Run! Hide! The robots are coming.
Well they are, but not quite as remorselessly or as swiftly as the movies might have conditioned us to imagine. And when the robot age does arrive, the impact on New Zealand — in jobs and economic disruption — may not be as apocalyptic as some future scenarios imagine.
At least that is the position of two leading robotics researchers, Armin Werner of Lincoln Agritech, and Bruce MacDonald, an Auckland University computer engineering specialist with over 30 years' skin in the robot game.
Werner, whose background is in precision agriculture that will have its most advanced developments in robotics, believes the broad use of automation will create more jobs, at least at the skilled end of the labour market.
His view is that the embrace of technology will lead to different forms of work, rather than making work more difficult. Werner is a co-ordinator of the $2 million first stage of a long-term research programme that aims to show how New Zealand can make and use robots in small-scale manufacturing and dangerous outdoor activities such as forestry.
He says the projected 10-year National Science Challenge spearhead project is all about settling on the technology and science needed to create a workforce of small, adaptable robots in a decade or so and big, tough machines to do the outside grunt work all by themselves — learning from their mistakes and from what other robots have "experienced".
Werner imagines that if everything falls into place, despite New Zealand's size and distance from major markets, it is not inconceivable that this country could have a flourishing robotics industry in 10 years.
The goal of his group is that after eight-10 years, New Zealand developed robots will be available for a range of duties in the indoor manufacturing and primary sectors. Small-scale manufacturing with adaptable robots is a contrasting approach to mass-produced, single-task robots.
The robots will need to be smart and adaptable - one day operating on steep, uneven slopes planting or felling trees, and soon after that working on flat terrain, undertaking different actions.
The current work is focused on the science. He gives the example of a one tonne forestry robot, which will need new, sophisticated sensors for outdoor work. The robot will need to distinguish, for example, the presence of a child in its path, unlike a factory machine surrounded by cages to keep workers out of harm's way.
Werner says giving a robot the ability to detect the child and do its job is not simply a matter of joining together a few sensors, but a far more complex task. "So what kind of sensing technology and what new methods in interpreting the resulting sensor data do we need? What do we have to develop? That is the challenge we face."
In a few years, Werner wants to be in a position where overseas investors beat a path to New Zealand because it has created a reputation for its small scale robot technology and scientific knowledge. The team with the job of delivering that vision includes specialists in artificial intelligence, sensor technology, engineers and software experts.
MacDonald is similarly upbeat. He has two major research strands under way — one in the growing healthcare robotics field and the other in horticulture, where his team has worked with Tauranga-based RoboticsPlus on technology that is already transforming kiwifruit production.
The first orchard machines worked under canopies, picking and sorting fruit with robotic arms. The new generation use cameras to identify suitable flowers before spraying pollen. Complex sensors designed to pick up and interpret signals from their surroundings tell the robots when to carry out their tasks, from pollination, weeding, crop thinning, fruit carrying to horticultural quality checks.
Like Werner, MacDonald thinks the transition to the robot age does not need to become an era of mass joblessness and deep poverty, as some suggest. Calculating the likely casualties of the artificial intelligence revolution has led to some alarming projections.
A survey by a team from Oxford and Yale universities concluded that all "routine" jobs — drivers, labourers, office staff and salespeople — could be obsolete in the next decade. "Nonroutine" skills would be next for the scrapheap — surgeons, writers, construction workers, police officers and nurses — as more tasks become automated by the 2040s.
The survey, which covered 352 experts in artificial intelligence (AI), concluded that by 2060 machines would be capable of doing anything that humans do. What's more, the AI workforce doesn't stop for lunchbreaks, works nights and weekends and learns during the job to do it better. Other studies have reached similar conclusions.
The consulting firm PricewaterhouseCoopers reported that 38 per cent of all US jobs were at high risk of automation by the 2030s, and the World Economic Forum forecast that robots would displace 5 million first world workers by 2020 — less than two years' time.
MacDonald doesn't buy it, but equally, he sees a pressing need to get some policy markers laid down to manage the growing presence of robots in our lives.
On the disruption side, MacDonald believes the shakeout from AI will not be as rapid as some studies conclude, and offers the novelty of self-driving cars as an illustration for his viewpoint.
He says such cars might appeal to the 10 to 15 per cent of consumers who are early adopters of technology. But without a suitable regulatory environment, autonomous cars remain off limits to investors. And even when the vehicles are considered compatible with road rules, the early adopters might only be tempted when they buy their next car.
"Others will want to see how it goes," MacDonald suggests. "Even then, I think you are looking at 20 to 30 years regardless of whether the cars are available tomorrow."
Even so, MacDonald believes New Zealand needs to be much more up to speed with the sweeping changes that will flow from automation. He says labour force planning will be necessary to ensure workers have the right skills for the new world of work.
"Rather than servicing cars, we'll need people to service robots," MacDonald says. "As things stand, we're a long way from preparing communities for where we need to be."
