Learning of Quartering in Digital Control of Refit

According to Industry 4.0 concept, digital industrial control is based on intelligent solutions such as Machine Learning, which is at the crossroads of control science and computer science. For digital control in large company, along with the use of machine learning, it is necessary to take into account the human factor. The paper examines the theoretical and practical issues that arise when applying machine learning technology with an edifier for estimation and management in a large industrial company. The model of learning bifurcation with the help of the edifier determines 2 alternative estimates of the random situation in the company. The integration of two such bifurcation models using systems engineering is proposed for learning quartering for company management. Incentives for company employees are provided on a learning quartering basis. In an environment of uncertainty, it is assumed that staff knows production capabilities better than management. Using this insight, employees can manipulate their activities to gain more incentives. Such unwanted activity can lead to the failure to use existing opportunities in which the company is interested. To solve this problem in the face of uncertainty, a mechanism for digital control of a two-level organizational system is proposed. This mechanism includes learning quartering with the help of an edifier and stimulation. Sufficient conditions have been found for the synthesis of such a mechanism, in which the staff makes full use of the available capabilities. The use of such a mechanism for the digital industry is illustrated by the example of control over the refit of railway locomotives.