Co-Designing an AI Factory Flow Manager for On-demand Apparel Manufacturing with Garment Workers

**商业需求***模式项目致力于将全球时尚产业转变为按需无库存模式。*我们的硬件和软件管道是为“拉动”动态模型而设计的，在这种模型中，生产是基于客户需求的自动化工厂流程。服装是个性化的或定制的，员工们接受过5件小批量的培训。*最困难的挑战是管理每周按需生产的机械师工作流程。我们目前为5个品牌生产12款，每周都有新品牌推出。每种风格都有9 - 30个独立的步骤，使用2 - 5种不同的机器。每周分批生产，每个款式的数量可以在1 - 50件之间变化。*我们根据订单队列手动安排生产花名册，并确定给定项目列表及其数量，并涉及机械师最少的切换时间。**新理念-人工智能工厂流程管理器**我们希望为按需服装制造创建一个人工智能工厂流程管理器，可以优化各种风格的生产过程。该人工智能将能够针对以下变量进行优化：*物流因素-服装批次的大小，期望的批次交货日期，机器设置和切换时间*服装风格因素-接缝完成的不同风格之间的协同作用*服装工人因素-机械师当前的技能组合，单个机械师的学习阶梯，机械师的相对速度，机械师当前的产品技能组合与生产线的弹性有关。质量生活，有意义的工作，应对挑战**效益***通过提高整体质量和生产力，使按需生产具有成本效益，为管理层服务。*让服装工人参与设计和控制工作，提高他们的积极性和工作满意度，从而使他们受益。*服务于整个组织的利益-通过建立一个正式的系统来鼓励学习，捕捉和交流方法的改进。灵感-[NUMMI植物过渡][0]。**EDI挑战**女性占全球服装工人总数的80%。尽管如此，男性仍然主导着制造技术/系统的发明和设计。这是令人担忧的，因为生产力软件、自动化和机器人的使用直接影响着服装工人的生活体验和他们的就业前景。**EDI的影响***通过充满活力和合作的文化，培训融入日常工作，提高服装工人的生活质量。*按需定制生产可以使劳动密集型行业的利润翻倍，从而保障和创造更好的就业机会。[0]: https://sloanreview.mit.edu/article/how-to-change-a-culture-lessons-from-nummi/

**Business Need*** Pattern Project is focused on transitioning the global fashion industry to an on-demand no-stock model.* Our hardware and software pipeline is designed for the "pull" dynamic model, in which production is based on customer demand with an automated factory flow. Garments are personalised or custom-fit, and staff is trained on micro-batches of five.* The most difficult challenge is managing the workflow of machinists for weekly on-demand production. We currently produce 12 styles for five brands and welcome new brands each week. Each style has between 9 - 30 individual steps and employs 2 - 5 different machines. Production is batched weekly, and quantity per style can vary between 1 - 50 pcs.* We manually schedule a production roster based on the order queue and determine, given a list of items and their quantities and involving the least switchover time for machinists.**New idea - AI Factory Flow Manager**We want to create an artificial intelligence factory flow manager for on-demand apparel manufacturing that can optimise the production process for various styles. This AI will be capable of optimising for the following variables:* Logistic factors - Size of garment batch, Desired date of delivery of the batch, Machine set up and switchover times* Garment style factors - Synergies across styles for seam finishes* Garment Workers factors - The current skillset of the machinist, The learning ladder of individual machinists, Relative speeds of machinists, Machinists' current product skillset concerning resilience in the production line, Quality of life and meaningful work, Response to Challenge**Benefits*** Serves management by improving overall quality and productivity and making on-demand production cost-effective.* Benefits garment workers by involving them in the design and control of their work, increasing their motivation and job satisfaction.* Serves the interests of the entire organisation - by creating a formal system to encourage learning and capture and communicate improvements in methods. Inspiration -[NUMMI plant transition][0].**EDI Challenge**Women account for 80% of all garment workers worldwide. Still, men continue to dominate the invention and design of manufacturing technologies/systems. This is concerning since productivity software, automation, and the use of robotics directly impact the lived experience of garment workers and their employment prospects.**EDI Impact*** Better quality of life for garment workers through a dynamic and cooperative culture, training integrated into daily work.* On-demand custom-fit production can double margins for the labour-intensive industry - safeguarding and creating better jobs.[0]: https://sloanreview.mit.edu/article/how-to-change-a-culture-lessons-from-nummi/