At the end of 2022, Saia Agrobotics received new capital to scale up the robot technology it is working on in Wageningen and make it market-ready. An important step is the construction of its own test greenhouse. It should be in place by early 2024.
Unlike many other players in the horticultural robotics market, Saia's experienced team of robotics specialists chooses to bring the plant to the robot instead of the robot to the plant. Roboticists have been fiddling with the latter approach for more than 20 years. The number of truly practical robots is still limited despite the fact that many companies worldwide are now working on robots to (partially) replace expensive human labor.
The founders of Saia, Dr. Ruud Barth and Bart van Tuijl, also tinkered for a long time with robots sent into the 'jungle' of plants in the greenhouse. The men were involved in the development of harvesting robots for cucumber and peppers by Wageningen University & Research. With Saia Agrobotics they have set up their own company. Formally in 2019, but the first steps were cautiously taken in 2017.
"For a long time, people thought we were doing mobile robots," says Ruud a few months after the funding news was announced. "However, we quickly switched to a completely different concept because we realized that mobile robots weren't going to achieve what we wanted. We didn't feel right with that business model."
Mobile plants instead of mobile robots
The concept being worked hard on behind the scenes in Wageningen is on the eve of scaling up. For that, it is important to build their own test greenhouse. In it, CEO Ruud and his team can show on a larger scale that what they have come up with works. A schematic video was used last year to lift a corner of the veil, but not everything could be shown in public yet. In early 2024, the test greenhouse should be there. More will become clear then.
Still, Ruud can indicate the direction in which it is heading. "We want to move towards a factory concept, where the plant goes 'through' the robots piece by piece. That way, we can make meters. In our concept, the plants are mobile and free. This allows the robots to see everything well and work better."
To get that plant 'mobile,' Saia developed proprietary technology. Ruud points out that cucumber and tomato plants grow very long, up to 40 meters, in the case of cucumber. The robot builders choose to hang the plants, which grow on gutters, differently, as Ruud describes it. "We have seen that the business model for this is good. The risk we are taking with this particular way of working is big, but if it works, then the impact is also big."
Car wash
When we spoke to Ruud this spring, the final nods are being made on what the 'greenhouse' will look like. "We are still in doubt whether we will put the greenhouse in a shed or whether we will work with natural sunlight above it," he says. Robots will take care of the crop. "We choose to have a single robot do several operations at the same time. What we actually do is strip the plant, leaf, and fruit at the same time."
Having multiple operations done by a single robot makes it economically attractive, the CEO stresses. But sometimes, it is also more complicated. What Ruud describes evokes comparisons with a car wash. He agrees. "The plant in our robot concept is like a car in the car wash. At each station, the plant gets a treatment. Sometimes, that is leaf picking and harvesting in one. That can also be done with one robot."
Cutting up gutters?
What people in Wageningen focus on most is keeping it simple. That might sound a bit contradictory after reading the particular approach above, but it is nevertheless true. "With us, the plant is completely free, the stem is bare. We then embrace the plant, so to speak, and the robots perform the crop operations." Saia will work with "standard varieties." "But that does not mean that there are not still a lot of gains to be made in breeding, from a robot perspective. However, we cannot wait seven years for that. The grower has been waiting long enough for a total solution."
Officially, the capital raised late last year from a funding round is not labeled Series A financing. Yet it is actually quite close, Ruud believes. "It is an investment to be able to take the step to this." Building a test greenhouse costs money. For now, the Saia concept lends itself best to new construction, it seems. In time, retrofitting existing greenhouses could also be an option. "In existing greenhouses, it is often logistically challenging to get the plant to the processing hall, also because it is often already full of machinery."
A question that has not yet been asked but which comes to mind when hearing the story is: how does the plant actually get to the robot? On a gutter? We ask Ruud if they would "cut up gutters" at Saia to make moving plants around easier. The answer to that question remains forthcoming. Too many details cannot yet be given. "When the test greenhouse is in place, you will see," he says.
Harvest forecast
One last bit about technology: What about harvest prediction? That should be quite possible with this concept, it seems. Ruud nods. "We see every plant very regularly. We work with track and trace. The plant gets a tag. We see from day to day how red the bunches are, for example, and can start using growth models to predict what ripe fruit will come off the plant. That is almost an exact prediction, I expect, especially with climate and weather data included. Then you know the plant balance, and you can also calculate when the next bunch is going to ripen."
How well harvest forecasting will work remains to be seen. Which brings us back to the test greenhouse after all. "With more than twenty years of experience in our pocket, we think we are on the right track. It is and remains a huge challenge, though; we are no longer building a few robots but an entire vegetable factory. We hope not to slip up."
If that does happen, it's not necessarily a bad thing. "Startups are there to take risks, we think. You have to dare to fail. Next year will show whether or not we have proven ourselves." Should the whole concept not work, Saia can always continue with the developed techniques Ruud expects. "We will soon produce the best data in cucumber and tomato country there is." Finally, some advice with a wink: "I can advise against anyone imitating us."
For more information:
Ruud Barth 
Saia Agrobotics
Tel.: +31 (0)317 700 205
www.saia-agrobotics.com