AgriPass launches AI robot that weeds like a human

The Israeli startup AgriPass introduced an AI-powered robotic weeding system that replicates human decision-making in the field, aiming to solve labor shortages and reduce reliance on chemicals. According to AgFunderNews, CEO Liron Cohen Yanay explained that the technology enables precise, selective weed removal at scale, offering a new approach to farm automation.

The company positions its solution as a shift toward “ag robotics 2.0”, combining artificial intelligence with mechanical precision to replicate how experienced farmers identify and remove weeds. Unlike traditional systems that rely on brute force, AgriPass focuses on context-aware decision-making, allowing the machine to distinguish between crops and weeds in real time.

“What we do is human weeding at the farm scale,” said Cohen Yanay. “We are replicating the human weeding process and making it affordable at scale.” He added that “AI handles contextual analysis and decision-making to identify which weeds to target and how, in the way an experienced farmer would in real time.”

The system integrates AI-driven vision and independent robotic arms, which act like multiple hands working simultaneously in the field. Each actuator performs targeted actions, removing weeds without damaging crops or soil structure. This approach contrasts with conventional mechanical weeders that cut indiscriminately.

A key advantage of the technology is its adaptability to changing field conditions. Soil moisture, crop type and weed species can vary daily, requiring different interventions. AgriPass claims its AI can adjust actions dynamically, mimicking human judgment rather than following fixed patterns.

The company is initially targeting small and medium-sized farms, particularly those growing high-value crops such as vegetables. These farms face frequent weeding cycles and often struggle with labor availability. The system is also designed for regenerative and organic agriculture, where chemical herbicides are limited or prohibited.

AgriPass has already tested its technology across multiple crops, including brassicas, tomatoes, squash and melons, with plans to expand to onions, carrots and other high-density crops. Training the AI model has become faster over time, dropping from months to just a few days thanks to improved data annotation.

From a financial perspective, the company emphasizes return on investment (ROI). Farmers using manual labor can expect payback in about one year, while conventional farms may see returns within two to three years. The system is priced below $200,000, a level the company considers accessible for many producers.

The business model combines equipment sales with a subscription-based software platform. This platform provides farmers with insights about field conditions, weed patterns and crop health, adding a layer of data-driven decision support.

AgriPass is also pursuing a partnership strategy with OEM manufacturers to scale production and distribution globally. The company has already partnered with Fyeld Group in Europe and is exploring similar agreements in the United States and Latin America.

Field trials are ongoing in regions including Italy, Georgia and California, with early commercial traction reflected in initial bookings. “We have people that believe in the product,” said Cohen Yanay, highlighting growing interest from farmers.

The broader vision behind AgriPass reflects lessons learned from the past decade of agricultural robotics. Instead of pursuing full autonomy immediately, the company focuses on practical, scalable solutions that farmers can adopt today.