Industrial Vision and AI Partnership

Jean-Michel Paré, Deputy General Manager at Machinex, oversees the development of cutting-edge technology solutions for the recycling and waste management industry.
Machinex designs, manufactures, and integrates high-performance sorting systems, leveraging artificial intelligence to enhance accuracy, speed, and adaptability in industrial environments.

Machinex wanted to identify and implement the most effective vision models to detect and classify materials along sorting lines.
The goal was to boost detection accuracy, robustness under real-world conditions, and adaptability to evolving material streams.

From the start, we established a continuous research and experimentation process built around:

• weekly iterations between Machinex and Updata’s R&D teams;

• selection, training, and benchmarking of computer vision models under realistic conditions;

• development of internal tools for data management, annotation, visualization, model training, and performance measurement;

• documentation and knowledge transfer to ensure long-term autonomy for the Machinex team.

This agile collaboration has fostered an innovation-driven environment, where prototypes are tested, compared, and improved continuously.

• Collaboration in establishing a complete research, training, and evaluation pipeline for industrial vision models.

• Integration of higher-performing and more robust models into sorting systems.

• Implementation of a data management strategy where Machinex continuously enriches its image dataset, while Updata analyzes and provides recommendations to improve model quality and relevance.

• Contribution to improving Machinex’s innovation cycle through structured methodologies and tailored tools.

• A long-term partnership built on trust, transparency, and technical excellence.