Artificial intelligence (AI) is also becoming increasingly common in agriculture. The southern German startup Karevo GmbH has now succeeded in designing an AI-supported optical sorting system for table potatoes and, after several years of development, is about to enter the market. "The response to our joint trade fair appearance with our partner Euro-Jabelmann at expoSE in Karlsruhe was particularly encouraging. We see a discrepancy in that many family businesses and direct marketers are forced to invest in automation due to a lack of personnel but are too small to afford conventional optical sorting lines. We aim to enable this target group to take the step towards automation with an inexpensive but functional process," explains co-founder and managing director Benedikt Keßler on request.
Johannes von Wittke and Benedikt Keßler from Karevo GmbH with Heiner Bosmann and Gitti Veurink-Bosmann from EURO-JABELMANN Maschinenbau GmbH at expoSE 2024
It has taken around four years from the initial idea and concept development to the official market launch, which will take place in November at Agritechnica. The biggest challenge was the compact, space-saving design of the process, according to Keßler, who grew up on his parents' potato farm and is therefore familiar with the challenges of the sector. "We have equipped several test farms across Germany with demo systems so that anyone interested can see the machine in action in their own region." Based on the inquiries received so far, the greatest potential for application is in direct-marketing potato farms, although larger agricultural businesses with 10-20 hectares of edible produce have also expressed interest.
The partial solution from Karevo GmbH is compatible with all standard sorting tables so that the existing infrastructure can continue to be used. Thanks to AI and 3D cameras, the process enables reliable detection of foreign objects, shapes, and defects (including rot, wireworms, deformities, and injuries).
Throughputs of up to 5 tons/hour
According to Karevo GmbH, the cycle rate of the sophisticated sorting process is at least 3 tons/hour. "Experience has shown that many companies also sort oversized and misshapen potatoes on the same conveyor belt, but in theory, throughputs of up to 5 tons per hour would also be realistic. Otherwise, the feedback from our test partners has been consistently positive. However, it turned out that the many adjustment options available to the user are not needed, and that a simplified user interface is preferred. Apart from minor adjustments, the basic settings are generally hardly ever changed."
Potential applications for carrots and onions
Depending on throughput and farm size, the system should pay for itself within three years. Although the process is primarily aimed at the potato industry, there is also potential for future applications in other areas. "Many family-run farms produce onions or carrots in addition to potatoes. Accordingly, it would be feasible to optimize the system so that it can also be adapted to these crops. There are also ideas for developing an even more compact line or a model with multiple outputs. But that's all still in the future," concludes Keßler.
Positive attitude toward AI as an opportunity for efficiency and sustainability
Cooperation partner EURO-JABELMANN also strongly believes in the use of AI to increase efficiency and improve quality in potato technology. According to managing director Gitti Veurink-Bosmann, Karevo's technological expertise is an important addition to the company's existing skills. "We are convinced that precise data analysis contributes to improving and automating cultivation, harvesting, and sorting processes, as well as increasing harvest quality through better detection and classification of potatoes. Looking ahead, we also see opportunities for integrating AI into monitoring and process control."
The recognized industry supplier based in Itterbeck (Lower Rhine) has a positive attitude toward AI as an opportunity for efficiency and sustainability, emphasizing Veurink-Bosmann. "We also see potential for other crops, as similar challenges exist in harvesting and sorting. For example, we could imagine similar AI-supported processes for improved quality control and sorting of onions, carrots, and other products. We also see this expansion of the technology to other crops as a logical further development of our collaboration with Karevo."
For more information:
Benedikt Keßler
Karevo GmbH
Lauberg 42
85417 Marzling
[email protected]
www.karevo.de
Gitti Veurink-Bosmann
EURO-JABELMANN Maschinenbau GmbH
Wilsumer Str. 19-21
D-49847 Itterbeck
Phone +49-5948-9339-0
[email protected]
www.euro-jabelmann.de