Deepack and AVK Plast have entered a strategic collaboration to implement next-generation AI-based quality inspection in industrial plastics manufacturing. The partnership is designed to explore how continuous, self-improving computer vision systems can support AVK Plast’s high standards for reliability and traceability.
Introduction
Deepack is an innovative AI platform built to automate visual inspection and quality control on production lines. Its proprietary model is designed specifically for manufacturing environments where flexibility, speed, and precision are critical. Unlike traditional vision systems that rely on manual configuration, Deepack’s AI continuously learns and adapts to defect patterns in real time.
AVK Plast is part of the global AVK Group and serves as a key supplier of high-performance plastic components for pallets, waste containers and climate products with special focus on recycled materials. Known for their precision molding capabilities and rigorous quality demands, AVK Plast operates at the intersection of innovation, consistency, and compliance.
Why This Partnership Matters
AVK Plast initiated this collaboration with Deepack as part of a broader effort to optimize quality assurance, reduce manual inspection burdens, and ensure documentation that supports traceability and audit requirements.
The pilot will focus on a specific production line where Deepack’s AI model will be deployed to inspect plastic components in real-time. The aim is to validate the system’s ability to detect both common and rare defect types, automatically classify them, and generate high-resolution data for process improvement and traceability.
This project is a vital real-world validation case and a unique opportunity to tailor the technology to the specific operational needs of AVK Plast.
The Technology and Its Value
Deepack’s AI model is developed in-house and trained specifically on manufacturing data. Key benefits of the solution include:
- Automatic defect detection and classification: Including previously unseen or unclassified issues.
- Real-time quality data: Enabling full transparency across shifts, batches, and product variants.
- Operator-assisted learning: The system improves with every labeled example and operator confirmation, building a robust dataset over time.
This approach not only streamlines visual inspection but also contributes to reducing false positives, eliminating process drift, and creating a unified quality reporting layer across AVK Plast’s production flow on the machine.
Statements from the Teams
“Collaborating with AVK Plast is a major milestone for us. Their commitment to quality, lean processes, and traceability makes them an ideal partner to validate and shape Deepack’s solution for industrial-scale deployment.”
— Carl Fagerlund, Co-founder, Deepack
Future Outlook
If successful, the pilot may pave the way for a broader rollout across additional lines or facilities within AVK Plast’s operations. The collaboration could also serve as a foundation for international scaling and deeper integration with other AI quality initiatives.
For Deepack, this project is not only a commercial milestone but a step toward shaping the future of visual quality assurance—one that combines human insight with self-learning AI to drive manufacturing excellence.
Stay tuned as the partnership evolves and helps redefine quality control in plastics production.
