Potato Processor Balances High Demand and Quality with Sorting Tech: ProFood World

How can a processor ensure consistent quality when managing high demand?

To meet needs for both high output and consistent product quality, Bem Brasil installed a Key Technology processing line capable of running 30 metric tons of frozen potato strips each hour, with a layered sorting system to find and remove any defects.

This is an AI-generated episode. Read the full featured article on ProFood World.

Transcript

SPEAKER_00 0:03

Welcome to Through the Line, the podcast exploring innovations and information across the packaging and processing landscape. From topics impacting consumer packaged goods and healthcare packaging to the latest technologies and food processing operations.

SPEAKER_02 0:22

Hi, I'm Casey Flanagan, Associate Editor with ProFood World. This AI-generated podcast episode covers how Brazil-based potato processor, BEM Brazil, uses integrated sorting to balance high production volumes and consistent quality. It explores the company's two-stage optical inspection setup featuring technologies like automatic defect removal and sort-to-grade software to maximize yield across high-capacity processing lines.

SPEAKER_03 0:46

This deep dive explores how Brazil-based potato processor, BEM Brazil, utilizes an integrated two-stage optical sorting system to balance high volume production with maximum product quality and yield. Because if you are managing a processing plant, you already know the inherent headache of agricultural products.

SPEAKER_01 1:05

Oh, absolutely. It is a constant challenge.

SPEAKER_03 1:07

Right. I mean, we are not dealing with uniform widgets perfectly stamped out of plastic here. We are dealing with natural potatoes that, you know, change in moisture, size, and quality hour by hour.

SPEAKER_01 1:17

Yeah, batch by batch, really. And how do you automate that level of variability without losing your mind? Or, well, your profit margin.

SPEAKER_03 1:25

Exactly. And that is the exact scenario Ben Brazil faced. Just to give you a sense of their scale, they produce over half of all the French fries consumed within Brazil.

SPEAKER_01 1:32

Aaron Powell, which is just a staggering amount of food.

SPEAKER_03 1:35

It really is, alongside a massive international export business. We're talking about an annual production volume of roughly 500,000 metric tons.

SPEAKER_01 1:44

Wow. And you know, operating at that kind of scale introduces logistical and mechanical hurdles that require incredibly precise engineering to overcome. The central operational challenge here is exactly what you just highlighted: natural variation. Right. Yeah. Celio Zero, the director of operations at BEM Brazil, noted that their existing production lines had simply reached a point where they could no longer keep pace with growing market demand. But when you are processing agricultural goods, simply turning up the dial to accelerate a production belt is never a viable solution.

SPEAKER_03 2:17

Because the incoming potatoes are entirely unpredictable, right?

SPEAKER_01 2:20

Precisely. Their quality fluctuates multiple times within a single shift. It depends on where they were grown, the soil conditions, how they were stored. Therefore, Bem Brazil did not just need a faster belt.

SPEAKER_03 2:30

They needed something smarter.

SPEAKER_01 2:31

Exactly. They required a processing system capable of adapting dynamically to these continuous product shifts without compromising the final output. That is the core friction in high-volume food processing, maintaining an aggressive throughput while standardizing an unstandardized raw material.

SPEAKER_03 2:48

So to resolve that tension, BEM Brazil partnered with Key Technology to install a processing line capable of running that massive 30 metric tons of frozen potato strips per hour. And when you are trying to push that much volume, you have to completely rethink how quality control is conceptualized.

SPEAKER_01 3:05

You really do.

SPEAKER_03 3:06

You cannot just place one giant inspection bottleneck at the end of the line and hope for the best. The sheer volume necessitates a highly sophisticated, layered sorting system engineered to detect and remove defects continuously.

SPEAKER_01 3:18

Yeah, exactly. When you're processing at that velocity, quality control must be deconstructed. It has to be separated into specialized, highly targeted stages.

SPEAKER_03 3:27

You have to attack the problem sequentially.

SPEAKER_01 3:29

Right. And this high volume requirement leads directly into the necessity of what happens first in the facility, which is the upstream wet area sorting. This is where the product is washed, cut, and you know, moving through the line before it ever sees a freezer.

