Maintenance work order assignment optimization

Sponsored by the office of Facilities Management (FM) in Georgia Tech

Research Motivation

The office of Facilities Management (FM) in Georgia Tech manages the overall Operation and Maintenance (O&M) of campus buildings. The FM department sets the procedures of creating daily work events and assigns the procedures into workforce. Unfortunately, It has been observed that the process of assigning work to facilities management employees is not systematical, which leads delay in overall process.

Key Idea

To perform work order optimization, we propose the following methodology:

1. Compile real datasets (e.g. working time) from the office of Facilities Management

2. Model the real scenario with DES and validate it with actual total working hours

3. Formulate the Traveling Salesman Problem (TSP)

4. Employ Genetic Algorithm (GA) and Simulated Annealing (SA) to solve the TSP

5. Perform DES with new (or optimized) work orders and calculate total working hours

6. Compare the optimum scenario with the baseline in terms of total working hours

Results

The results show that the optimized scenario reduces the total working hours by approximately 3.5% relative to the real scenario.

Publication

Presented in the Conference of Computational Interdisciplinary Science in 2019