2.2.2. The 3D Data Model for Underground Utility Networks

The CityGML utility network Application Domain Extension (ADE) [32] focuses mainly on three aspects: (i) the general 3D geometric of network components; (ii) the 3D topographical structure of the entire utility network; and (iii) the functional information of different types of the network [32,33]. Based on the general concepts of the utility network, different types of utility networks can be implemented with their specific function [32]. Moreover, the interdependence between utility network features and city objects can be presented in 3D space [34]. Because this data model is an extension of CityGML [35], which is the popular standard for 3D city modelling (e.g., building), it is beneficial to integrate information of utility networks to the infrastructures to support urban planning and the other city studies. However, it does not consider the accuracy of the data. Some works begin to extend the existing data model to consider many more details about utility networks, such as [36], represent geographical uncertainties of utility locations based on CityGML Utility Network ADE.

The Industry Foundation Classes (IFC) utility model [37] is an ISO standard for data exchange of buildings in the architecture and civil engineering domain [32]. In the utility part, it describes 2D and 3D geometries of utility elements. Meanwhile, two different ways of connection are defined to describe the relationship between supply service components within the building, which is a logical and physical connection. In addition, it has a comprehensive semantic definition of utility network objects. However, this standard only focuses on the building level and lacks spatial information.

The INSPIRE Data Specification on Utility and Government Services—Technical Guidelines [38] organize the basic information of utility networks and administrative services of utility networks in a city or country range. It is a part of INSPIRE, which is a standard of the European Union to describe the spatial information of infrastructures. However, the INSPIRE Utility networks are lacking a definition of 3D geometric information of utility networks.

ESRI Utility Network model [39] provides a GIS-based utility solution to represent the basic logical and physical structure of all types of utility networks, which is composed of edges and junctions. This model is a general utility data model to represent the 2D geometric information and connections of the utility networks.

Until now, there has not been an international standard that has been widely used for 3D modelling of underground utility [40]. Although some existing standardized data models have been developed to integrate multi-utility networks, they can not guarantee the information to be reliable [3]. In order to develop a comprehensive utility database, we have the challenge to integrate different types of utility datasets from multiple surveying techniques, as well as the existing 2D data. Table 2 compares four popular utility data models relevant to the objectives of this work. Obviously, most of the existing utility data models are to focus on the 3D representation, and include 3D geometric and topological information. The existing data models provide a good reference to describe the geometric and spatial information of utility networks in 3D. Nevertheless, none of them considers the accuracy of data of underground utility networks. On the one side, the survey technique directly impacts the data accuracy. However, industry service providers are not usually aware of these extensive standards [3]. On the other side, different applications might use data at different levels of accuracy. Hence, we need an ideally 3D utility data model to support mapping procedures and control accuracy of underground utility network data.


**Table 2.** Comparison of model characteristics.


On the basis of their discussion and the situation of Singapore, it is necessary to register the utility segments as the legal objects in the land administration system, which helps to identify the ownership of underground utility. An integral approach needs to be developed based on legislative and technology solutions. It is essential to establish a degree of reliability and consistency between data produced by different service providers. It is essential to standardize the practices regarding the use of those techniques and various information management. In the underground utility data model, land parcel, as an important role in the land administration, should be connected to the underground utility networks [18,21].
