These days, the headquarters of G’s Fresh — a large farming complex on the outskirts of Ely, East Cambridgeshire — is oddly quiet. In the office building, administrative staff work in open-plan spaces and walk through bluish-carpeted corridors, rubbing elbows with the occasional truck driver just in from Spain with a fresh load of broccoli or tomatoes. Internal windows show unloading bays dominated by towers of crates filled with supermarket-ready salad bags. The nearby residential building, which can accommodate hundreds of vegetable pickers during harvesting season, is almost empty. Only a few people helping with planting operations dawdle around between shifts. In a couple of months things will look very different. Hopefully.
All around is a green expanse of celery and lettuce fields. Right now, the fields are empty. But in June – G’s Joshua Pugh Ginn tells me as we walk through the building – they will be where all the action happens. There will be harvesting rigs – wheeled, tarpaulin-covered moving storehouses equipped with conveyor belts – trogging across the land, like oliphaunts constantly fed by teams of workers laying neatly cut vegetables on the conveyor belts. “People would be inside the rigs, under the tarpaulin roof, or walking in front, and on the sides,” Pugh Ginn says. If you were there on the field, you probably wouldn’t hear much English being spoken: the near totality of G’s seasonal workers, the ones taking care of the harvesting, are from Eastern European countries, with Romanians and Bulgarians making up the biggest cohorts. That is not an isolated case: according to the National Farmers' Union, less than one per cent of seasonal workers in British farms are UK-born – a drop in the ocean of workers from Romania, Bulgaria and the so-called EU8 countries.
But there is a problem, Pugh Ginn – a young Cambridge-educated history PhD who takes care of Brexit-related matters for G’s – tells me. Following the 2016 referendum, and the government’s decision that freedom of movement from the EU will stop at the end of 2020, Eastern European migrant workers have already started shunning the UK. For the harvest, G’s needs around 2,000 workers, 600 of them in Cambridgeshire alone. The company is currently grappling with a gnawing question: come June, will the fields crawl with people deftly cutting heads of iceberg and romaine, or will the company find itself in the unenviable position of not having enough pickers? More importantly, when the UK leaves the EU for good, how will farmers across the country have their fruits and vegetables picked?
Labour shortages in agriculture are not a recent problem – nor a specifically British one: farmers from Canada, to Australia, to California, have been complaining for years about recruitment pains. In wealthy countries, less and less locals are willing to toil in the field on seasonal contracts: G’s attempts to recruit seasonal workers from the UK’s most high-unemployment areas have proved routinely unsuccessful.
Problem is: as the global economy grew more prosperous, people in countries that were previously a reliable reservoir of seasonal workers are also becoming less keen on embarking on short-term fruit-picking expeditions to richer countries. “The problem with labour started before the referendum,” says Sharon Cross, who, as G’s ethical working director, is in charge of recruiting seasonal workers. “The labour supply was getting shorter, as unemployment levels in Eastern Europe, and in the UK, had been going down and down and down.” That was particularly marked in Bulgaria – where unemployment levels have dropped from 12.3 per cent in 2012 to 6.2 per cent today. In the same period, joblessness in Romania dropped from seven per cent to five per cent. “We saw this coming,” says Alison Capper, the chairman of National Farmers' Union’s (NFU) horticulture and potatoes board, and a grower of apples and hop in Stocks Farm, Worcestershire. “But Brexit has exacerbated it.”
On June 24, 2016, when it had become clear that the country had voted to leave the EU, Cross walked to G’s hostel – a 400-people residential structure right next to the firm’s main office – pondering the situation. “I was thinking: ‘Flip, what does this mean? What on Earth does it mean?’” It was harvesting season: workers coming back after eight hours of veggie-picking still had the energy to pump iron in the structure's gym, or to play ping-pong in the common room, or tennis and basketball on nearby pitches. Others were cooking their meals in a massive brushed aluminium kitchen (G’s used to have a canteen, but it was shut down as it struggled to cater to the workers’ variegated gastronomical tastes.) Others went to rest in their six-bed rooms. Initially, Cross says, nothing happened: workers had not yet zeroed in on Brexit’s possible repercussions on their lives. “After a couple of days, with conversations, and stuff in the press, more and more people came forward with more and more questions about what was going to happen,” Cross says. “And that was really difficult because nobody knew.”
