Is the future of agriculture autonomous?

For as long as we have grown crops, farming has required farmers. But a new breed of technology is taking some of the human factor out of agriculture. These advances come as the world requires ever more food to feed a growing population – some estimates say we’ll need to grow 70% more by 2050 – but also as there is a clear need to cut greenhouse gas emissions from agriculture, as well as the use of other inputs such as pesticides, herbicides and, most crucially, water.

With improvements in artificial intelligence, machine learning and computing power, a growing number of tasks on the farm are being automated, according to Dr Duncan Robertson, CEO of Dogtooth Technologies, a company that makes fruit-picking robots. 

And Rob Edwards, marketing manager for Kubota UK’s tractor business unit, says that autonomous agriculture, as it is known, “can be as basic as automatic steering guidance, and as complex as robotic machines that use AI to make agronomic decisions based on weather patterns, soil conditions and plant recognition”.

There’s clearly a lot of potential. The World Economic Forum predicts that autonomous agriculture will be a $95bn market by 2027.

WEF says that agricultural drones, self-driving tractors and seed-planting robots are among the innovations that could be key to future food supplies, as autonomous farming promises to produce more crops with less effort and reduced impact on the environment. 

Driverless tractors, controlled via an app on a phone or computer, can free up farmers from spending hours in the field. Drones and autonomous farm machinery can also be used to monitor factors such as crop condition, soil moisture and fertility, as well as carrying out precision, targeted pesticide spraying. Sensors placed in the soil can tell farmers where they need to use fertiliser most and to irrigate and spray only the crops that need it. 

Robertson says that Dogtooth’s robots are able not only to pick delicate berry fruits without damaging them, they can also determine when they are ripe, inspect them for defects and put them in punnets according to supermarket sorting criteria.

Growers around the world face a shortage of farm workers and the most commercially important applications are likely to be in areas where farms depend on a large number of seasonal agricultural workers, such as fruit harvesting. 

While there will inevitably be some concerns around changing agricultural sector labour requirements, recruitment in the sector has been challenging for some time. By reducing the need for the particularly difficult-to-recruit seasonal labour, autonomous technology cuts staff turnover and costs. 

Autonomous agriculture can also help in the many markets, such as Japan, where the farming population is ageing and retiring, with fewer people taking their place. 

Because autonomous farming increases efficiency and crop yields, as well as reducing labour costs, it makes farms more sustainable, Robertson says. Robots are pre-programmed and use precision technology, therefore they produce more consistent and accurate results. They further improve productivity because they can work longer hours, in all weathers and at night, as well as performing multiple tasks at the same time, such as planting seeds and monitoring soil conditions.

Their ability to collect data in real time allows farmers to react to changing conditions, enabling them to reduce waste and optimise yields. Norway’s Yara says that its water sensors can cut water use on farms by up to 20%, for example.

Using AI “swarm” technology it is possible to deploy a higher number of smaller, lighter vehicles instead of one enormous tractor, reducing soil compaction. The sector is also seeing the spread of “farming-as-a-service” models. Chicago-based Sabanto hires out fleets of small, autonomous tractors that allow farmers to experiment with the new technology without the expensive capital outlay for new machinery, for example. This is especially suited to smaller farms, allowing them to take advantage of technology that would otherwise be out of their reach.

There are inevitably a number of barriers to autonomous technology implementation at scale, including the high upfront cost and the need to retrain farmers to use the new technology. 

But, echoing the WEF predictions, Prof James Lowenberg-DeBoer, chair of agri-tech economics at Harper Adams University, says that robotics and automation are going to be part of creating food security around the world, including the developing world where farm labour shortages are a problem just as everywhere worldwide. 

As he concludes: “Low-cost robotics and automation could help farmers in the developing world intensify production without using more pesticide and fertilisers.” The technology thus has the potential to make a real difference in the markets where a revolution in productivity is perhaps most acutely necessary.