Cutlery Detection System

AK Wrap, a well-established Croatian automation company working with high-end cruise industry clients such as MSC, had a real operational need inside its automated cutlery wrapping workflow. To support different spoon, fork, and knife types within the same process, the system required a more intelligent way to detect, classify, and localize cutlery reliably under changing real-world conditions.

Without machine vision, process variability often leads to additional mechanical complexity, reduced flexibility, and more difficult adaptation when different utensil types, positions, or presentation conditions are introduced. In high-throughput industrial environments, this creates pressure on reliability, increases maintenance burden, and limits how scalable the automation can become over time.

Hartera Robotics developed the Cutlery Detection System as a computer vision and system-integration solution designed for real industrial automation. The platform was built to detect different cutlery types, provide positional information and depth data, and support the broader machine workflow with robust verification logic and fallback handling.

The solution combined computer vision, ROS 2-based system architecture, real-time processing, embedded and computing integration, and machine-level coordination into one structured engineering workflow. Special attention was given not only to detection performance, but also to reliability under operational edge cases, system communication, and practical deployment inside an existing industrial process.

The result was a more capable and adaptable automation workflow that allowed the machine to work with greater awareness, precision, and robustness. By introducing vision-based detection, position estimation, depth-aware logic, and controlled fallback behavior, the system reduced dependency on rigid mechanical-only solutions and improved the machine’s ability to operate under variable conditions.

For Hartera Robotics, this project represented an important step in applying robotics, computer vision, and software-driven automation to a real industrial use case with demanding standards. It also demonstrated how advanced engineering solutions can create measurable value in production environments, especially when developed around practical needs, system reliability, and integration with existing automation platforms.