AI-powered optical sorting machines for food and industrial products
Optimum Sorting manufactures optical sorting hardware using deep learning (TensorFlow, PyTorch, YOLO, Vision Transformers) embedded in production-line systems. The tech stack—Siemens PLCs, OPC UA, HALCON vision libraries, C++/Python—reveals a hardware-software company, not a pure software vendor. Engineering dominates hiring (12 of 16 open roles), skewed toward mid-level talent, indicating they're scaling production and vision-system integration rather than hunting senior architects. Internal pain points (inventory minimization, production delays, planning bottlenecks, documentation friction) suggest they're managing rapid hardware scaling while wrestling with the operational overhead of custom vision deployments.
Optimum Sorting designs and manufactures optical sorting machines (Novus, Ventus, Magnus product lines) that use AI-enhanced vision to inspect and separate food and non-food products. The company operates across frozen foods, fresh produce, pet food, confectionery, and minerals. Founded in 2017 with roots tracing to earlier optical-sorting work, Optimum Sorting now employs roughly 100 people distributed across Belgium (headquarters in Hasselt), the Netherlands, Thailand, and the United States, with 24/7 global support. Core technical work involves building vision systems that integrate into customer production lines, training defect-detection models, and optimizing hardware for variable real-world conditions.
Windows and Linux hosts; Siemens S7 PLCs; Microsoft 365; deep learning frameworks (TensorFlow, PyTorch, YOLO, Vision Transformers); HALCON vision library; OPC UA and Ethernet/IP for machine communication; C++ and Python for core development.
Three optical sorting machine lines: Novus, Ventus, and Magnus. Each handles frozen/fresh foods, pet food, candy, nuts, seafood, potatoes, and minerals using embedded AI vision for defect detection and product separation.
Other companies in the same industry, closest in size