Explore chapters and articles related to this topic
The Discipline of Systems Engineering
Published in Lory Mitchell Wingate, Systems Engineering for Projects, 2018
Although the processes that ultimately make up systems engineering were practiced in various forms throughout history, the first documented use of an approach that took the system into consideration was in Bell Telephone Laboratories. Their use of “systems engineering” as a term to reflect their methods was documented in the 1940s.1 Bell Laboratories was involved in military action optimization studies during World War II. Scientists and engineers there were using operations research methods, specifically optimization modeling using calculus, linear algebra, and other techniques, as well as stochastic processes such as queuing theory and probability theory. It was not until 1962, however, that “Arthur Hall published his first book on systems engineering: A Methodology for Systems Engineering. Hall was an executive at Bell Laboratories and was one of the people who were responsible for the implementation of systems engineering at the company.”2 During the same era, the RAND Corporation, ideated by a newly formed United States Air Force, developed a process for systems analysis that would become an important part of the systems engineering discipline. Systems analysis is an approach that reviews a problem in logical steps, and describes the system thoroughly and explicitly. Using computing resources to perform systems analysis and optimization modeling provides a solid scientifically based approach for performing systems engineering.
General introduction
Published in Adedeji B. Badiru, Handbook of Industrial and Systems Engineering, 2013
Modeling the logistics of construction present interesting challenges which include crew scheduling under uncertainty (use of union vs. nonunion workers, regular time and overtime assignments, multishift and rolling shift, etc.). Many areas such as supply chain, transportation, material dispatching and inventory planning and control may be modeled and solved by operation research tools. The current tools for dealing with uncertainty in project management are overly simplistic and are dependent on unrealistic assumptions, such as the independence of completion times for each activity. More robust methods to analyze the distribution of completion times for large-scale construction projects are needed. Furthermore many issues related to automation, such as robot path planning, collision avoidance, image and signal processing, etc., would require OR modeling.
Introduction to Optimization Models
Published in William P. Fox, Nonlinear Optimization, 2020
The optimum-seeking methods are also known as mathematical programming techniques (specifically, nonlinear programming techniques) and are generally studied as part of operations research or applied mathematics. Operations research is a branch of mathematics that is concerned with the application of scientific methods and techniques to decision-making problems, and with establishing the best or optimal solutions. Table 1.1 lists various mathematical techniques used in the areas of operations research. As operations research is ever evolving, the list is always growing.
Understanding and addressing complexity in problem solving
Published in Quality Engineering, 2021
Roger Hoerl, Willis Jensen, Jeroen de Mast
Anderson-Cook (2017) provided some specific tools, such as Pareto fronts, to be able to address this complexity in decision making. If we can rigorously define the criteria that matter and collect the right data, then these tools can be quite valuable. Operations research (OR) provides another set of decision-making tools that can be helpful, such as discrete event simulation, queueing theory, and linear and quadratic programming. OR and Business Analytics discern several complexity classes of decision and optimization problems, such as P, NP and NP-hard. These classes characterize whether the problem’s complexity allows an algorithm to solve it in polynomial time (P), or to verify a potential solution in polynomial time (NP, for nondeterministic polynomial time). NP-hard is the class of problems that are even more complex than NP (Arora and Barak 2009).
Application areas and antecedents of automation in logistics and supply chain management: a conceptual framework
Published in Supply Chain Forum: An International Journal, 2021
Benjamin Nitsche, Frank Straube, Maximilian Wirth
By reading those definitions and understandings it becomes obvious that replacing a task performed by a human being with a machine or a computer is the focus of automation. During the early stages of automation in production this mainly meant that a machine is replacing or supporting a human being to fulfill the task more efficiently and safely, but also to perform a task that the human could not handle (e.g. lifting heavy components). Due to the advancing industrialisation and digitalisation, the aspect of supporting tasks that humans alone cannot solve becomes even more important. In modern supply chains, intelligent algorithms support decision making and make problems solvable that would be too complex for humans alone. Therefore, operations research is an integral part of research in the field of logistics and supply chain automation since it addresses the support of very complex decisions by means of advanced analytics.
Data-driven stochastic optimization approaches to determine decision thresholds for risk estimation models
Published in IISE Transactions, 2020
Gian-Gabriel P. Garcia, Mariel S. Lavieri, Ruiwei Jiang, Michael A. McCrea, Thomas W. McAllister, Steven P. Broglio
Operations research has also been applied to medical diagnosis decisions, where such problems typically optimize pre-diagnosis decisions or follow-up decisions after an initial diagnostic test. For example, Bayati et al. (2018) determine the least-cost set of biomarker tests that allow for sufficient diagnostic power while Ayvaci et al. (2012) and Zhang et al. (2012) optimize biopsy follow-up decisions for cancer. However, our work focuses on the actual diagnosis decision at hand instead of pre- or post-diagnosis decisions. To this end, Ayvaci et al. (2017) and Ahsen et al. (2019) study when and how bias-inducing information should be incorporated in breast cancer diagnostic decisions. Similar to Ahsen et al. (2019), we also study the incorporation of clinical decision support systems in diagnostic decisions. However, they focus on the design of such systems whereas we focus on the interpretation of these decision support systems, i.e., a risk estimation model. Furthermore, their work focuses on balancing two sources of information (i.e., mammogram risk and clinical-risk information) and deriving one diagnostic threshold, whereas our work focuses on a single source of information (i.e., risk estimates) and deriving two diagnostic decision thresholds.