Robots must be affordable to be useful

“Food prices will not spike so much that you can afford to have a $1 million machine picking five bins of kiwifruit a week,” he says.

The innovation centre is looking at service models, where the robots are sent in on contract, costing the same as labour with fewer hassles.

Companies at the innovation centre include Robotics Plus, which has won a $7.7m MBIE research grant for its work on robotic fruit harvesting and sensing technologies.

Saunders told the NZ Horticultural Conference that he is not a ‘techo’; he is just a kiwifruit farmer with a vision for where we need to go in horticulture.

“Horticulture is facing a lot of challenges,” he says. “Comparing horticulture to the broad-acre food crops in the US – they are highly mechanised, huge acreage, machine driven, GPS guided and genetically modified. [The ethos is] let’s mow it down and get it into the food system as fast as possible.

“But there is change happening in the fruit and vegetable world. Grains and those sort of things are low-nutrient, as opposed to fruit and vegetables which have dense nutrient make-up. So there is a swing in favour of fruit and vegetables; and to scale up to meet our audacious goal of $10 billion in exports by 2020… we need more labour and to be more efficient with what we have.”

Saunders says labour is a big problem for the sector – whether it’s lack of availability or hygiene.

“We’ve got the scalability issue: as we grow we have more reliance on labour; where are we going to get that labour? For us robotics plays a huge role in that formation, into the future.”

It is no use building a robotic machine that nobody can afford, so they are looking at service modelling.

“How do you get people to adapt to technology? The easiest way is to say, ‘you don’t have to pay [to buy it]; we will do the task for you: robots arrive, job happens, you pay. Our goal is you pay what you would now for labour; you just get rid of the labour hardship’.”

He says they start with an end user-focused outcome.

“This is what we pay to pick a bin of apples or kiwifruit; how can you build a machine which costs no more than at present to pick kiwifruit? To build an affordable machine we have to see whether there is a good return on investment for building it.”

They also look at consistency: in a harvesting situation people can be erratic in their actions.

“Just prior to Psa we tested our robotic kiwifruit harvester with a small fund Zespri gave us to look at man versus robot.

“We did every part, comparing man versus robot picking kiwifruit. We tracked that through cool storage and through all the damaged fruit. The robot was more consistent and the fruit quality better all the way through the system. It was constant: there was no reaction similar to how people react when they are harvesting fruit.”

Saunders says technology is happening very fast and technology costs are decreasing.

“Technology that can be adapted to robotic systems is now incredible compared to four-five years ago. The speed of new technology enables fast adaption.”

He had recently visited an operation Salina, California, where they harvest 101ha of lettuce a day. Their big concern is labour. They were “freaking out” at the prospect of a US$15 minimum wage.

“There are a lot of areas within that value chain that can be improved with the use of sensors, technologies and robotics,” Saunders says.

“We are looking at robotic harvesting, wireless orchard sensors, the use of drones, pollination systems and then moving into the post-harvesting to get more simplified systems.”

With drones they can do kiwifruit orchard flyovers to look and report on factors such as budding, canes or pruning.

They have also been doing trials with robotic apple packers with Compass Packhouse, Nelson, as early adopters. They can pack apples more quickly and efficiently than people standing on the lanes.

“At Compass they now have a machine that packs 120 apples a minute; it orientates the apples; it can outperform people and fits onto a standard sorter.”

He says a number of robotics projects are underway with his company, including a robotic kiwifruit harvester. The base technology was developed several years ago, but the whole project got stalled because of the Psa outbreak. It can pick 24 hours a day.

The innovation centre also has the universities of Auckland and Waikato and Plant and Food Research involved in a four-year, contestable science project with $10.5m; of this, $7.6m is from the government.

Saunders says they are looking at automated platforms for a variety of tasks, such robotic picking and spraying biologicals direct to a target.