Enhance Food Safety and Quality Assurance with AI
Detect and eliminate more foreign materials and product defects with Borde's AI. Achieve advanced levels of quality and safety by leveraging the power of Borde's Plant and Cloud Advisor products.
Every Quality Oversight Impacts More Than Just Sales
In the food industry, quality oversights can lead to significant health risks and erode consumer trust. Maintaining the highest standards of quality control is not just regulatory—it's a commitment to consumer safety and brand integrity.
of food recalls in the past year were attributed to contamination and quality oversights.
of consumers report they would avoid purchasing from brands that have experienced even a single food safety incident.

Deploy in hours.

Integrate with PLCs, Databases for full automation.

Both Borde Plant Advisor and Cloud Advisor solutions adapt and become more efficient and effective every day.
Frequently Asked Questions
The Borde Inspect System combines a patented AI software (aka Borde Inspect Advisor) and AI hardware to recognize a wide range of product defects and foreign objects. Borde Inspect Advisor is our premier AI inspection software, that is designed from ground up in the Silicon Valley- not only to handle 1 Million objects per second so that each foreign object and defect can be inspected, but also provide the most accurate inspection system to avoid recalls and reduce human labor.
Yes, the Borde Inspect System conducts single-sided inspection of the product that flows past the system. However, as Borde Inspect sees every kernel, in direct comparison with sample data from customers, Borde Inspect achieves high accuracy across defect classes with the added bonus of seeing every piece of foreign material and outperforming manual sampling for defect classes that are less likely to show up in a manual 500 gram sample (gummy, shrivel, etc.). Because it is single sided on all almonds, Borde statistically normalizes some defect categories and those results have been proven to match on exact bin count or bigger sample size QC measurement.
The Borde Inspect System is volume agnostic, capable of inspecting up to 20 tons an hour per line. Processors should feel comfortable running their lines as fast as they want with no data/detection loss from processing speed.
The Borde Inspect System maps to the governing standard by which the product is graded. As such, each model contains data for whatever product it is inspecting. In almonds as an example, the Borde Inspect System captures and categorizes kernels into edible kernels and multiple dozens of defect and foreign material subcategories. It then returns that data with a USDA grade and breakdown by piece count and percentage by weight.
Yes, the Borde Inspect System can size product. One of our most popular deployments for almond processors is on the sizer, returning bin level quality control reports with accurate CPOs through use of a proprietary sizing model. For walnuts, the Borde Inspect System can also size product into halves and pieces with a high degree of accuracy.
One of the benefits of the Borde Inspect System is its ability to be fully automated. Integration with check weigh scales, PLC signals upstream or downstream and 3rd party ERP/inventory management systems creates a seamless end-to-end process flow, capturing images and creating quality control reports on a lot, bin or box level without the need for a human in the loop. For lines where PLC automation is not available, a manual start/stop of the Borde Inspect System is available through the UI that all Borde customers have access to. Database export remains automated at the end of the report.
At present, the Borde Inspect System operates independently from sorting machines. However, the Borde Inspect System captures and returns quality control data in real time anywhere in the plant, allowing for faster operator action in adjusting sensitivities on sorting systems to account for deviations in product quality. Looking forward, Borde expects the market to move towards allowing real-time closed loop feedback to sorting systems for tighter control and automation within the plant, minimizing the need for expensive human labor.
The Borde Inspect System tends to outperform manual sampling or other 500 gram sampling AI quality control systems. As both operate on outdated sampling methodology of 500 grams on 2200lbs, the quantity of product inspected is much more representative of the product that is actually running through the plant. Frequently, foreign material and certain defect categories will not show up at all in a traditional sample, but that doesn’t mean they aren’t there. When they do show up in a sample, extrapolated out, the numbers indicate that there are pounds of foreign material or certain defects in a bin which is also not true. The Borde Inspect System’s ability to inspect every kernel ensures that processors are receiving highly accurate quality control data that is significantly more representative of the overall product quality. What we find is that as the sample size or sample rate increases, the average numbers from the increased sample size or rate trend more closely to the numbers the Borde Inspect System returns.Additionally, the Borde Inspect System is a fully in-line solution that requires minimal to no change to existing process flows. The Borde Inspect System can be flexibly deployed anywhere in the process flow where a single layer flow is present or where processors need quality control data, capturing and returning quality control data in real time. Compare that to existing sampling methodology or Qcify technology that is not in-line and still requires a human to pull a sample from the line.
A single Borde Inspect System can be installed in as little as 6 hours. For larger deployments such as a full sizing line, the installation takes about a week.
The Borde Inspect System can detect any defect or foreign material the human eye can see. Our models are trained to detect all relevant defects and foreign material particular to the crop. The Borde Inspect System can also grade based on color.





