The fruit-picking robot that could replace humans
A robotics breakthrough with potential to revolutionise the fruit picking industry, seems likely to fuel the debate over the future of immigrants and contractual labourers working during harvest.
Cambridge Consultants is developing a bot that is able to pick and sort through irregular organic fruit and vegetables, and detect and remove weeds found among crops.
Chris Roberts, head of industrial robotics at the research company, explained that the challenge is achieving automation when dealing with uncertainty. The innovation involves a blend of existing technologies, as well as new signal processing techniques, to produce a “radical” new system, “poised to disrupt the industry.”
For a long time, scientists have been able to achieve stable robotic mechanisms for use in controlled environments. However, the struggle is working with changing conditions – in a fruit and vegetable warehouse or orchard for example. In this fruit-picking case, the robot has to be equipped to deal with different object types and sizes, calculate picking orders, gauge grip requirements, lift and place items.
Companies such as San Diego-based Vision Robotics and Belgian automation institute ACRO, have proposed robotic fruit harvesters to replace seasonal labour, but their designs have struggled to attract financial backing.
They have also faced objection from farmers’ unions who predict damage from mechanical interference. “There are already machines that will pick wine grapes, but the high end wine growers don’t use them, because they want the quality,” said United Farm Workers’ spokesman Marc Grossman, commenting on the Vision product in 2007.
Confident of the new innovation’s potential, Roberts explained: “Our […] industrial sensing and control team has combined high-powered image-processing algorithms with low-cost sensors and commodity hardware to allow ‘soft’ control of robots when the task is not rigidly defined.”
“The system is capable of handling objects for which no detailed computer-aided design (CAD) model exists – a necessary step to using a robot with natural objects which, although they share some characteristics, are not identical,” he continued.
A video demonstration of the technology [see below], shows the robot using machine vision and smart software to identify which piece of fruit is placed on top.
The system then plots this information into co-ordinates and positions its arm to pick the chosen fruit, while avoiding touching the other pieces. The ‘hand’ is able to adapt to the form of the fruit and securely grasp it without causing damage. The robot can also sort the fruit by colour, sorting red apples from the green for example.