Grainswest - Tech 2025

Tech 2025 grainswest.com 27 Make the grade While it works to perfect its on-combine tech, Ground Truth Ag has produced a benchtop unit. It can grade grain for protein, moisture and defects. When Kyle Folk launched his business in 2022, he had a simple yet audacious goal: To benefit farmers, make grain grading more accurate and consistent. Traditional grading is not perfect and never has been. However, he was convinced that mechanization could make the process better than ever. He first sampled grain as an 11-year-old driving the family’s grain truck at their Holdfast, SK, farm. Since then, he has come to realize the value of sending samples off to buyers, but also that the results can be imprecise. The grading risks his father shouldered always stuck with Folk and would eventually lead to the creation of Ground Truth Ag. From the start, the feedback was clear. An attempt to build automated grading machines with improved accuracy had been tried and failed, so do not waste your time. “Today, it’s the opposite,” said Folk, who sought to prove the skeptics wrong. “Everybody has seen what we’ve done, and they all see that it works. It just keeps getting better. It’s no longer a mountain that can’t be climbed.” Ground Truth Ag The company had initially focused efforts with on-combine grain grading. While it still works to build that market to a commercial release, it has largely shifted focus with the introduction and subsequent popularity of its benchtop unit used at country terminals. The unit is in the early commercial phase of development. An 80 per cent accuracy rate is the generally accepted guideline for personnel who carry out grading at CGC- licensed grain elevators. Though not a pretty number, it’s the reality. “Grain buyers don’t want their grain to be subjectively graded, they want it objective, but it’s hard when humans are responsible for grading the grain,” said Folk. “Inevitably, you end up with this process that gets questioned.” Today, the Ground Truth benchtop unit can grade grain for protein, moisture and defects through visual and non-visual grading factors. This is achieved through NIR spectroscopy, machine vision and machine learning technology that can analyze the inner and outer kernel. As an example, Ground Truth’s unit initially correctly graded CWRS at a rate of 85 per cent, which has only gone up as the machine has been fine-tuned. The product is sensitive enough to determine, for instance, an accurate mildew damage rating of light, medium, heavy or severe. “We are just getting started,” said Folk, of the machine’s expanding capability. Feedback from grain buyers has been overwhelmingly positive, and they are pleased to see a machine that produces results with greater accuracy, he added. The benchtop unit can now grade spring wheat, red lentils and soybeans. This year will see the addition of corn, durum and oats to its grade modelling algorithms. In a world increasingly powered by AI and machine learning, Folk believes greater accuracy and consistency in grain grading can become a reality for farmers. “This tech was created to provide consistency to the process and remove fragility in the supply chain,” he said. “Until this is mainstream, I feel like the job is not done. We need to keep pushing forward and get the product out there, get it used, adopted and be standard issue. If we get buyers to start using this, then everyone in the industry will use it.”