SARAL-Bot autonomously trims diseased leaves in strawberry fields
The SARAL-Bot was developed at the University of Nevada, Reno, and is specifically designed for soft fruit cultivation in tunnels or greenhouses.
The robot moves autonomously between raised beds, detects withered or diseased leaves, and selectively prunes them using a robotic arm. This allows it to take over repetitive manual trimming tasks.
The key benefits are labor savings and precision. SARAL-Bot can work 24 hours a day and takes time-consuming tasks off growers’ hands. Early detection and removal of diseased leaves helps prevent the spread of fungal infections and other plant health issues.
The robot also collects data via sensors, including leaf color, blooming stage, moisture stress, and plant development. This information can be used in crop registration systems or linked to site-specific crop management tools.
Autonomous mobility and targeted pruning
The robot drives on four aluminum mecanum wheels with rubber grippers, which allow it to turn in place and move accurately even on plastic mulch. Navigation is based on a RealSense 3D camera and RTAB-SLAM, an open-source mapping and localization system.
At the rear, a liftable camera scans the crop canopy. Once abnormal leaves are detected, the software controls a five-degree-of-freedom robotic arm with a gripper and small blade. The arm reaches up to 40 centimeters into the crop without disturbing healthy foliage.
SARAL-Bot runs entirely on open-source ROS 2 software and is powered by three separate battery units: one for drive motors, one for the robotic arm, and one for the cameras and processors.
High accuracy and minimal downtime
In test trials, the robot navigated reliably with a positioning accuracy of about 2 centimeters. Leaf detection accuracy exceeded 95%, even under variable light conditions.
The robot prunes one leaf approximately every 12 seconds, including movement and positioning time. One full battery charge allows about two hours of continuous operation. The design allows for a quick battery swap in just 30 seconds, minimizing downtime.
Preparing for real-world deployment
SARAL-Bot is intended for growers dealing with labor shortages during trimming season, who are looking for automation without having to redesign their entire cropping system. The prototype was built on a relatively low budget. In small-scale production, the developers estimate a unit cost of around €35,000 ($37,500).
The team is now seeking grower partners and industry collaborators for commercial pilot trials in standard 120 cm-wide strawberry beds.
