At the Automation and Robotics Conference 2025, hosted by Fizyr, Fredrik Malmgren, CEO of Cognibotics, discussed a clear shift in industrial automation: vision technology is moving from support function to central intelligence. “The vision system used to be just a tool to identify where to pick,” he said. “Today, it’s taking over the whole dexterity — everything that happens in the image.” The Rocking Robots team sat down with him to discuss trends in robot vision.
Traditionally, machine vision served one purpose: locating and identifying parts for a robot to handle. But according to Malmgren, the technology has now evolved far beyond that. “Vision systems are detecting errors, identifying materials, classifying products, and even evaluating quality,” he explained. “They are taking on the complexity that used to be handled by separate systems.” This deeper visual understanding means robots are becoming more autonomous and adaptable. With vision-driven intelligence, systems can adjust to variations in shape, color, or texture. challenges that once required custom programming or extensive human supervision.”
The growing intelligence of vision systems also accelerates how quickly automation can be implemented. “We’re moving towards systems that can practically set themselves up,” Malmgren said. “Deployment times are getting shorter, because vision takes care of more of the configuration and calibration that used to be manual.” This shift reduces integration costs and increases flexibility — two of the biggest barriers to adopting robotics in smaller production environments. As Malmgren put it, “If we want automation to really scale across industries, we need to make it easier and faster to get systems up and running.” Ultimately, the transformation of vision technology ties directly to industrial goals. “At the end of the day, it’s about productivity,” Malmgren said. “Being quicker, more accurate, and delivering consistent results is what makes automation valuable. Vision is what makes that possible now.”
See also
The eye of the robot: Video Report from the Automation and Robotics Conference 2025
