BIG Interview: Andrea Saadeh on replacing slow fruit sampling with faster decisions at intake 

Packhouses do not lose time only on the grading line. They lose it earlier, while trying to understand what has just arrived: which sizes are in the load, how colour is distributed, what should be packed first, what can wait, and how sales teams should act before sorting is complete. 

Hectre gives packhouses a faster way to assess incoming fruit at this critical stage, replacing slow, labour-heavy sampling with a more immediate digital view of each load. The aim is simple: help teams make better decisions while choices around storage, grading, sales, and grower feedback can still make a real difference. 

Fructidor spoke with Andrea Saadeh, Regional Director EU at Hectre, about how the technology works, why the company says its digital sampling can reach around 95% accuracy, what it changes for packhouse operations, and why France is becoming an important market in its expansion strategy. 

Fructidor: How does Hectre’s digital sampling technology operate? 

Andrea Saadeh: Our system analyses bins or boxes captured via fixed cameras or mobile devices of any kind — a smartphone, a tablet, or other compatible equipment. Thanks to our AI technology, we can determine the percentage of size distribution and colour distribution. 

So, for each size, we provide the percentage, and for each colour, the distribution. The high-precision data we detect provides results for every single load that has been processed. This is what we describe as digital pre-sorting. 

Does this mean packhouses no longer need sorting machines? 

No, this is something different. Packhouses still need sorting machines, especially because of retailer and market requirements. But before sorting, when they receive fruit, they already need to understand what is entering the packhouse. 

Today, this sampling is often done manually. Some larger companies have small sampling machines, almost like a dedicated mini-sizer used only for sampling. But then you need to buy that machine, and you still need people to unload the fruit, check it, and manage the process. 

What we intend to do is reduce this cost for packhouses by giving them higher sampling accuracy in a much more efficient way. 

How much can a conventional sampling machine cost compared with your solution? 

A sampling machine, depending on what you are looking for, can easily cost €100,000 to €150,000. On top of that, there is the cost of the people needed to operate it and check the fruit. 

With Hectre, the goal is to reduce the overall cost of sampling while improving the accuracy of the information packhouses receive. 

How accurate is the system? 

With AI and the photo or video data we capture, we can achieve, on average, around 95% accuracy compared with the real fruit in each load. 

Once customers have our results, they can compare them with the final sizing results from their real sizer. In practice, they can see that our error is usually 2–3%, and generally not more than 5%. 

Do customers trust AI in this kind of decision-making? 

I would say it is not really a matter of trusting AI. The results are clear. 

As I always tell customers: do not trust me — trust your eyes, trust what you see. We make the test with our technology, then the customer sizes the whole load through the real sizer, and they can immediately see the accuracy for themselves. 

Compared with manual sampling, our data is objective. Manual sampling can also be expensive because it requires many people, a lot of time, and trained staff. With us, you do not need so many people or expert operators, because the technology does the work. 

Beyond accuracy, what practical benefits does this bring to packhouses? 

Packhouses can optimise the space in the backhouse and reduce costs in general. It is also a major help for sales teams. 

After we process a load, the sales department can know within about a minute exactly which sizes and which colours are present in that specific load. That means they do not need to wait until final sorting is complete. They can start building their commercial strategy and preparing sales decisions from the moment the fruit enters the packhouse. 

What would you say directly to a potential customer considering Hectre? 

We do digital sampling of fruit using artificial intelligence and camera vision. We are able to determine the size and colour of fruit within seconds, in a way that works almost like radar. 

Our data is about 95% accurate, and it can be used for storage decisions, sales planning, and feedback to growers — helping them improve harvesting decisions and avoid mistakes during the harvest period. 

Customers still need to sort their fruit, but with our data they can know what to grade first, what to leave for later, and what should not wait. This helps them avoid unnecessary stock build-up and food waste. They can also speed up the grading line because they choose the most efficient slot and prepare goods much faster than usual. 

So the main advantage is saving time? 

And money. There are labour savings as well. For example, if a packhouse does not need to run the grader for 16 hours and can instead run it for 12, that is four hours saved. It may also mean not needing an additional 16 people working for those extra four hours. When you multiply that by labour costs in different countries, the impact becomes very clear. 

Which crops does Hectre currently work with? 

Apples, pears, cherries, stone fruit, and citrus. We are also always open to building knowledge and doing more research on different fruit types and crops. 

What are the next product developments in the pipeline? 

By the end of the year, we expect to be ready with external defect detection. It is not a different sampling process, but an extension of the information we can provide. 

Other crops are part of our second development path and should be ready next year, although today we do not yet have a specific date. We are working on kiwi, blueberries, and tomatoes. 

Our priority is external defects first, and after that also internal defects, meaning the internal quality of the fruit. 

Hectre is also targeting France. What is special about this market for you? 

In France, we are still relatively new, but it is one of the key markets we want to develop, with a particular emphasis on apples, stone fruit, and cherries. 

Is the solution already available internationally? 

Yes. We work in 22 countries around the world. Just a few weeks ago, a Swiss company joined us. We have strong coverage in the US market and in Chile. We are also growing in Europe and South Africa, and we are looking forward to what the future will bring. 

Packhouses looking to reduce manual sampling time and make earlier grading, storage and sales decisions can send their inquiry to discuss how Hectre's solution fits their operation.