SPEAKER_03 3:40

Let's just examine that first layer because the hardware involved is fascinating. In the upstream wet area, Bem Brazil installed three V Erie YX B210 sorters.

SPEAKER_01 3:50

Those are serious machines.

SPEAKER_03 3:52

They are. They are equipped with off-axis cameras arranged in what key technology calls a tilted X configuration. So I am picturing this setup, and if you just place a potato strip on a standard flat conveyor belt, a camera looking directly down only sees the top surface.

SPEAKER_01 4:08

Exactly. You completely miss the bottom and the sides.

SPEAKER_03 4:10

Yeah. So this tilted X configuration is basically putting the fry through a calibrated hall of mirrors. Right. Because agricultural defects can materialize on any side of that three-dimensional potato strip. This specific camera geometry ensures there are absolutely zero blind spots.

SPEAKER_01 4:27

It's a full 360-degree view.

SPEAKER_03 4:29

Right. It provides a true full surface inspection of every individual piece moving through the machinery at high speed.

SPEAKER_01 4:35

That hall of mirrors analogy is exactly right. But understanding why this full surface inspection occurs specifically in the wet area is critical if you're designing a processing line. Why is that? Well, placing this rigorous level of inspection early in the overall process fundamentally alters the dynamics of the entire plant. By eliminating major defects upstream, the system creates a significantly more uniform product stream before the potatoes undergo further processing, like freezing or flavoring. This uniformity acts as a massive force multiplier. Think about the downstream equipment. Your freezers, your packaging machines, and your secondary sorters inherently performed there require less maintenance and experience far fewer operational jams.

SPEAKER_03 5:17

Because the input stream is clean and standardized.

SPEAKER_01 5:20

Exactly. If you push defective, clumped, or misshaven product into a freezer, you create mechanical nightmares down the line. Wet area sorting is a proactive approach to quality control, stopping the chaos before it compounds.

SPEAKER_03 5:33

Wait, that brings up a massive operational pushback, though. If that hall of mirrors is catching every single tiny blemish and we are kicking all those fries off the line right at the start, aren't we throwing away tons of perfectly good potato?

SPEAKER_01 5:46

That is a very valid concern.

SPEAKER_03 5:48

Because if you are managing a plant's profit and loss statement, that yield loss would be terrifying. Rejecting an entirely usable French fry just because it has one localized dark spot fundamentally hurts the bottom line.

SPEAKER_00 6:00

Yeah.

SPEAKER_03 6:01

So how does an operation balance the need for that pristine stream with the absolute financial necessity of maximizing product yield?

SPEAKER_01 6:09

That is the precise operational friction that Bam Brazil solved. And they did it through the integration of an automatic defect removal system, specifically the ADR exos.

SPEAKER_03 6:19

Okay, how does that work?

SPEAKER_01 6:20

Well, this piece of equipment is strategically placed immediately downstream of those initial wet area VREX sorters. Instead of routing the rejected strips directly to a waste bin, the system diverts them into the ADR Exos.

SPEAKER_03 6:33

Oh, interesting.

SPEAKER_01 6:34

Yeah, this machine physically accepts the defective strips and utilizes a highly precise automated mechanical cutting system. It essentially maps the fry, isolates the exact location of the blemish, mechanically excises only that specific dark spot, and then takes the remaining clean portions and automatically returns them to the primary line.

SPEAKER_03 6:51

Wow, that is an incredible mechanism for waste recovery. It operates on the philosophy that a single blemish should not condemn an entire product to the garbage.

SPEAKER_00 7:00

Exactly.

SPEAKER_03 7:01

And importantly, this yield optimization strategy begins even before the product reaches those wet area sorters. Bem Brazil employs another vital layer of stream cleaning at the very beginning of the process.

SPEAKER_01 7:12

Right, the sizing equipment.

SPEAKER_03 7:14

Yeah, they utilize three machines called sliver sizer removers. These do not use optical cameras, they physically and mechanically grade the incoming product to reject small slivers and nubbins. This ensures that only properly sized potato strips move forward into the Viurix sorters in the first place.

SPEAKER_01 7:31

This creates a highly engineered sequential approach to stream management that, you know, any plant manager can appreciate. First, the sliver sizer removers eliminate the geometrically incompatible pieces, preventing them from cluttering the optical sensors.

