search
Promoted Content

Fresh fruit picked from the tree by Israeli robots

10 percent of fruit worldwide rots on the trees. Tevel Aerobotics combines aeronautics, robotics and artificial intelligence to pick it

Tevel has developed an autonomous driving platform with several tethered robots that fly up and pluck fruit from the trees (Tevel)
Tevel has developed an autonomous driving platform with several tethered robots that fly up and pluck fruit from the trees (Tevel)

The harvest is a stressful time for fruit farmers. After planting, watering and weeding their fields, they need to quickly hire additional workers and coordinate the picking and packing of their crops within a few weeks before the oranges, apples or berries begin to decline in quality.

“This is hard, seasonal work,” says one apple grower in northern Israel, who, like many farmers, also needs to arrange housing, insurance, transportation and work visas for seasonal harvest workers. “Costs are rising all the time.”

Industry experts estimate that about 10 percent of the world’s fruit crops rots on the trees and goes to waste because there are not enough workers to pick it.

So this apple farmer is testing out a robotic fruit-picking system from Tevel Aerobotics Technologies, a local startup founded by veterans of Israel’s aerospace and electronics industries. Tevel has developed an autonomous driving platform with several tethered robots that fly up and pluck fruit from the trees. Instant artificial intelligence-based analysis of video of the trees allows the robots to pick only the fruit that is ripe. As the robots work, the system constantly updates farmers through a mobile phone app on how many pounds of fruit have been picked, and how much time it will take to finish the harvesting job.

“It solves the farm labor shortages,” said Yaniv Maor, founder and CEO of Tevel, based in Gedera in central Israel.

It has other advantages. The flying robots are more accurate and work longer hours than people. They can also carry out other tasks, like thinning and pruning trees, bringing down the cost of fruit production by about 30 percent.

“But the main advantage is simply that it can be there when there are not people to do these jobs,” Maor said.

The twin forces of urbanization and the desire for more steady jobs have reduced the percent of the global workforce engaged in agriculture to 26 percent from 43 percent in 1991, according to the World Bank. In developed countries like the United States, Europe and Australia, harvesting jobs are relatively low paid, temporary, physically challenging and often filled by migrant workers, who also must deal with changing immigration policies. The decline in farm labor comes as the world strives to increase food production to meet the needs of a growing population.

Tevel’s system constantly updates farmers through a mobile phone app on how many pounds of fruit have been picked, and how much time it will take to finish the harvesting job (Tevel)

Tevel’s solution is still in the testing phase, but the company expects to introduce it soon to an eager global market. The agricultural robotics sector, worth about $4.6 billion, is estimated to be growing more than 30 percent each year as farmers look to solve labor shortages and harvest costs. The Covid-19 pandemic, which has exacerbated the labor shortage, with countries closing borders and halting or limiting the issuing of visas and work permits, has further increased interest in robotic harvesting. In addition, there have been several Covid-19 outbreaks reported among seasonal agricultural workers in the United States, Europe and elsewhere.

“The coronavirus has definitely pushed the sector forward,” Maor said. “But even beyond the pandemic there is a large need that people really want to fill.”

Maor, a specialist in robotics and vision systems, who holds two engineering degrees from the Technion Israel Institute of Technology, has never worked in agriculture. But about a decade ago, while director of research and development at the touch-screen technology company Lumio Inc., he happened to see a documentary about harvest workers that highlighted the difficulties for farms as well as laborers, who often face long hours and harsh conditions.

“I said to myself, this is a huge problem that we must be able to solve with technology,” he recalled. But after spending some time visiting orchards to observe how fruit-picking works, he reluctantly concluded that technology probably couldn’t help much because robots could not know which fruits to pick and which to leave on the trees.

“But I said back then that one day I will return to this problem when the technology is better,” he said.

That day came in 2016, when he realized that artificial intelligence had advanced enough to be able to analyze video and determine which pieces of fruit were ripe enough to pick. So he founded Tevel, and with the help of a team experienced in robotics, machine learning, data analysis and other cutting-edge technology, he began designing the picking system. They ultimately decided to use flying robots tethered to an autonomous vehicle rather than fly-alone drones to avoid having to deal with the problem of short battery life.

“This way there is always a power source on the ground for the robots,” Maor said. Each ground unit vehicle contains six flying disc-shaped robots, with arms that jut out to gently pick fruit. For now, a team of two company technicians accompanies each fleet deployed to farms, but in the future, Tevel expects to rent out the units for farmers to operate themselves.

Backed by the Jerusalem-based equity investment platform OurCrowd, Tevel’s team boasts top Israeli tech and business professionals, including its board chairman Eyal Desheh, formerly CFO at Check Point Software Technologies and Teva Pharmaceuticals.

“Tevel offers a 21st century way to pick fruit,” said the Israeli apple farmer, who began searching for a technological solution a few years ago after struggling to hire enough reliable fruit pickers. “This is our future.”

For more information about Tevel, click HERE.

read more: