Today’s food processors expect more from their sorting machines than simply identifying visible defects. Issues such as insect damage, early-stage mould growth, or subtle structural irregularities can be difficult to distinguish from good product. This is precisely where conventional laser sorting systems reach their limits: they can interpret only a limited number of signals at a time, causing valuable information to be lost.

With Topological Data Analysis (TDA), Optimum Sorting takes a fundamentally different approach. By combining multiple laser signals simultaneously and analysing them as a unified dataset, the system creates a far more comprehensive picture of every product. The result is more consistent sorting, fewer machine adjustments, and greater confidence in what the sorter is truly capable of detecting.

The limits of conventional laser sorting

Does your sorting machine really see everything moving through the line? In reality, the biggest challenge is rarely obvious defects such as foreign material or off-colour products. Those are relatively easy to detect. A black stone among light-coloured nuts, for example, stands out immediately. The real challenge lies in subtle defects: issues that closely resemble good product, such as early-stage mould development, insect damage, or slight structural variations.

This is where conventional laser sorting reaches its limits. Most laser sorters on the market can analyse and combine no more than three signals at the same time. Each signal captures a different characteristic of the product, such as colour, fluorescence or structure. Because only a limited number of signals can be processed simultaneously, operators are forced to choose which data to prioritise and which data to ignore.

Those decisions often rely on experience and can make sorting performance highly operator-dependent. A setup that performs well on one shift may deliver different results on the next. As a result, valuable information is left unused, not because it is unavailable, but because conventional sorting systems cannot interpret all relevant signals simultaneously.

TDA: turning six signals into a single, complete product view

TDA starts from a fundamentally different principle than conventional laser sorting. Every laser signal contains valuable information. Some signals are more sensitive to colour differences, while others reveal structural characteristics or subtle product variations. To make reliable sorting decisions, those signals need to be analysed together rather than in isolation. That is exactly what TDA does.

TDA processes six laser signals simultaneously, allowing every available data point to contribute to the sorting decision. Each pixel is analysed using its full spectral response, creating a unique fingerprint for every object. This fingerprint combines all relevant product characteristics into a single profile. Rather than relying on simplified 3D models, TDA analyses data in a high-dimensional space, where advanced clustering algorithms reveal patterns that would otherwise remain hidden.

The impact becomes clear when sorting real products. A nut affected by mould may show little or no visible colour difference, making it difficult for conventional sorting systems to identify. TDA can detect subtle correlations across multiple laser signals, bringing small but meaningful deviations to the surface. When six independent laser signals all suggest the presence of a defect, the evidence becomes far more compelling than when only three signals are available. By shifting the focus from individual signals to pattern recognition, TDA enables more robust, consistent and data-driven sorting decisions.

Understanding topology: the science behind TDA

Topology is a mathematical way of analysing shapes and patterns without focusing on exact dimensions or individual details. Instead of asking, “How big is it?” or “How hard is it?”, topology asks: “What does the pattern look like, and how are its different elements connected?”

So, what does this mean for sorting technology? Traditional sorting systems evaluate a product using three signals, whereas TDA expands this to six. These additional signals provide a much richer set of data, making subtle differences easier to detect and improving contrast between products. While minor variations can sometimes be difficult to distinguish when only three signals are available, six signals add valuable context and create a more accurate representation of the product. As a result, TDA can identify defects and abnormalities with greater reliability, even when the differences are extremely subtle, whether in colour, texture, or biometric characteristics.

How does TDA improve your sorting line?

TDA proves its value where conventional sorting systems struggle most: detecting subtle defects without adding complexity to daily operations. A few examples from everyday sorting applications:

TDA also simplifies day-to-day operation through the Unknown Filter. Anything that doesn’t match the learned product profile is automatically flagged as “unknown”, giving operators an additional layer of confidence without the need for constant fine-tuning. The result is straightforward: fewer adjustments, smoother product changeovers and more consistent sorting performance, even when products, operators or operating conditions change.

How TDA makes sorting easier and more reliable

TDA gives you greater control over your sorting process by using more relevant data and interpreting it in a smarter way. The biggest benefits are reflected in ease of use. You’ll quickly notice that getting started with the sorting process becomes easier for everyone involved.

This technology integrates seamlessly into your existing Ventus or Novus sorting machine. No layout modifications, additional floor space, or capacity changes are required. Your sorting process continues to operate exactly as it does today, with TDA adding an extra layer of intelligence on top of your current setup. For operators, day-to-day operation remains largely unchanged. Behind the scenes, however, TDA delivers more consistent sorting performance, reduces variation in reject rates, and minimises the need for manual adjustments.

How food processors benefit from TDA

At Optimum Sorting, we build our machines around one clear belief: technology should support the day-to-day reality of food processors, not make it more complex. With TDA, we enhance operators’ knowledge and expertise with richer, more reliable data. We do not see TDA as a replacement for human expertise, but as a way to make that expertise more consistent and predictable. By providing deeper insights into the sorting process, TDA helps food processors achieve more stable production and maintain greater control over their operations, without having to change the way they work.

Interested in a free demo?

Get in touch with the sales team of Optimum Sorting and have your products tested at one of our demo centers around the globe.
Experience firsthand the benefits our machines can bring to your business!