When it comes to robotics, artificial intelligence (AI) is bridging the gap between people and their robotic counterparts.

“AI is like this giant evolving sponge that keeps absorbing critical information based on repetitive human behavior and soon enough, emotional accuracy,” said Jamie Bobyk, marketing manager, Apex Motion Control. “With machine learning in the baking environment, it will move slow enough to garner acceptance and make robotics and automation in our production facilities non-obtrusive and better, and it will happen in the background as this industry evolves.”

Overall, there are two primary types of AI, noted Hunter Schultheis, North Central sales manager, BluePrint Automation. Generative AI creates new content from scratch, such as text generated by ChatGPT, while analytical AI operates within a defined dataset, continuously analyzing and refining information.

“AI vision systems in bakery and snack production are a form of analytical AI,” he pointed out. “These systems constantly evaluate images to assess product quality, orientation, count and other critical factors, enabling robots to intervene in real-time to ensure production efficiency and consistency.”

With 3D vision, robots can accurately assess the shape, size and positioning of products, allowing them to adapt in real-time to changes in the production flow.

“Products can now arrive more chaotically or in groups to the robot eliminating costly alignment systems that take up valuable floor space in the plants,” explained John Weddleton, automation product manager, Harpak-Ulma Packaging. “AI algorithms further enhance this capability by learning from the data, enabling robots to optimize their movements and decision-making, leading to smoother integration with upstream operations and greater overall efficiency.”

Felix Pang, robotic solutions specialist, ABI Ltd., said AI-powered vision systems, such as ABI’s EYE-Q quality control system, have significantly improved robotics' ability to operate more efficiently and adapt to changes in production or packaging lines.

“By providing more accurate and detailed analysis, these systems enable robots to perform complex tasks with greater precision,” he said. “This shift allows robots to make qualitative comparisons — focusing on factors like texture, shape or visual quality — rather than just relying on basic quantitative metrics such as basic size metrics or weight. As a result, robots can better handle variations in products, optimize processes and adapt to different production conditions in real-time, leading to increased efficiency and flexibility on the production line.”

Within the past year, Schultheis said AI and machine learning have also made robotics more adaptable. For instance, AI-driven robots can now adjust pick-and-place patterns based on real-time product flow and positioning, minimizing errors. He added that machine learning helps predict component failures before they happen, reducing downtime.

Meanwhile, machine learning enables robots to optimize their own movement and pressure, improving precision with each production run, noted Wes Bryant, product group leader for AMF Workhorse, an AMF Bakery Systems brand.

Marcus Kurle, cofounder of AAA20 Group, echoed that AI and machine learning have made significant strides in enhancing the efficiency and adaptability of robotics, especially in complex production and packaging lines. On an international basis, however, US manufacturers are still catching up to global standards.

This article is an excerpt from the April 2025 issue of Baking & Snack. To read the entire feature on Robotics, click here.