The Vegebot can identify iceberg lettuce heads that are ready to pick and make a precise cut without damaging the crop. Photo by University of Cambridge
July 8 (UPI) -- Scientists have finally automated the job of picking iceberg lettuce. As detailed in a newly published study, the vegetable-picking robot designed by engineers at the University of Cambridge proved itself a capable hand in the field.
The picking of dozens of fruit and vegetable crops has been automated, but engineers have struggled to design a robot and the necessary algorithms to effectively harvest iceberg lettuce.
The so-called Vegebot has previously picked iceberg lettuce in the lab, but the latest test results -- published this week in the Journal of Field Robotics -- mark the first time the robot has successfully picked iceberg lettuce under field conditions.
Though Vegebot isn't yet as efficient as a human, the breakthrough marks an important step in the development of sophisticated agricultural robotics.
Some crops like potatoes and wheat are relatively easy for robots to harvest. But lettuce presents unique problems for robots. Lettuce heads are relatively flat and are low to the ground. Lettuce is also easily damaged.
Automated machines plant lettuce seeds and tend to their needs as they grow, but harvesting still requires manual labor. Picking lettuce is difficult work, but with a few tweaks, it could soon be the work of a robot.
"Every field is different, every lettuce is different," Cambridge engineer Simon Birrell said in a news release. "But if we can make a robotic harvester work with iceberg lettuce, we could also make it work with many other crops."
Harvesting lettuce presents two main challenges to a robot: selecting and cutting. Vegebot relies on a computerized vision system to identify the target crop and determine whether a head of lettuce is ready to be harvested. Vegebot's cutting system allows the robot to slice the head from the roots without damaging the lettuce.
A machine learning algorithm perfected in the lab helps Vegebot determine what a ready-to-cut head of iceberg lettuce looks like. One camera, coupled with its machine learning software, helps Vegebot find ready-to-pick lettuce, while a second camera guide's the robots blade, ensuring a clean cut. A gripping arm holds the head with just the right amount of pressure, so as to allow an efficient slice without bruising or crushing the lettuce.
Researchers suggest the Vegebot can adjust its software and gripping-cutting technique to harvest other types of lettuce and above-ground crops.
Vegebot could help reduce food waste by ensuring only ripe lettuce heads are picked. The robot could also help farmers improve their farming strategies.
"We're also collecting lots of data about lettuce, which could be used to improve efficiency, such as which fields have the highest yields," said lead researcher Josie Hughes. "We've still got to speed our Vegebot up to the point where it could compete with a human, but we think robots have lots of potential in agri-tech."