Quantifying Lot Variation for Enhanced Sustainability - AI for INJ MLDG powered by MAZIN

Advancing Sustainable Manufacturing with Low-Grade Resins.



MAZIN, Inc. (President and CEO: Takanori Kadoya, located in Chuo-ku, Tokyo; hereinafter referred to as "MAZIN"), has begun providing algorithms for quantitatively evaluating the variability of material lots.

In the future, we will actively enhance our efforts to quantitatively evaluate and support the selection of recycled materials, in collaboration with chemical manufacturers and material trading companies.



The increasing interest in sustainability and stricter environmental regulations are accelerating the use of recycled materials. Particularly in the automotive industry, centered around Europe, proposed new regulations are mandating the use of recycled materials, aiming to ensure the availability of high-quality recycled materials and their effective utilization.


Challenges Faced by the Industry

Expanding the use of recycled materials contributes to reducing environmental impact. However, significant variation in quality between batches poses a major challenge. As interest in sustainability continues to grow, securing small, high-grade recycled materials with minimal variation, such as PIR (post-industrial recycled) resins, becomes increasingly difficult.


For both factories molding materials and chemical manufacturers and material trading companies supplying materials, establishing techniques to reliably mold even large-batch variation resins such as post-consumer recycled (PCR) resin is becoming increasingly important for achieving sustainable production.


Our Initiatives

MAZIN is engaged in the development of algorithms that analyze sensing data obtained from manufacturing sites to detect production abnormalities and make adjustments to production conditions.


Rather than focusing on the development of technology for producing higher-quality recycled materials, which major trading companies and chemical manufacturers are pursuing, we are dedicated to supporting improvement activities from a perspective closer to the production site. This includes addressing how to deal with the variation in batches of recycled materials occurring within the injection molding process and improving productivity and defect rates.


The graph above represents clusters formed by our algorithm based on data obtained from molding different batches of the same resin variety. It confirms the formation of distinct clusters for each batch.


In this case, there is a high likelihood that the consistency of molding quality is compromised with each change in batch.


By providing algorithms to quantitatively evaluate the variability between batches, we can support material selection with less batch variation. This assistance aims to facilitate stable molding using recycled materials.


Future Developments

In addition to the above technology, we are developing algorithms aimed at solving various challenges that arise in molding processes, such as estimating resin viscosity. In the future, we will expand our collaboration with injection molding machine manufacturers, sensor manufacturers, chemical manufacturers, and material trading companies both domestically and internationally, based on these technologies.


# About MAZIN, Inc.

- CEO:Takanori Kadoya

- Founded:June 2018

- Location:3-3-6 Nihonbashi Honcho, Chuo-ku, Tokyo, Waka-Sue Building 2F

- Capital JPY:430,616,734 (including capital reserve)

- Business Description:Research, development, and sales of manufacturing AI

- URL:https://www.mazin.tech




| Small | Normal |
| Big | Original |


| Small | Normal |
| Big | Original |


| Small | Normal |
| Big | Original |





  • ※購読している企業の確認や削除はWebプッシュ通知設定画面で行なってください
  • SNSでも最新のプレスリリース情報をいち早く配信中