SPEAKER_03 7:45

Very taking out the trash early.

SPEAKER_01 7:46

Exactly. Second, the VaryYX sorters identify the surface defects using that full surface optical inspection. Third, the ADR Exos recovers the usable mass from those defective pieces through mechanical cutting.

SPEAKER_03 8:01

It is a very comprehensive system.

SPEAKER_01 8:03

It is. By the time the product stream exits the wet processing stage, it is aggressively clean, highly uniform, and yield maximized. The heavy lifting of quality control is essentially done, which prepares the product for the final, most rigorous hurdle, the frozen area.

SPEAKER_03 8:19

Let us step into that downstream frozen area, which is located immediately prior to the packaging phase. Here, the operation utilizes three V AEX B175 sorters.

SPEAKER_01 8:29

And these are quite different from the upstream ones.

SPEAKER_03 8:31

Yeah, they are distinct from the upstream wet area sorters we just discussed because they're fitted with both cameras and lasers. This dual sensor setup performs a highly complex three-way sorting process. It actively separates short potato strips into a completely dedicated alternative stream for different product lines, while simultaneously targeting and removing any remaining foreign material or latent product defects right before the final product enters the bagging phase.

SPEAKER_01 8:56

The hardware in the frozen area is certainly impressive. But the true processing capability is governed by the software functioning as the brain of this stage. Key technology utilizes a proprietary program called Sort to Grade, or STG.

SPEAKER_03 9:11

I was reading about that. It's fascinating.

SPEAKER_01 9:13

It is crucial. STG is not a simple pass or fail algorithm. It does not just look at a fry and say good or bad. The STG software dynamically evaluates the precise dimensions and the specific quality profile of every single individual strip passing through the sorter in real time.

SPEAKER_03 9:29

I love how the mechanics of this sort to grade software work. Think of it like a forgiving academic professor who is grading a difficult exam on a curve.

SPEAKER_01 9:37

That is a great way to look at it.

SPEAKER_03 9:39

Right. Instead of applying a rigid binary standard to every single fry, the software evaluates that specific fry based on exactly how its inclusion will impact the final aggregate quality grade of the entire batch. Yes. For example, if a plant needs a bag of fries to be 95% free of minor defects to meet a fast food customer specification, and the STG software calculates that the current batch is tracking flawlessly at 99%, it will mathematically determine it can safely accept a strip with a very minor imperfection.

SPEAKER_01 10:13

Exactly. It lets that fry pass because it knows the overall bag will still hit the 95% target. This dynamic calculation maximizes overall yield continuously, without ever failing the customer's strict quality expectations.

SPEAKER_03 10:26

That algorithmic flexibility is brilliant for protecting the bottom line.

SPEAKER_01 10:30

It really is. And that software is then paired with another crucial technological detail: key's pixel fusion technology. Pixel fusion operates right at the sensor level. It doesn't just take a picture from a camera and a reading from a laser and compare them later.

SPEAKER_03 10:43

Weed, so it processes them together.

SPEAKER_01 10:45

Yes, exactly. It combines the data streams from both the cameras and the lasers simultaneously at the exact pixel level. This fusion of data dramatically dials in detection accuracy. It is specifically engineered to identify the most notoriously difficult to find foreign materials.

SPEAKER_03 11:04

I can imagine that in a frozen state, distinguishing between different materials is incredibly difficult.

SPEAKER_01 11:09

Oh, absolutely. Think about the physical characteristics of the product at this stage. You have frozen potato strips, you might have chunks of clear ice that have formed during the freezing process, and you might have a clear piece of foreign plastic that accidentally entered the stream.

SPEAKER_03 11:24

Which is a huge safety issue.

SPEAKER_01 11:25

Exactly. To a standard camera, clear ice, clear plastic, and highly frosted potatoes can look virtually identical. A laser might detect the density difference, but it might lack the spatial resolution to accurately map the object's boundaries.

SPEAKER_03 11:39

I see the problem.

SPEAKER_01 11:40

Right. So Pixel Fusion combines the spatial resolution of the camera with the material identification properties of the laser perfectly. It ensures these dangerous anomalies are identified and ejected immediately ahead of packaging, safeguarding the final consumer from finding a piece of plastic in their fries.

