capture detailed imagery across the field. With eight cameras operating at 1 millimetre resolution, TerraScout can generate billions of data points per acre, allowing it to identify specific objects, including individual weeds."

"That imagery is processed using AI models trained on datasets collected across millions of acres, creating a detailed view of field conditions at the plant level. The machine carries a 60-foot boom and operates at speeds of up to 15 miles per hour. Depending on field conditions, TerraScout is expected to cover 50 to 85 acres per hour, with the ability to operate continuously, day or night."

"From a weed management perspective, the system provides a digital interface that can inform herbicide timing, tank mixes, and application rates. Data can also be exported, allowing growers to use existing equipment for spot spraying."

TerraClear introduced TerraScout, an autonomous scouting robot that moves through fields collecting high-resolution data for actionable farming insights. The robot features eight cameras operating at 1 millimeter resolution, generating billions of data points per acre to identify individual weeds and crop conditions. Built on machine learning models originally developed for rock detection, the technology now extends to weed identification and crop monitoring. The robot carries a 60-foot boom and operates at speeds up to 15 miles per hour, covering 50 to 85 acres per hour continuously, day or night. Output provides practical guidance on herbicide timing, tank mixes, and application rates, with data exportable for spot spraying with existing equipment.

Read at Realagriculture