Artificial Intelligence and logistics at A&T in Torino

AI and logistics are becoming increasingly interconnected thanks to solutions that streamline warehousing, ensure product traceability and integrate quality control processes At the AT&T trade show, ItaliaImballaggio collected some interesting case histories

Battaglio: industrial vision and Quality

Through a project developed by CIM 4.0 (Competence Industry Manufacturing 4.0) and Accenture, Battaglio, a leading player in the import-export of fruit and vegetables, has integrated industrial vision technology to enhance the efficiency of its logistics chain. This innovation accelerates quality control processes and optimises product logistics, from ripening chambers to the end of the line. The new packaging line automates quality control and product conformity, ensuring alignment with customer requirements.
This process is now manually managed by staff, who handle and visually inspect crates weighing more than 20 kg before dispatching them for palletising and final processing. With the new automated solution, production rates, which today handle 300-400 packages per hour, will increase to 1200 pieces per hour. As part of this process, fruits enclosed in transparent wrappers, sometimes made opaque by condensation, are manually inspected by staff to assess ripeness based on colour and determine whether they meet customer specifications.
To enhance efficiency, CIM 4.0 and Accenture have implemented an AI-driven algorithm capable of sorting products by colour, significantly optimising the process. Once up and running, the line will use robots equipped with cameras to handle, inspect, sort and discard products as needed. After inspection, the products are sent to the pulping area, ripening chambers, or final palletisation stages. It is important to note that AI does not replace human operators but instead supports them in cases of uncertainty, such as difficulties related to colour variations, irregular product placement or condensation. This makes the solution more effective than traditional pixel-based control methods.

State-of-the-Art labelling for Venchi

With an annual production of 4 million bars, Venchi is one of Italy's most renowned confectionery brands. Its production lines, particularly the labelling phase for traceability and expiry dates, face critical challenges related to printing accuracy, positioning and legibility. These issues often result in product returns and rejects, affecting both saleability and leading to increased economic and environmental costs. To address these issues, Venchi collaborated with Inventio to develop a quality detection system equipped with AI-powered in-line cameras trained for visual product detection.
The industrial vision system captures an image of each chocolate bar, checking for print compliance, errors or misalignments, and directing defective products to an ejector. Operators then manually intervene, re-introducing these bars into the production line for correct packaging.
AI plays a key role in verifying batch and expiry dates, generating data that, in line with Industry 4.0 principles, is sent to the Manufacturing Execution System (MES) for data-driven process control. The system operates with rapid recognition times, associating data with each product and identifying non-compliant items.