How 3AS is helping to increase your overall equipment effectiveness for lot-size-1 manufacturing

Lot-size-1 manufacturing

The efficient manufacturing of high-quality products for small lot sizes is very challenging due to the lack of history to stabilize the production processes. Consequently, the capability to profile the machine tool for optimized planning and operation already at the first batch is a game changer to increase overall equipment effectiveness (OEE). Existing solutions for machine health and key process attribute monitoring have limited accuracy, are difficult to adapt to process variation, respectively require several production batches to stabilize.

The Challenge

Together with our partners we are aiming to develop an innovation platform for digital transformation relying on data-driven machine tool profiling for lot-size-one production. The generated profile of your machine tool can be used for prediction of your machine operation (e.g., prediction of execution duration during offering process), for quality monitoring and for machine health monitoring.

The platform will consist of three main assets:

  1. an ideation tool for the identification of business value of the customers’ data, and an online data analytics platform for efficient data exploration.
  2. a method for automated generation of an adaptive and accurate profile of your machine tool, integrating average energy consumption, time duration and tool wear information for each NC instruction set.
    1. An innovative data analytics platform (3AS), providing
    2. privacy awareness for management of sensible data,
    3. openness and customizability for ease use of ready-made models and analytics functionality,
    4. extensibility by onboarding data scientists and engineers whenever applicable, and
    5. cost awareness for tailored exploration and deployment at different scales based on SaaS business model.

Problems 3AS is solving in this project

Through efficient scoping, we are empowering the user to identify the business value of their data through efficient integration of all stakeholders. Business specialists, product managers, and IT experts can efficiently contribute with their expertise to identify and specify the appropriate data product(s).

Through the generation of an accurate machine profile, significantly higher monitoring and diagnosis accuracy can be achieved. Furthermore, the model is very fast to set-up (1 measurement is sufficient) and remove all confidentiality issues since the full NC file is completely abstracted during model creation.

Furthermore, the domain expert can now become a data expert as well. Through our privacy-awareness approach, data products relying on confidential and/or private datasets according to regulations in place can be created efficiently.