Mann Packing

Broccoli Robot

s the industry fuses its current time-tested operations around the upcoming technological advances of tomorrow, the automated digits of a robot may soon be extending in tandem with human hands in broccoli fields across the globe. 

Researchers at the University of Lincoln in England are looking to extend the capabilities of field harvesting even further by using 3D technology to catapult a fully-automated robotic harvesting system into fields of perfectly plucked florets of broccoli.

The technological improvements are rooted in the 3D cameras, which the entire prospective robotic system will revolve around. The cameras use identification systems to properly pinpoint where broccoli heads are located in a field of foliage, using technology similar to that seen in motion-sensing video game consoles; specifically Microsoft’s Kinect console, which first made its lasting splash into the electronics market in 2010.

The cameras use 3D algorithms through its eyes to focus on one location, fixating fully on a head of broccoli in order to relay to the robot where to pick a bundle of florets ready to harvest. 

Temporal filtering combines each dimension of detection to form multiple frames, localizing each individual broccoli in the field. To obtain these in-field and in-soil images, the cameras are pulled along by human counterparts on a tractor; moving through the rows of greens to collect field data information.

Researchers stated that although the camera’s depth plane was the hardest to tune, all respective levels of “x, y, and z” are fully functional. 

With its camera now focused, the project plans to move onto its next step of research: cutting the identified heads through an established partnership with Lincolnshire, England-based horticultural consultants R. Fountain & Son Ltd., who will be responsible for creating the broccoli-slicing device. 

This system of identification boasts an accuracy rate of up to 95 percent, marking a significant step towards reducing the production costs associated with time-sensitive harvesting.

The ongoing project, led by Professor Tom Duckett and Dr. Grzegorz Cielniak of Lincoln’s School of Computer Science and Professor Simon Pearson from the University’s National Centre for Food Manufacturing (NCFM) at Holbeach, could see the development of a commercial system within the next few years, with plans to extend the technology even further than broccoli to other categories within the varied fields of fresh produce.

Will harvesters soon have a Wall–E type companion reaching out its arms to help them harvest crops? As the industry looks to streamline production costs towards greater efficiencies, these arms could be just the ones to provide a nice balance of costly reprieve and the techy tools to do so. 

All photo credit | University of Lincoln, England