SPEAKER_03 11:56

That is incredible. Now, beyond the optical sorters, the lasers, and the software, there is connective equipment that ensures this entire system functions cohesively. The unsum heroes of this specific production line at Bim Brazil are keys ISOFlow vibratory conveyors.

SPEAKER_01 12:13

Yes, those are essential.

SPEAKER_03 12:14

While sorting technology often receives all the industry focus, optical sensors are completely useless if the product is clumped together. If you have fries stacked on top of each other, that hull of mirrors we talked about earlier cannot see the fry on the bottom.

SPEAKER_01 12:27

It's physically impossible.

SPEAKER_03 12:29

Right. These vibratory shakers evenly spread, distribute, and meticulously regulate the flow of the product as it transitions between the different processing steps. This precise distribution is vital because it guarantees the VR Ionic sorters have a clear, unobstructed, single-layer view of the product stream.

SPEAKER_01 12:46

Furthermore, we must consider the human element and the operational user interface of this integrated line. It is easy to get lost in the automation, but human operators still run this plant.

SPEAKER_03 12:57

Oh, for sure.

SPEAKER_01 12:58

Based on the feedback from Celio Zero, the software interface is designed to be exceptionally clear and responsive. When an operator makes a setting adjustment, the equipment responds exactly as expected, without severe lag or unpredictable behavior.

SPEAKER_03 13:13

Which is critical when you are moving that much product.

SPEAKER_01 13:16

Exactly. This predictability is critical on a facility processing 30 metric tons an hour. Because the raw agricultural material fluctuates, Bem Brazil must execute fast recipe changes multiple times a day. If a new truckload of potatoes comes in with higher moisture content, the line must adjust instantly.

SPEAKER_03 13:35

So the intuitive interface really saves them.

SPEAKER_01 13:37

It does. A responsive, intuitive interface minimizes downtime during these transitions and reduces operator error, which directly protects that massive high-volume output.

SPEAKER_03 13:47

Now, the ultimate metric for any piece of processing equipment is proven reliability over time. A machine can look great on paper, but how does it handle a 24-hour production cycle?

SPEAKER_01 13:59

That is the true test.

SPEAKER_03 14:00

It is. This integrated system reached full production capacity at BIM Brazil in 2022. Since that installation, it has consistently demonstrated its operational effectiveness across multiple processing seasons, effectively managing the inherent variations of different potato harvests year after year.

SPEAKER_01 14:18

That is a solid track record.

SPEAKER_03 14:19

Yeah. This sustained high-volume performance has firmly established key technology as a trusted foundational partner for BEM Brazil's future expansion and growth initiatives in the global market.

SPEAKER_01 14:30

Synthesizing this entire case study reveals a fundamental takeaway for any processing professional managing a facility. Scaling food production is not merely a matter of moving raw material faster down a wider belt.

SPEAKER_03 14:42

It takes a lot more than that.

SPEAKER_01 14:44

It really does. True scalability requires a holistic, full-line approach. It demands that mechanical sorting hardware, intelligent, dynamic software like sort to grade, and sophisticated waste recovery systems like the ADR Exos operate an absolute lockstep.

SPEAKER_03 15:00

Right, they all have to communicate.

SPEAKER_01 15:01

Exactly. Only through this high level of integrated engineering can an operation effectively manage the inherent chaos and variation of natural agriculture while simultaneously meeting strict, uncompromising commercial demands.

SPEAKER_03 15:15

That is so true. And you know, as automated sorting technologies become increasingly precise at evaluating every single product on a granular level to influence a final aggregate batch grade, a fascinating, larger implication emerges for you to think about.

SPEAKER_01 15:29

Oh.

SPEAKER_03 15:29

Yeah, if we can utilize predictive sort degrade logic to intelligently accept minor imperfections without compromising the final product standard in potato processing, how might this exact mathematical philosophy be applied to other areas of global manufacturing to fundamentally eliminate the entire concept of usable waste?

SPEAKER_00 15:48

Thank you for listening to Through the Line Packaging and Processing. You can listen to more episodes on all streaming platforms. Be sure to visit us at packworld.com, profoodworld.com, and healthcarepackaging.com for more packaging and processing news. This podcast was edited by Bree Guns.