Road infrastructure performance is closely associated with passengers and freight transportation systems and socio-economic development. The performance of road infrastructure is commonly measured by sensor-monitored indicators, and the ability of monitored indicators in revealing actual performance is generally determined by decision makers and road users. However, it is usually unreliable to directly apply monitored indicators in road performance evaluation, due to the limited aspects of individual sensor-monitored indicators, and potential biases and uncertainties of human experience. To address the issues, this study proposes a model-driven fuzzy spatial multi-criteria decision making (MFSD) approach to derive a comprehensive and accurate indicator of sustainable road performance. In this study, the MFSD approach is applied in exploring the road network in the Wheatbelt region in Western Australia, Australia. Spatial variables of road properties, traffic vehicles and climate conditions are used as criteria in the decision making. Four sensor monitored indicators are collected for estimating contributions of criteria. Results show that the MFSD-based indicator can more comprehensively and accurately characterize sustainable road infrastructure performance. In the study area, the MFSD-based indicator can improve 30.46% of the correlation with road maintenance cost compared with roughness, which is the optimal sensor monitored indicator. At the local government areas, the MFSD-based indicator can explain 45.8% of practical road maintenance cost. Sensitivity analysis from multiple aspects indicates that MFSD is a reliable and accurate method in decision making. The proposed method and analysis have broad potentials in the network-level sustainable infrastructure management.

道路基础设施的绩效与客运和货运系统以及社会经济发展密切相关。道路基础设施的性能通常由传感器监控的指标衡量，而监控指标显示实际性能的能力通常由决策者和道路使用者确定。但是，由于各个传感器监控的指标的局限性以及人类经验的潜在偏见和不确定性，通常在将路标直接应用于道路性能评估时并不可靠。为了解决这些问题，本研究提出了一种模型驱动的模糊空间多准则决策方法（MFSD），以得出可持续发展道路性能的全面而准确的指标。在这个研究中，MFSD方法用于探索澳大利亚西澳大利亚州Wheatbelt地区的公路网。道路属性，交通车辆和气候条件的空间变量用作决策标准。收集了四个传感器监视的指示器，以估计标准的贡献。结果表明，基于MFSD的指标可以更全面，准确地表征可持续道路基础设施的绩效。在研究区域中，基于MFSD的指标与道路养护成本相比，可以改善30.46％的相关性，这是最佳的传感器监测指标。在地方政府区域，基于MFSD的指标可以解释实际道路维护成本的45.8％。多方面的敏感性分析表明，MFSD是一种可靠，准确的决策方法。所提出的方法和分析在网络级可持续基础设施管理中具有广阔的潜力。