Should we be using robots to pollinate plants?
Following up on my piece a few days ago about using robots in agriculture to reduce the cost and impact of pesticides and herbicides by applying them locally, here’s another example of how robotics is changing farming: using machines to pollinate plants, a response to the drastic fall in bee numbers around the world and Colony Collapse Disorder (CCD).
Farmers have sought solutions to the problem by renting bumblebee colonies from specialized companies, a procedure common in greenhouses for many years. Some robotic alternatives are reminiscent of Black Mirror, including drones, which Walmart is hoping to patent in a bid to control its supply chain, along with large robots with articulated arms that operate between rows of plants, shaking flowers they identify through computerized vision.
This video shows prototype robots that are still unable to recognize flowers through computer vision and instead use QR codes as they patrol a greenhouse.
Different plant species have their own pollination systems, which vary enormously, while there are any number of adaptation mechanisms that have led to co-evolution, along with highly specialized examples, such as some orchids. Blackberries and raspberries self-pollinate, so robots need only locate their flowers and gently vibrate them, which is sufficient for the pollen to detach from the stamens and reach the pistils. For other plants, the robots approach a flower, attracting the pollen using a gel with an electrostatic charge and then visit other flowers to deposit it, a more complex task that, among other things, requires mapping the flowers so as to maintain an order. At the same time, using robots makes it possible to eliminate flowers with malformations or that are likely to produce fruits with little commercial viability.
Given nature’s ability to come up with myriad ways to pollinate plants, the need to resort to the use of expensive robots merely highlights the mounting environmental problems we face. There are many species of insects specialized in the pollination of different species that have happily been at work for millions of years: the idea that a robot can do a better job is illusory. One thing is a controlled environment with smooth ground conditions such as a greenhouse, but pollinating an almond orchard is quite another. Developing robots to help is across as wide a range of activities as possible obviously makes sense, but in the case of agriculture, a better alternative is to fight against the disappearance of the organisms that carry out pollination naturally.
(En español, aquí)