And it's not just the workforce who will need to pick up their game on the AI front.
Callaghan Innovation argues that New Zealand firms need to be more aggressive in adopting robots and AI. The publicly-funded agency believes NZ firms are dragging the chain in the uptake of AI and robotics technology.
Callaghan's chief executive Victoria Crone says too many firms "have their heads in the sand" over the adoption of new digital technologies and risk being left behind by competitors in China and elsewhere in Asia.
A discussion paper on innovation, launched recently by Callaghan, reported that over the next two to eight years, high impact technologies would emerge in animal health monitoring, robotic surgery, personalised health, energy saving and autonomous cars, trucks and tractors.
Crone hopes the report will jolt firms into action, rather than pretending change isn't happening.
According to Callaghan's survey, agriculture will experience an "extreme impact", with automation replacing many tasks now done by human hands.
Lincoln Agritech's Werner says the primary sector is certain to see more robots. The technology, he says, holds the promise of reducing costs, increasing exports and filling the labour shortfall.
He says growers will often tell him that they find it hard to hire staff, and when they do, there is no guarantee they will all perform to the required standard.
Robots, on the other hand, do what is required of them. They work to the same exacting standards every time, which leads to better management of a crop, less waste and the promise of higher returns. Werner also points out that people will be needed to supervise and train the robots.
The Callaghan paper identifies a comparative advantage which New Zealand holds over other, larger economies. It says co-operation between corporates, innovators, academics and funders is possible in New Zealand, but hard to achieve in bigger markets.
MacDonald says part of his work in the personal health field rests on research relationships which he believes give New Zealand robotics an edge. He engages with a team which includes a psychologist and a gerontology nurse.
MacDonald has been involved with a project which teams Korean robot technology, Auckland science and Selwyn Retirement Village residents. The groundbreaking work is moving to create robots which can detect and respond to human emotions and moods.
Part of the challenge, says MacDonald, is devising ways for robots to work with humans - and for humans to work with robots. He notes that humans are used to relating to other humans, and with animals. "Relating to robots is a kind of new game."
At a basic level, robots can remind elderly people when to take their medication, monitor blood, weight and breathing, and send a message to family members when something appears to be amiss.
The next step is creating a relationship between patient and robot that goes beyond simple companionship. The robot will get to know their human friend, talk to them and react to their needs. MacDonald says it's not about replacing human contact, but hopefully increasing it.
Both Werner and MacDonald say there are barriers in New Zealand to the adoption of robotics, and believe these obstacles, including limited funding, restrain the economy from getting the full value of automation.
In the rural field, Werner says that unlike European farmers, New Zealand producers are hesitant to embrace smart technology which would deliver clear benefits. He says that while robots have an edge in any industry which employs staff for repetitive tasks, the technology is not available for all sectors.
But where it is, he wishes that growers were bolder and there were more companies willing to invest in developing robots because innovation would happen faster. He also wishes the industry had better networks to share information.
"We don't have enough farmers using new technology," Werner observes. "Farmers learn from farmers. New Zealand farmers are very selective. They want to be sure that what they pay for gives the best benefit or profit."
The benefit, he says, is not always immediately obvious so understanding what the technology can do is vital to the uptake of automation. When farmers see their neighbour using GPS or sensors and data to improve decision-making and profits, then they get convinced of the advantages.
MacDonald wants to see more done to prepare New Zealanders for a future with robots. He says we shouldn't fear them, or create impediments to their deployment - for example, by taxing robots.
That is an idea raised in a report to the European Parliament and endorsed by business titan Bill Gates. Behind the notion is the belief that the impact of AI and robots will be far deeper than policymakers concede. In other words, a job apocalypse is coming down the assembly line and one way to fund the transition to a robot working world is to make them pay tax. Rise of the Robots author Martin Ford says mass joblessness will be widely felt and won't be confined to low-skill workers.
MacDonald says New Zealand is not prepared for the new robot world and ought to be doing more. He has confidence in the ability of the workforce to change and adjust, but he sees a need to ensure that communities are properly equipped for the robot future.
He cautions against getting too carried away with Cassandra-like predictions. Says MacDonald: "If you're going to make a prediction, be careful when you make one about the future."
Robots already among us
Robots are already starting to plug in to the New Zealand economy, particularly in the rural sector. Robotic milking sheds allow dairy farmers to run operations remotely, with limited human input. Robots work on mussel farms, in packing sheds, on kiwifruit orchards and in meatworks.
They can also be found in a few retirement homes, as companions to elderly New Zealanders.
These clever machines could herald a leap in the development of New Zealand's land-based industries and help meet the Ministry for Primary Industries' aim of doubling the value of primary exports to $64 billion by 2025.
But what exactly is a robot? University of Auckland robotics professor Bruce MacDonald says true robots are autonomous, are programmable and usually, but not always, have arms that can work in many directions. He says this definition rules out cranes, which have arms but not the full flexibility of robots, but includes self-driving cars, which don't have arms but can move in many different directions.