New Scientist – Automation is the key to precision farming. Specifically, say agricultural engineers, precision farming needs robot farmers.
Soon, we might see fields with agribots that can identify individual seedlings and coax them along with tailored drops of fertiliser and measured sips of water. Other machines would distinguish weeds and dispatch them with a microdot of pesticide, a burst from a flame gun or a shot from a high-power laser. These machines will also be able to identify and harvest all kinds of ripe vegetables.
Robots could bring major changes, too, in jobs and how we work, in the soil and its quality, and in how much energy, and thus carbon, goes into farming. They could reduce pollution and water use. The most visible change, though, for ordinary people, could be in how farmland looks. Crops could be planted in small, geometrically arranged fields, while fruit farms are filled with arrays of two-dimensional trees. Robofarmers might even influence the type of fruit and vegetables that reach our shelves.
When fully autonomous systems take to the field, they will look nothing like tractors. With their enormous size and weight, today’s farm machines have significant downsides: they compact the soil, reducing porosity and killing beneficial life, meaning crops don’t grow so well. Compaction also increases erosion by rainwater run-off. “Why do we plough? Mainly to repair the damage that we have caused with big tractors,” says Blackmore. “Up to 80 per cent of the energy going into cultivation is there to repair this damage. Surely there is an opportunity to do things in different ways.”
Fleets of lightweight autonomous robots have the potential to solve this problem, Blackmore believes. Replacing brute force with precision is key, he says. “A seed only needs one cubic centimetre of soil to grow – if we cultivate just that we only put tiny amounts of energy in and the plants still grow nicely.”
These lightweight robots could remove the need for ploughing altogether, significantly reducing the amount of energy, and thus carbon dioxide emissions, coming from farming. And with less compaction, the soil keeps its structure and beneficial organisms, and is able to absorb more water and stay fertile for longer.
Agribots need to have three key abilities: to navigate, to interpret the scene in front of them, and to be able to help the farmer, by blasting a weed, applying a chemical or harvesting the crop.
Navigation – RTK-GPS
Navigation systems are the simplest part of the equation, particularly with the emergence of a high-precision satnav technique called RTK-GPS, which enables machines to locate themselves to within 2 centimetres. Arno Ruckelshausen from the University of Applied Sciences in Osnabrück, Germany, is developing this for a modular robot farmer called BoniRob. This four-wheeled field rover uses spectral imaging cameras to pick out green plants against brown soil. It then records the location of individual plants and repeatedly returns to each one during the season to monitor its growth.
Precision Weed Elimination and Fertilizing
Eliminating weeds is a particularly desirable aim, since they reduce yields in some crops by more than 50 per cent. So next, Ruckelshausen intends to fit this robot with a precision spraying system – based on an ink-jet printer – that can apply microdots of herbicide to the leaves of weeds. He calculates this could cut chemical use by up to 80 per cent. Even taking into account the initial investment in the robot, this would end up being cheaper than conventional weeding.
Similar savings are possible with fertiliser: field trials have shown that by using sensors to assess an individual wheat plant’s nitrogen levels, a robot can tailor the amount of fertiliser it gives and reduce the overall amount used by more than 80 per cent, with no loss in yield