Forecasting of Store Maintenance in Retail Chain Operation with Neural Network Method: The Case of True Corporation Public Company Limited


  • สุกิจ สวัสดิ์แสงสน Management, School of Management, Shinawatra University
  • Eksiri Niyomsilp Management, School of Management, Shinawatra University


Artificial Intelligence, Neural Network, Forecasting Model, Retail Chain Operation


This research aimed to study the repair data within the maintenance department of True Corporation Public Company Limited by using the repair period as a variable and measurement. The 236 branches and 3,103 total repair alerts in 2017 and 2018 were collected as the research data to perform the mathematical process in the form of artificial intelligence (AI: Neural network) in order to predict the repair in the future.

The process used in this research was based on the Neural Network, a basic model of machine learning that is popular in many industries. Most of them are used in forecasting and classified, including integrating with Image processing technology. By creating a single-layer Perception feed-forward neural network model using two large amounts of data from data collected from repair notifications of 3,103 repair notifications; namely, the date and time of the repair notification, the number of repair notifications comes through a mathematical process based on function x plus the randomized weight and through the training of the AI ​​process to get results that are close to the actual values.

The results of this research may be used as a guideline for significant improvements and efficiency in stores in the maintenance process so that management can be made in terms of budget preparation, maintenance management costs each year. When adjust properly, it can adjust maintenance strategies from a passive strategy to an active strategy and can prevent preventive maintenance more efficiently.