Long admired for its rolling hills and colourful sprawling meadows, Britain’s countryside is under threat. Though on the surface, it may seem that there is still a wide variety of flower and animal species, their numbers are rapidly declining. The cause for this is the intensification of farming over the last several decades, which has seen increased and indiscriminate use of pesticides aiming to raise productivity as the global population surges. However, the trend is not yet irreversible. To halt the decline in biodiversity and to improve farming efficiency while minimising adverse environmental effects, innovative measures are being taken that will revolutionise the way large-scale farming is done – thanks to the rapid advancement in Artificial Intelligence (AI).
While the world has zeroed in on the advancements of AI in sectors like finance and manufacturing, AI also holds immense potential for oft-overlooked sectors like agriculture and farming. If one of the fundamental questions of our time is how to increase food and land productivity without causing further damage to the environment, AI might just be the answer.
Given that the agriculture sector is such a critical industry, the stakes are high. In the UK, the value of food, feed and drink exports continues to increase year over year. The U.S. Environmental Protection Agency (EPA) estimates agriculture to add some $330 billion of annual revenue to the economy. And in poorer countries, increasing agricultural productivity is a critical element in meeting Sustainable Development Goals of ending global hunger by 2030, with organisations working at all levels to boost farming output where it is needed most.
No surprise, then, that agricultural players were fast in embracing the new technological opportunities that come with AI. The Food and Agriculture Organization of the UN (FAO), for instance, has placed AI for productivity increases high on its agenda and is raising awareness for its use. Agricultural robotics – where AI undertakes essential tasks like harvesting, soil and crop monitoring using deep-learning algorithms – as well as predictive analytics are among the most pioneering applications.
However, AI’s added benefit is that it also offers a means to evolve old techniques. Take the issue of pesticides and herbicides. Pesticides have been used for millennia but have become controversial in recent years for their potential negative effects on the environment. Some of the current practices involve blanketing entire fields in a bid to target pests and weeds, also affecting not explicitly targeted animal or plant species in the process. However, because pesticides are indispensable for modern agriculture to function, banning them is hardly an option. Instead, reducing the amount that is sprayed through precision delivery systems is the most widely considered solution.
Enter AI technology: with new applications, precision-targeting of micro-dosages can reduce herbicide use by up to 75 percent, according to researchers from Denmark’s Aarhus University. By mounting cameras on the back of vehicles that patrol crop fields, sprays are only activated once individual weed plants have been identified. Another future option is the development of robots patrolling fields and plantations autonomously, delivering tiny amounts to individual plants that need them, and thereby reducing unwanted effects on the surroundings to a minimum. While the development of such robots would require much higher rates of investments than current levels, some farming experts believe the robots could become commercially available in as few as three years.
From that it is clear that a lot is to be achieved in merging AI with agriculture, and considering the game-changer this technology promises to be, whoever succeeds in making precision-targeting systems available at a large-scale for the first will make gigantic profits. Unsurprisingly, the race is already on as major global companies are working overtime to get their foot in the door of the AI revolution. Amid all this, the AI market in the agriculture industry is forecast to grow by 23% through 2021.
As it shifts more and more to collecting and analysing data to feed AI software, the industry is rapidly consolidating. For example, the Climate Corporation, purchased by Monsanto in 2013, uses data collected from 2.5 million locations to simulate weather patterns and their effect on local climate and soil – factors that determine how much and what kind of herbicide or pesticide is needed on a given day. Using this information, farmers can adapt their pesticide use responsibly and even raise their harvest yields.
Furthermore, farm equipment giant John Deere acquired computer-vision specialist Blue River Technology late last year, adding to the firm’s catalogue of data-driven, high-tech services that are becoming ever-important for players in the agricultural industry. The acquisition opened the door for the company to equip farming machines or robots with advanced machine learning algorithms that allow analysis of visual data to make on-the-spot decisions on whether herbicides should be delivered or not.
These developments signify an evolutionary step in agriculture. AI promises to turn data into an environmental dividend, where pesticides will be used more precisely and selectively than ever before. The result would be an unprecedented return of flora and fauna to their traditional habitats in the UK and elsewhere in Europe. If biodiversity is to be saved, new AI technologies are the way forward.