The Five Phases of Operational Research
Operational research is an interdisciplinary idea that uses thinking and also methods such as mathematical modeling, statistics and algorithms to arrive at optimal or near optimal solutions to complex problems in an organization. In an operational research project there are 5 phases, namely, problem identification, formulating problem, deriving solution validating model and lastly, the implementation of results.
The first phase is to identify the problem through the observation of its symptoms if it is not obvious. Relevant questions to what the problem is about are asked to clear any doubts and the objectives, limitations and requirements of the client are established. In addition, delineation of the subproblem to be studied will be conducted if the problem is too complicated, large or time consuming.
In the second phase, formulation as a mathematical model would take place based on the problem in the previous phase. The choice of the model is crucial to success so a lot of considerations need to be made for the restrictions, limitations and assumptions to be put in place. The presence of the ‘data barrier’ by people trying to block the project or working in a data –poor environment would hinder this formulation process and a trade-off between time and costs should be done in order to obtain the closest information and at the same time, protecting the clients interests.
Model validation, the third phase of the project, involves running the algorithm for the model on the computer. This is to ensure that the input data or computer program is free from errors and correctly represents the model. It is also to check and ensure that the results from the algorithm seem reasonable.
The fourth stage would be to derive the optimal solution of the model through the use of a standard computer package, or specially developed algorithm. Different numeric scenarios will be used to establish sensitivity. In addition, checks should be done to ensure the speed of
The first phase is to identify the problem through the observation of its symptoms if it is not obvious. Relevant questions to what the problem is about are asked to clear any doubts and the objectives, limitations and requirements of the client are established. In addition, delineation of the subproblem to be studied will be conducted if the problem is too complicated, large or time consuming.
In the second phase, formulation as a mathematical model would take place based on the problem in the previous phase. The choice of the model is crucial to success so a lot of considerations need to be made for the restrictions, limitations and assumptions to be put in place. The presence of the ‘data barrier’ by people trying to block the project or working in a data –poor environment would hinder this formulation process and a trade-off between time and costs should be done in order to obtain the closest information and at the same time, protecting the clients interests.
Model validation, the third phase of the project, involves running the algorithm for the model on the computer. This is to ensure that the input data or computer program is free from errors and correctly represents the model. It is also to check and ensure that the results from the algorithm seem reasonable.
The fourth stage would be to derive the optimal solution of the model through the use of a standard computer package, or specially developed algorithm. Different numeric scenarios will be used to establish sensitivity. In addition, checks should be done to ensure the speed of