SACMI Imola is leveraging artificial intelligence (AI) to redefine quality inspection in PET preform manufacturing, with a new generation of vision systems designed to detect defects more accurately, simplify operations and minimize production downtime, according to Massimo Gani, rigid packaging business unit, SACMI Imola.
Speaking at the 11th Injection, Blow Moulding, PET & Recycling International Business Summit & Exhibition, organized by ElitePlus Business Services in Mumbai, Gani said AI is transforming quality control from a simple inspection process into an intelligent production management tool capable of identifying defects, anticipating problems and supporting operators in real time.
Addressing delegates, Gani said it was his fourth or fifth visit to the conference and his third as a speaker. He describes the event as an important platform for engaging with India’s plastics processing industry.
He outlined SACMI’s global presence, noting that the company, founded 107 years ago by nine people, has grown into a multinational group with around 70 companies and nearly 5,000 employees worldwide. The group serves multiple industries, including ceramics, chocolate processing, packaging and rigid packaging, with operations in India through its Mumbai office.
Within its rigid packaging division, SACMI manufactures complete turnkey production lines for plastic and metal closures, plastic containers and PET preform. Gani said the company’s vision inspection business has evolved over nearly 35 years, developing solutions for a wide range of packaging applications.
A key highlight of the presentation was SACMI’s latest automatic PET preform inspection system, designed to inspect every critical component of a preform, including the top seal, thread and body. The system also performs weight verification, optical character recognition (OCR) and AI-powered inspection using polarized light.
“The most important feature is the fully automatic format change,” Gani said. “There is no need for manual adjustment, which significantly reduces the possibility of human error.”
Gani said SACMI’s AI platform comprises three integrated tools. The first, CVS Suite, provides a simplified operator interface that allows production settings to be changed using a single parameter, reducing setup time and making the system accessible even to non-expert operators.
The second tool, Classy AI, applies deep-learning algorithms to classify defects instead of merely identifying them as acceptable or defective.
“It is no longer just a black spot,” Gani said. “The system recognizes the actual defect, helping manufacturers understand the root cause and improve product quality rather than simply rejecting defective parts.”
The third AI-based application, Gilda-AI, functions as an intelligent virtual assistant embedded within the inspection system. Operating without an internet connection to meet cybersecurity requirements, Gilda analyses operating manuals and provides real-time guidance whenever alarms or operational issues arise.
“When an alarm appears, Gilda opens a chat interface and explains in simple language what has happened and why,” Gani explained.
He said all of SACMI’s software and AI algorithms are developed in-house, enabling the company to continuously update its systems without relying on third-party software providers.
Concluding his presentation, Gani said the objective was not to create increasingly complex inspection equipment but to simplify quality control for manufacturers.
“Complexity is easy. What is harder is making complexity easy,” he said, adding that SACMI’s goal is to deliver intelligent quality control solutions that help customers reduce defects, improve efficiency and enhance production quality.
