Visual Quality Assessment Of Urban Infrastructures Through Malfunction Incidents

Urban infrastructures are critical,as they influence citizens’ life and urban development.They may suffer a variety of malfunction incidents every day,and citizens who are influenced by these incidents will report malfunction issues they witness to the non-emergency number of urban infrastructures.On the one hand,incidents are real reflections of urban infrastructure quality.On the other hand,the quality can feed back into incidents’ happening.This close relationship of the two gives us inspiration that quality assessment of urban infrastructures can be based on the study of malfunction incidents.However,this quality assessment can be challenging,because incidents are high-dimensional,and temporal-spatial varied.This article proposed a visual quality assessment method.This method was divided into three levels: analysis of malfunction incidents,of their causal relationship and of their temporal-spatial evolution.Firstly,we extracted incidents from telephone issues through our temporal,spatial and semantic clustering algorithm,and designed stack-up histogram and petal diagram to visualize the temporal and spatial patterns of incidents.Then we used the Bayesian network algorithm to extract the causal relationship among incidents,and did topology optimization on the causal network.In addition,the differences of incidents’ spatial distribution in different periods were also compared in this article.Through the differences,we examined the evolution of infrastructure’s quality over time.We not only designed a hexagonal heat map to show the hot spot regions with large variation,but also designed a pattern gallery that could automatically recognize valuable evolutions.Among these tools,the greatest contribution is to analyze the temporal and spatial evolution of incidents.Because the analysis of incidents’ cardinal number and growing rate made the quality assessment more objective.Through these tools,managers can find areas of poor quality of infrastructure service from the historical issues data;can know the most fundamental cause of the incident to improve the efficiency of troubleshooting;can master the infrastructure’s whole quality situation.