MAS

This project aimed to modernize the manual embellishment tracking system at MAS Linea Aqua Hanwella. To initiate the project, staff members and a couple of students met with Linea Aqua's IT department at their Hanwella branch. This meeting was crucial in understanding the challenges faced in the current process, such as manual tracking of embellishment demands, supplier bookings, cut readiness, and package statuses, leading to delays and inefficiencies.

To address these challenges, a comprehensive digital system was developed to automate these processes. The system allows for uploading demand plans, booking supplier capacities, tracking cut readiness, monitoring supplier processes, and notifying stakeholders about package statuses. This new system improves efficiency, accuracy, and communication, reducing delays and manual errors.

 An agile methodology was followed throughout the project, ensuring flexibility and adaptability to evolving requirements. Regular meetings with Linea Aqua's IT team and key personnel provided invaluable insights and feedback, ensuring the project's success and alignment with Linea Aqua's needs.

Apparel companies manually operate the garment finishing process in a sample room. When experienced apparel technicians migrate abroad, inexperienced newcomers struggle to analyze sketches to determine optimal construction methods. Manual interpretation takes extensive time, which leads to higher sampling expenses due to missteps.

MAS consulted DIM on automating their sample room process. DIM developed an AI ensemble leveraging computer vision and large language models (LLMs) to annotate sketches and suggest sewing sequences from drawings to slash costs. By training on past image-operation pairings, the model reliably recommends stitch types, tools and steps for new sketches. This intelligence augmentation assists in upskilling novices while increasing development efficiency. It also decreased sampling costs by 30% and duration by 25% over manual approaches in trials. Experiencing iterations with automated validations enables agile adjustments that are closer to market needs.