Risk-Based Optimization of Building Maintenance Sequencing in Multi-Plant Factories Using Genetic Algorithm

Akhmad Firdos Khoiril Khitam (Teknik Sipil, Fakultas Sains dan Teknologi, Universitas Islam Nahdlatul Ulama Jepara)
Mochammad Qomaruddin (Teknik Sipil, Fakultas Sains dan Teknologi, Universitas Islam Nahdlatul Ulama Jepara)
Fatchur Roehman (Transportasi, Fakultas Teknik, Universitas Maritim AMNI)
Bahak Mujtabak Yufid (Teknik Sipil, Fakultas Sains dan Teknologi, Universitas Islam Nahdlatul Ulama Jepara)

DEARSIP: Journal of Architecture and Civil
Vol. 6 No. 01 (2026): Desember 2025 - Mei 2026

Building maintenance management in multi-plant factory environments is a complex and challenging task due to the geographical distribution of facilities, limited maintenance resources, and varying levels of operational risk across plants. In practice, maintenance sequencing decisions are often based on expert judgement, which may lead to subjective and suboptimal outcomes. This study proposes a risk-informed optimization framework to determine the optimal sequence of building maintenance activities in a multi-plant factory by modeling the problem as a travel network optimization task. The objective of the proposed model is to minimize total travel cost associated with mobilizing maintenance equipment and materials, while incorporating risk-based maintenance priorities into the sequencing decision. The proposed approach is demonstrated through a case study of a multi-plant factory located in Kudus, Central Java, Indonesia, consisting of eight plants. Each plant is represented as a node in the travel network, and inter-plant movements are associated with travel costs expressed under 2025 cost conditions. Risk values are assigned to each plant based on expert judgement, considering routine maintenance importance, audit compliance requirements, production impact, and safety and product quality risks. A single-objective optimization model is formulated and solved using a Genetic Algorithm (GA). The results show that the proposed risk-informed GA produces a stable and efficient maintenance sequence that balances travel cost efficiency and risk prioritization. Compared to the initial maintenance plan proposed by the building administrator, the GA-based solution achieves a travel cost reduction of approximately 2.74%, while providing a more systematic and transparent prioritization of high-risk plants. Although the no-risk baseline scenario yields a slightly lower travel cost, it fails to adequately prioritize plants with higher risk levels. These findings indicate that the proposed approach offers a practical decision-support tool for maintenance planning in multi-plant industrial facilities, where modest cost savings combined with improved risk awareness can lead to more defensible and effective maintenance decisions. Future research may extend the model to larger-scale systems, multiple maintenance teams, and dynamic risk conditions.