Optimize Billing Routes for Distance Minimization Using Qaco: Qgis Mapping and ANT Colony Optimization (ACO) Algorithms

Abstract

The acceleration of PLN’s cash-in performance is significantly influenced by customers’ early payment behavior. Therefore, PLN requires a solution to shift the payment pattern of customers who frequently delay payments into making timely payments before the due date. One of the main causes of payment delays is the limited availability of payment counters, as well as customers forgetting to pay their electricity bills. This study analyzes customer electricity payment patterns over a six-month period at PLN ULP X, which is part of PLN UP3 XYZ. The analysis revealed that 7.87% of customers are recurring late payers, dispersed across various locations. To address this issue, a preventive collection strategy is required to reduce the occurrence of late payments. However, the implementation of preventive collection faces time constraints, as billman officers have limited availability due to other responsibilities. Thus, a supporting tool in the form of route planning is needed to minimize travel distance. In designing the collection route, the Traveling Salesman Problem (TSP) model is applied using geographic mapping through QGIS and optimized with the Ant Colony Optimization (ACO) algorithm. QGIS is used to map the distribution of late-paying customers based on geographic conditions. The study results show that the ACO algorithm successfully reduced the total travel distance to 1,022 km, compared to the existing condition of 4,300 km, resulting in a distance saving of 3,278 km (76.23%).