Operator Challenge Response

Navy Special Warfare is looking to improve the way they conduct resource allocation. Current allocators cannot compare their needs across departments, research, and military equipment. Our proposal includes a two-pronged approach to optimize future utility for the military force. Data on consumer preferences is gathered directly from operators by a simple survey then compiled into utility parameters. These parameters are then input into a large linear programming model that uses Kuhn-Tucker conditions and the simplex method to optimize utility. The survey will be reusable each quarter, as will the linear programming model once fresh data is acquired. This allows for allocators to receive complex technical insights before making their decisions, without requiring them to painstakingly perform it.

NSW currently conducts their resource allocation at quarterly meetings with various department heads. These meetings currently take 2-3 hours and scheduling is difficult because of the status of the meeting attendees. Additionally, the department heads, who are allocating resources to the factor of millions of dollars, have little input from operators and the rest of the chain of command as to which resources to prioritize. Lastly, these department heads all compete against each other for a limited budget. Prioritization is challenging because attendees need to determine how to adequately compare research, development, and a wide range of military equipment.

Our current solution is a two-pronged approach of changing the meeting structure as well as providing data-driven analysis to help department heads make more informed decisions to optimize future utility for the force.

First, our plan is to administer a survey to gauge operator preferences then use those preferences to create utility parameters. The survey will consist of a randomized list of items for each operator to choose between with prices and a budget. It’s just a short test to see how they would spend money. It is critical that the survey mimics the resource allocation decisions that department heads face, while also simple enough for operators to express their true preferences.

After collecting data we plan to use Kuhn Tucker conditions and the Simplex method to optimize total system utility subject to budget constraints. Once we create a base equation, we will implement it into an independent program in python to run off the survey data. This will allow it to be reusable between quarters.

Lastly, we plan to present this analysis to department heads so that they can make more informed decisions.

Current road blocks we are facing include our lack of experience with research survey creation as well as the level of linear programming knowledge required to perform significant analysis.