The first problem was the pound’s freefall compared to the Euro, which took a toll on workers’ savings practically overnight. “Some guys here said that from when they arrived in May to when they left in October the value of their earnings had dropped by a third because of the exchange rate going down,” Cross says. (In 2017, G’s raised wages from £7.50 to £8 per hour.) Some decamped: according to NFU’s statistics, 42.9 per cent of the seasonal workers left earlier in July 2016, compared to 14 per cent the previous month. Seasonal workers usually go fruit-picking for two to three summers in a row, putting aside the money and coming back the following year, until they reach a certain goal – they might want to earn enough money to buy a house, or fund their university studies back in Romania. But amidst a weakening pound, perceived anti-immigrant hostility, and a lot of uncertainty regarding whether they will be able to come back working in the UK over the next years, some of them resolved to build relationships with growers in places that are not in the throes of hazy divorce with the world’s largest trading bloc. Put simply, people now prefer veggie-picking in Germany.
Since the referendum, G’s has needed to work thrice as hard – with an intense campaign of recruitment drives in Eastern Europe’s high-unemployment areas – to get a lower number of seasonal workers than it did pre-2016. Returnee workers, which in 2015 made up 72 per cent of the company’s seasonal workforce, accounted for 42 per cent of the whole in 2017. In 2015, Cross says, there was a waiting list of 750 people who wanted to come to work at G’s over summer; in 2017, there was no waiting list. Even if Cross thinks they might have managed to recruit enough workers for this year’s harvest, she will have to wait until May to see whether that is true.
“Last year we had a higher no-show rate than we'd ever had before,” she says. “We expect the same impact this year but we're preparing for it and we got a little bit of extra people.” But as hard, no-free-movement Brexit gets closer every day, Cross is bracing for reckoning. “We will start our recruitment for 2019 in July. I am hoping that between now and then we'll have some clarity about what's gonna happen in terms of hiring foreign seasonal workers,” she says. “We can't say ‘Everything will be okay, you'll be able to come back’, because we don't know.”
Many in the farming sector pine for the good old days of Seasonal Agricultural Workers Scheme (SAWS) – a system that, from 1948 to 2013, allowed growers to recruit migrant seasonal workers for up to six months. It was binned in 2013, on the grounds that UK growers could simply access EU labour. Stocks Farm’s Alison Capper says that, with the UK’s farming industry needing some 80,000 seasonal workers every year, a SAWS-like scheme would be of crucial importance to stave off a Brexit-triggered crisis. Otherwise, she says, crops will rot unpicked in the fields.
And it’s already happening: last year, she had to send 100 bins (or about 35,000 kg) of eating apples to juice, because they were too ripe. That cost her £30,000. ”I know of several growers – of blueberries, strawberries – whose crops are going to waste because of lack of labour,” she says. “They don’t want to say it publicly, to stick the head above the parapet, as they fear that would spoil their reputation.”
So far, the government has not come out in favour of any new migrant-labour schemes. Although, according to Pugh Ginn, ministers have been quite understanding in private conversations, public declarations tend to not even acknowledge that the UK has a labour shortage problem. In a document published by parliament’s select committee for Environment, Food and Rural Affairs, in early 2017, government ministers were quoted saying that there were only “anecdotal stories” supporting claims of a labour shortage issue. “Certainly, to date, there is no suggestion that there is a problem,” the then immigration minister Robert Goodwill said.
More or less in the same period, leading Brexiteer and then secretary of state for Environment, Food and Rural Affairs Andrea Leadsom suggested that a possible solution might come from technology: in the future, berries would be picked by robots. The government, and the Department of Environment, Food and Rural Affairs (Defra), now under the stewardship of Michael Gove, seems to be quite into the idea: the government recently earmarked £90 million for a ‘Transforming Food Production’ programme, aimed at harnessing AI, robotics and satellite data to innovate the agri-food industry; Defra itself has announced a £40m grant for farmer investing in new technologies, such as robotics. “The government has commissioned advice from the Migration Advisory Committee to better understand the reliance on EU migrant workers across the economy and we will be working closely with our food and farming industry to consider their needs,” a Defra spokesperson wrote in an emailed statement. “For the longer term, Defra is working with industry to explore the potential for innovation and automation in contributing towards meeting future labour needs.”
Whether turning to robots is or is not a viable fix depends on whom you ask. G’s Pugh Ginn says that operations such as packaging and washing have been effectively automated over the last few years, whereas the actual harvesting – the hunkering, the judging the right size and ripeness, the clean cutting or delicate picking – is still the bailiwick of human pickers.
Alison Capper, who has also turned to automation to speed up some aspects of apple-picking, is similarly sceptical. “Picking soft crops is so difficult to automate, and there isn’t even a field-trialled robot ready for deployment yet,” she says. “I think the timescale is five to ten years.”
