| dc.description.abstract | This thesis investigates the question, “How can manufacturing companies effectively handle
events that are unavoidable in real life uncertain and complex manufacturing environments?”
within the context of production scheduling and control. The research was initiated as a part
of an 8-year project called NORMAN (Norwegian Manufacturing Future),
www.sfinorman.no, and supported by two other ongoing research projects called SoundChain
and MIX with industrial partners. The PhD study took place as a part of the NORMAN task
addressing intelligent handling of events in production control. This task aimed to analyze
real-life production control practices, and develop decision support models and systems for
multi-level real time control of plants and supply chains.
The motivation of the PhD study stems from the challenges created by events that are
unforeseen during the initial production planning and scheduling decisions. Events are
caused by situations such as material shortages, machine breakdowns, rush orders, quality
problems and due date changes, when executing and controlling the production operations.
These causes can be categorized into resource-related causes (e.g. machine breakdowns) and
order-related causes (e.g. order priority changes). Some efforts are made during the initial
planning and scheduling decisions to mitigate against events, such as keeping safety stocks.
However, real life schedulers still use a great deal of effort to monitor and handle events in
production execution.
“Handling events” refers to a decision making process that consists of identification of the
event, evaluation of possible reactions, and choosing the most appropriate action. The
decisions taken in response to events are referred as “reactive control decisions”. Effective
reactive control decisions for events are critical to avoid unnecessary plan deviations and
keep the overall performance high. High performance means keeping commitments to
customers while sustaining operational efficiency, which together builds the competitiveness
of the firm.
Handling events in production scheduling and control is a well-studied research topic. Many
methods, optimization algorithms, and tools have been developed and proposed in the
scientific literature and commercial systems. However, the majority of these theoretical
developments are based on simplified and well-defined models of the real life production
scheduling and control practice, which in reality is much more complex and ill defined. The
real life production scheduling and control practice is much broader than an optimization
problem, incorporating other aspects such as iterative information sharing practices between
different parties in the organization. Many scholars recognize this issue as one of the main
reasons for the gap between the theory and practice of production scheduling, leading to
implementation challenges when attempting to apply theoretical developments in real life
practice.
The scope of this research encompasses these issues with the aim to contribute to better
understanding of how events can be handled effectively in real life production scheduling and
control practice. The study is built on the hypothesis that the design and development of
decision support principles and systems for production scheduling and control should
consider the dynamics of event handling in real life practice. This hypothesis was inspired by
the pioneering studies of McKay and Wiers (1999) and Aytug et al. (2005) that discussed the
gap between the theory and practice of production scheduling. Advancing the knowledge on
how events are handled and what aspects of real life practice are involved is crucial for
bridging the gap between theory and practice. In turn, this will improve the implementation of
the decision support principles and systems in real life uncertain and complex manufacturing
environments.
The research was conducted primarily through the case study methodology, involving four
manufacturing companies. Two of the companies were members of the NORMAN consortium
and two were partners of SoundChain and MIX projects. Initial studies within the two
NORMAN companies provided in depth understanding of the decision making process and
interactions that take place when handling events in real life production scheduling and
control practice. Research further focused on the real life production scheduling and control
aspects that influence the effectiveness of the decision making process for event handling,
through a literature study. The combined findings were then tested and evaluated in all four
Norwegian manufacturing companies. Finally, the research addressed the implications of the
real life aspects on designing decision support systems for production scheduling and control
practice. These implications led to the generation of guidelines that were used to develop and
propose prototypes of decision support systems for the two NORMAN companies involved in this study.
This research contributes both to theory and to practice by increasing the knowledge about
the decision making process in real life production scheduling and control practice as well as
the knowledge on decision support requirements. The findings of this research can be applied
further to develop more useful decision support principles and systems, addressing the
requirements of real life practice.
The main contributions of this thesis are summarized as follows.
• A task model for event-driven production control, which describes the stages that are
taken to classify and handle different types of events in real life practice.
• A decision making framework for event handling in production control, which
classifies the decision making patterns when handling events in production control.
The framework can guide the development of event handling principles better suited to
the decision making patterns and behavior of real life decision makers.
• An organizational maturity model for effective production scheduling and control
practice, which addresses the needs of an organizational environment for establishing
and performing effective decisions in production scheduling and control. The model
can be utilized to evaluate the effectiveness of the production scheduling and control
process in real life practice, as an initial step in the improvement efforts.
The auxiliary contributions of the thesis are summarized as follows.
• A framework for characterizing the events, which can be utilized to model the events
in production scheduling and control.
• A model for categorizing the events based on their potential impact, which can be
utilized in development of event specific principles and algorithms for taking reactive
control decisions
• A set of guidelines for design and development of a practical decision support system,
which analyses and addresses the requirements of real life production scheduling and
control practice.
• Real life case studies on the topic of production scheduling and control, that provided
additional insights and field-based learnings to the literature
The findings and contributions of this thesis imply the following issues to be carefully
considered in order to provide more relevant contributions from the production scheduling
theory to real life practice:
• Prior to developing principles and tools for production scheduling and control, there
is need for a thorough understanding of the production scheduling and control
practice in a given context,
• The real life production scheduling and control aspects that influence the
coordination and information sharing efforts when making effectiveness-oriented
reactive control decisions are often unique and deserve thorough analysis,
• The scheduling context and decision support systems should be modelled and designed
in accordance with the identified aspects of the real life production scheduling and
control practice.
Overall, this PhD project has achieved its goals by contributing to broader understanding of
the real life production scheduling and control practice which is significantly influenced by
the real time events. This thesis aims to be helpful to the future work on designing and
developing decision support models and systems better suited with the real life production
scheduling and control practice.• The scheduling context and decision support systems should be modelled and designed
in accordance with the identified aspects of the real life production scheduling and
control practice.
Overall, this PhD project has achieved its goals by contributing to broader understanding of
the real life production scheduling and control practice which is significantly influenced by
the real time events. This thesis aims to be helpful to the future work on designing and
developing decision support models and systems better suited with the real life production
scheduling and control practice. | nb_NO |
