How Orbem’s industrial MRI technology is eliminating guesswork, waste, and labour dependency from fresh produce quality control
The cost of hidden defects
For decades, the only reliable way to assess the internal quality of fruit like watermelons and avocados has been to cut them open. This practice runs throughout the entire supply chain, from growers to supermarket shelves, and results in up to 10 per cent of a batch being sacrificed as samples, then discarded.
The shopper’s experience mirrors this uncertainty. A watermelon in a supermarket is often a gamble: you simply cannot know what is inside until it’s too late.
Seeing fresh produce from the inside out
Orbem, a Munich-based deep-tech company founded in 2019, has industrialised Magnetic Resonance Imaging (MRI), combining it with Artificial Intelligence (AI). The German group has transformed the same technology used in hospitals for decades into an application that is fast and affordable enough to deploy on a production line.
“We basically create a digital 3D twin of the product and identify every internal characteristic our customer cares about, including maturity, bruises and much more,” explains Johann Steinert, director of new markets at Orbem.
How is that possible? Thanks to state-of-the-art machine learning, Orbem can turn complex MRI data into simple yes or no decisions. Working closely with each customer, Orbem trains its AI models to recognise specific defect signatures such as bruising, internal browning, mould, as well as ripeness stage.
Speed and scale built for industry
One of the central challenges in bringing MRI to industrial use cases is throughput. A technology that slows the line adds no value.
Orbem’s system is engineered to keep pace with industrial volumes. This means processing watermelons, for example, at roughly 4,000 fruits per hour, equivalent to 20,000kg per hour in a large facility. The AI models that power the sorting decisions achieve industry-leading accuracy which also continues to improve with every additional scan the system processes.
Fully automated scanning and sorting
The entire process, from fruit entering the scanner to being sorted into various quality classes, is end-to-end automated.
Orbem’s solution for the fruit sector reaches penetration depths that are unparalleled. Seeing clearly all the way to the core is no longer a dream but an advantage in operational efficiency available to producers now.
“If you can guarantee consistent quality to your customers, that pays off – for growers, packers and for the whole supply chain,” says Steinert.
A system that keeps getting better
Unlike conventional quality control technologies available, Orbem’s AI models improve continuously. Industrial production generates tonnes of scan data every day, and every scan makes the model more accurate.
Addressing the labour challenge
Beyond waste reduction, Orbem’s technology speaks directly to an increasingly urgent operational problem: labour.
Quality assessment for watermelons and avocados today often requires employing highly trained staff with years of experience. As labour markets tighten across European agriculture, this kind of skilled, repetitive inspection is becoming harder and more expensive to sustain. Orbem’s technology removes the dependency entirely.
Expanding into fresh produce
Orbem first brought its AI-powered MRI technology to the poultry industry, where it is used to determine the sex of a chick embryo while still inside the egg. The company now has around 200 employees and 25 active installations across several European countries.
Find out how Orbem’s technology can support your operations.
Connect with the team at Macfrut: Pavilion B1 Stand 021
Read more about Orbem’s solution here:
