Those monster machines are scary, both to normal farmers that cannot afford them, and to the millions of poor migrant labourers that pick the fruit today.
But after questioning 500 fruit experts, the prospects of many multi-million dollars fruit-picking robots seem uncertain: a device costing U$D 3 per tree can replace them. 46% of responders said the little device could replace the expensive robots. 38% were undecided, 12% favoured the natural, human hand labour and 4%, who happened to be working in software for the fruit-picking robots, said the robots were better.
Approaches to mechanical fruit collection
Lets take a look at the contenders.
On the expensive range of the field, Carnegie-Mellon Robotics Institute and Japanese researchers have been developing immense machines with a few dozens of eyes and thousands of robotic arms with hands and fingers. These robots locate the mature fruit, take them delicately with their fingers, spend a few seconds deciding whether the fruit is mature and deposit it in a basket.
An important part of these robots is the software that allows the robots to tell a mature from a still green apple. Or pear, or peach, because each different fruit poses a different challenge. Those expensive machines need to be very fast, because Nature does not give us much time to collect ripe fruit, and when you rent a monster like that you pay monstruous bills.
The most common aproach to automated fruit picking is the tree shaker. Shakers are very aggresive approach to fruit picking, because they apply a lot of energy to trunks to produce a good crop. That energy is expensive in terms of fuel and also damages the bark, trunk, branches and roots. Software is also an important part of these type of fruit picker. There is a software that evaluates each tree and decides how much energy (wave amplitude and frequency) needs to be applied for optimal collection and minimal damage.
The third approach is "combing", a system that also requires precision, large number of combing arms, sincronization, amplitude of movement and artificial intelligence to locate the combs near the branches and move them.
The fourth approach would be "beating", using robotic arms to impact the branches and dettach fruits. Some machines combine shaking and beating. Of course, it is the most damaging method and only applies to nuts and fruits intended for juice or other processing.
This video of a Colossus 26-ton olive picker robot is a sample of how a brute-force approach is applied to a seemingly simple task: http://www.youtube.com/watch?v=lxS7s3BzIUk&noredirect=1 . Similarly, this is a monstruous olive picker:
Simple is better.
The existence of some many expensive and complex approaches to fruit picking is a sign of no clear cut winner. For that reason, a most simple solution by a non-fruit scientist is winning adepts: The Bendable Tree Project.
This diagram shows the principle:
A wooden press forces the tree to develop a narrowing that allows easy shaking.
The method efficacy is now being confirmed by scientists at the University of Buenos Aires, School of Agronomy.
The information about the pressing device and a contact form is found at http://netic.com.ar/bendable-tree/ .