That is a view shared by most farmers who deal in soft vegetables and fruits – such as lettuce, asparagus, apples, berries, broccoli; cereals are hardier, and much more easy to reap automatically. They either say there is no working technology, or fret about the potential costs of buying a platoon of automata.
Researchers are slightly more optimistic. A small number of projects in the UK and elsewhere have built prototypes able to pick various crops – from strawberries, to cauliflowers, to apples, to broccoli – with some accuracy. Among roboticists, the general view seems to be that the technological pieces of the robo-farmer jigsaw are all there. Computer vision, to spot crops and decide whether they are ready to be harvested; 3D sensors, to create digital images of a robot’s surrounding; AI, to give the machine autonomy; and advanced robotics, to avoid that the robot’s arm squish the fruits by exerting too much pressure. It is all about combining them all in an effective way – and that is no easy feat. That the fate of British farming might hinge on scaling up a bunch of prototypes to a feasible workforce is exciting or terrifying, depending on your penchant for futuristic ordeals.
Martin Stoelen, who teaches robotics at the University of Plymouth, and has developed robots for picking raspberries and cauliflowers, is confident picking robots will hit the market soon. Although his raspberry-plucking robot only picks one berry every 10-15 seconds – much slower than a human – Stoelen is confident that speed will improve as the machine gets more training.
“It'll get faster. And the key thing when we look at comparing performances, is that this systems may be able to operate for much longer maybe 20-22 hours a day,” he says. “If they work at night it could be even easier because you can control the lighting. And each robot might have four to eight arms, so you can multiply the yield.”
Cost, in Stoelen’s opinion, is not an issue: even if each robot would cost “tens of thousands of pounds”, he thinks that the most convenient mode of use would be robots as a service – with farmers outsourcing harvesting to robot-owning companies during the picking season. As of today, of course, no such service exists. “I am pretty confident that we'll get systems that are commercially and economically viable within a couple of years” he says.
Not everybody agrees on that. Duncan Robertson, the founder of strawberry-picking robotics firm Dogtooth believes that jump-starting mass robot production is nearly impossible. “We would need to manufacture many tens of thousands of robots to meet the demand for picking,” he says. “It's unlikely we're gonna be able to do that in time for the next few years in which the effects of Brexit are gonna be felt.”
Timescale is not the only issue about which there is no consensus. Robertson, for instance, thinks that robots must be designed to operate on our oft-messy, cramped farms, not the other way around. Simply put: make the robots work in our world, don’t change farms to accommodate. Stoelen, on the contrary, sees this as an opportunity to rethink farming as a whole. It could be about minor tweaks – placing a dark background behind strawberry plants, so that berries are easier to spot for computer vision algorithms; but it could also lead to more radical moves. “When we talk to farmers about picking vegetables, they want to keep the big tractors and instead of having people picking the vegetables, they just want robot arms,” he says. “That might still give them a benefit, but once you take the person out of the machine if might make sense to just scale that down and have multiple systems moving around in the field and spread themselves out.”
How heavy you want to go on robots also says something about what kind of farming sector you think one should be promoting. Kate Pressland, programme lead at non-profit farmer networks Innovative Farmers, fears that ditching migrant labour in favour of automation could end up penalising smaller farms. “Smaller producers don't necessarily have the investment levels for robotics – yet there is some evidence that smaller farms have a higher productivity. They could potentially lose out even if they are run very well as farms,” she says. “Technology is being given the strongest focus, but that way you could end up with fewer, massive farms.”
What everyone seems to agree on, though, is that robots will not be able to replace every single fruit-picker. Even in the best-case scenario — Robertson and Stoelen concur — robots could “alleviate” the labour shortage crisis to come, by enhancing the capabilities of human workers. And according to Robertson, to some extent the question is whether throwing technology at the problem is a solution at all. “People provide a lot of benefits – robots are very bad at lateral thinking, at communicating. The interface growers have now is the one they want: they want to speak with their workers, they want to talk face-to-face with them,” he says. “I am not sure that doing without people is the right goal.”
That is certainly not the goal that growers across Britain — their crops tumefying in the fields — are trying to reach; but it is the goal the government seems to have set for them, in order to stick to its immigration red lines. Either that will change, or farmers will need for a robot-revolution to happen before the harvest of 2021.
Updated 19.04.18, 09:35: Martin Stoelen’s robot picks raspberries, not strawberries.
This article was originally published by WIRED UK
