More human than human?

In the 1982 cult film Blade Runner, technology mastermind Eldon Tyrell describes his latest robot creations as “more human than human”. The movie’s dystopian future depicts human-like “replicants” used as free labourers. Back then, the idea seemed unimaginable, but in agriculture Tyrell may have called it sooner than anybody expected.

Fast-forward to 2017 England, and a young collaborative dubbed “Hands Free Hectare” told the press they had grown and harvested wheat without any humans in the field by using an array of robots. In Japan a lack of farmers coupled with a shrinking workforce necessitated robotic advances, with lettuces now being grown for fully automated indoor harvesting.

Such examples remain rare, but they demonstrate the huge progress artificial intelligence and robotics have made in recent years. These are exciting, if ambiguous times in robotics. On the one hand breakthroughs in technology, particularly camera-based decision making, have propelled the discipline forward, but on the other, it’s unclear how soon they will become commercially viable to food suppliers who face increasing pressures.

Well-documented seasonal worker shortages in Britain and Europe, and rising wage costs have emphasised the need for much leaner production capabilities. According to the latest National Farmers’ Union survey, 60 per cent of producers did not secure the number of workers they needed in 2017. It’s therefore unsurprising that an AHDB report on robotics released in October stated 84 per cent of growers said they were planning to invest in robotics.

James Simpson, managing director of topfruit suppliers Adrian Scripps, says he has invested millions in automation and robotics as a result of pressures from labour shortages and wage increases. “We are adding case-filling and pallet-stacking robotics to our packing lines to complement the robotic loading of packing lines and automatic pack filling we have installed over the past three to four years,” he said.

“We are also looking at robotic harvesting of apples, but they are realistically several years away from commercially viable equipment. The intention is to answer the pressures placed on the business by labour shortage and wages increases.”

Debbie Wilson, head of AHDB horticulture knowledge exchange, concurs that from seeding to the packhouse, robotics and automation will likely take over in more uniform production processes.

“There are gaps to fill in both of these areas where routine manual tasks could be automated within our lifetimes assuming the investments stack up,” Wilson says.

“Recent changes in labour availability and cost are both influencing the question of economic payback. Harvesting is another labour-intensive area and has systems already in development. It seems realistic to expect this technology to become mainstream for at least the bigger crops or more uniform production sites within our lifetimes.”

It’s no surprise that agriculture has emerged as one of the key drivers of robotic research. According to industry analysts Technavio, the global agri-robotic sector will increase in value to $28 billion by 2025. In Europe, they predict the autonomous tractor market will expand by 50 per cent in the next three years.

“Agriculture really appealed as an industry because we could see there was a technological gap at the time,” says Ed Herbert, founder of UK company Dogtooth Technologies. “Anything you could do to drive efficiency would create profits for everyone involved.” His team is developing a robot that will be able to harvest strawberries using a patented technique that picks the berries by the stalk. At the other end of the supply chain, Paul Wilkinson, business development manager at Pacepacker Services, says he has noticed an uptake in interest in their services, stating a “good chunk” of his business is now in agriculture.

Niall Mottram, head of product development, industrial and energy at Cambridge Consultants, echoes similar anecdotal evidence: “I have been here two and a half years. When I came on board we had one [agricultural] client and now we have six to seven.”

He adds that recent political developments like Brexit are not necessarily key drivers of this trend, saying that a mixture of commercial, technological and political factors are combining to push developments in robotics.

“This has really only come on in the last 24 months or so. Irrespective of the political climate, putting aside labour shortages etc, the thing that generally drives technology is when the numbers add up.”

Likewise, Herbert insists however that the narrative about replacing labour is a “red herring”. He says that labour shortages have long been a problem, and robotics would be more likely filling a gap, rather than replacing workers. Agricultural jobs, he says, would shift instead to higher skilled roles that are managerial and information based, rather than physical or manual.

Instead, Herbert believes the real revolution will be in the extensive data-gathering potential that robots offer. “To me that’s the key reason for doing this, the data is the thing that really adds value and will be transformative.” Other robots specifically designed for data gathering, such as the Spanish-built “Hamster Ball”, flying drones and autonomous tractors, will soon help create ultra-smart farms with large data sets pushing farming innovation.

Mottram agrees: “A lettuce farm, for example, may have around a million lettuces. No-one is going to look at all those, but you could have an automated drone that could capture information about every single lettuce. You can then decide ‘here are the ones we harvest or water etc’. They have the potential to make greater impact in the near term.

“Since the dawn of man, farming has been an art,” he concludes. “Now it’s turning into a science, and it’s the larger amount of high-quality data that’s turning it into a science.”

So are we on the cusp of a fully automated robotic revolution in horticulture? “I think the desire is there. I don’t think any grower will go for a big bang of switching all their picking labour from people one year to robots the next,” says Herbert. “That’s the normal story with the introduction of technology, these predictions of overnight industry changes, that’s the science fiction bit, not the technology itself.”

Simpson adds: “An 80-90 per cent automated topfruit farm is a real possibility; parts of the puzzle are readily available, others are a few years away. Fruit picking is coming closer, but is realistically five to ten years away before commercially viable machinery will be in orchards.”

But even when robots do become commercially feasible, it may ironically be their “minds” that prove more valuable over their labour, generating huge data sets that could spark the next transformation in farming. Then perhaps they will earn the accolade of being “more human than human”.

Data

Dubbed the “Hamster Ball”, after being developed by Universidad Politécnica de Madrid, the Rosphere rolls over farmland like tumbleweed, gathering data on soil and crops using a multi-rotor motor inside to propel it over different terrains. While not yet autonomous, its adaptable design could make it popular as data becomes increasingly vital.