Process Architecture as a Neural Network in Digital Manufacturing. Main Characteristics, Elements and Metrics

The article considers the analogy between process architecture in digital production and neural networks. The author details the main characteristics, elements and metrics by which parallels can be drawn between the two concepts. It is shown that process architecture, like neural networks, has the properties of interconnectivity, adaptability and hierarchy. Analogies between the inputs/outputs, transformations, connections and control of processes and the corresponding elements of neural networks are examined. Key metrics such as performance, quality, flexibility and robustness are also compared. The paper then presents analogies between business process modelling and neural modelling, and between process architecture modelling and neural network design.