Ensuring the Robustness and Reliability of Data-Driven Knowledge Discovery Models in Production and Manufacturing.

The Cross-Industry Standard Process for Data Mining (CRISP-DM) is a widely accepted framework in production and manufacturing. This data-driven knowledge discovery framework provides an orderly partition of the often complex data mining processes to ensure a practical implementation of data analytics and machine learning models. However, the practical application of robust industry-specific data-driven knowledge discovery models faces multiple data- and model development-related issues. These issues need to be carefully addressed by allowing a flexible, customized and industry-specific knowledge discovery framework. For this reason, extensions of CRISP-DM are needed. In this paper, we provide a detailed review of CRISP-DM and summarize extensions of this model into a novel framework we call Generalized Cross-Industry Standard Process for Data Science (GCRISP-DS). This framework is designed to allow dynamic interactions between different phases to adequately address data- and model-related issues for achieving robustness. Furthermore, it emphasizes also the need for a detailed business understanding and the interdependencies with the developed models and data quality for fulfilling higher business objectives. Overall, such a customizable GCRISP-DS framework provides an enhancement for model improvements and reusability by minimizing robustness-issues.

Periodical capacity setting methods for make-to-order multi-machine production systems.

The paper presents different periodical capacity setting methods for make-to-order, multi-machine production systems with stochastic customer required lead times and stochastic processing times to improve service level and tardiness. These methods are developed as decision support when capacity flexibility exists, such as, a certain range of possible working hours a week for example. The methods differ in the amount of information used whereby all are based on the cumulated capacity demand at each machine. In a simulation study the methods' impact on service level and tardiness is compared to a constant provided capacity for a single and a multi-machine setting. It is shown that the tested capacity setting methods can lead to an increase in service level and a decrease in average tardiness in comparison to a constant provided capacity. The methods using information on processing time and customer required lead time distribution perform best. The results found in this paper can help practitioners to make efficient use of their flexible capacity.

Influence of dispatching rules on average production lead time for multi-stage production systems.

In this paper the influence of different dispatching rules on the average production lead time is investigated. Two theorems based on covariance between processing time and production lead time are formulated and proved theoretically. Theorem 1 links the average production lead time to the "processing time weighted production lead time" for the multi-stage production systems analytically. The influence of different dispatching rules on average lead time, which is well known from simulation and empirical studies, can be proved theoretically in Theorem 2 for a single stage production system. A simulation study is conducted to gain more insight into the influence of dispatching rules on average production lead time in a multi-stage production system. We find that the "processing time weighted average production lead time" for a multi-stage production system is not invariant of the applied dispatching rule and can be used as a dispatching rule independent indicator for single-stage production systems.