Carbon Robotics launches plant identification AI for agricultural robots

A new AI model enables farm robots to identify and eliminate weeds instantly, giving farmers real-time control without retraining machines or relying on chemical herbicides.
Trained on millions of plant images, Carbon Robotics’ latest AI model allows autonomous robots to recognise unfamiliar weeds directly in the field.

Carbon Robotics has introduced a new artificial intelligence system designed to identify plant species in real time, changing how weeds are managed in agriculture. The model allows farmers to decide what should be removed directly in the field without retraining machines.

Called the Large Plant Model, the system is trained on more than 150 million plant images gathered from farms across 15 countries. The technology now powers Carbon AI, which controls the company’s autonomous LaserWeeder robots.

Earlier systems required new data labelling whenever unfamiliar weeds appeared or conditions changed. With the new model, farmers can instantly flag unwanted plants through the robot interface, even if the species has never been seen before.

Carbon Robotics says continuous data from its machines will further improve accuracy over time. Backed by more than $185 million in funding, the company aims to scale AI-driven weed control while reducing reliance on herbicides.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!