Optimization of training programs for competency-based training - COMFORT model description

FFI-Report 2020

About the publication

Report number

20/00002

ISBN

978-82-464-3250-2

Format

PDF-document

Size

1.1 MB

Language

English

Download publication
Eirik Løhaugen Fjærbu
The Royal Norwegian Air Force (RNoAF) is in the process of phasing in new F-35 combat aircraft, which will take on a wide range of roles. Efficient pilot training is crucial in order to make full use of this investment. Simulator training will be an integrated part of the training program for the F-35 pilots. The simulators provide a means to better adapt the training to the needs of the pilots. One possible approach to improve training efficiency is competency-based training, where training is planned according to competency requirements rather than requiring pilots to complete specific missions. Competency-based training is a promising approach that may improve the training outcome and reduce costs. At the same time, the scheduling and resource allocation can be challenging due to the high degree of flexibility required. In this report, we make use of optimization algorithms to assess these challenges. We consider a regime where the set of missions that the pilots can carry out and the set of required competencies are given. Our aim is to minimize the total cost of the pilot training according to a given cost function, while ensuring that the pilots train each competency a given number of times. Two cases are considered: one where the training outcome of each mission is fixed and one where it can vary between different repetitions of the same mission. Both problems are formulated as Constrained Optimization Problems (COPs). For the case where the training outcome varies, we compare two equivalent formulations. We implement a Java program named COmpetency-Mission Frequency Optimizer for Readiness Training (COMFORT) that solves the COPs using the software package OR-Tools. OR-Tools includes implementations of three different COP solution algorithms. We compare the three solution algorithms in terms of computation time for a set of example parameters and highlight some of their general strengths and weaknesses.

About the publication

Report number

20/00002

ISBN

978-82-464-3250-2

Format

PDF-document

Size

1.1 MB

Language

English

Download publication

Newly published