Traditionally dynamic segmentation has been used as a method to view linear assets on a geographic information system (GIS) map. The principle being that continuous segments with homogenous condition can be viewed as a single segment. These segments are viewed as a line with unique attributes. This concept is now being expanded to enable modern Asset Management software to analyze networks more efficiently. There are several methods of dynamic segmentation, including set banding and tolerance ranges. Each of these has advantages and disadvantages, depending on the desired application. In light of this a new system has been developed, combining these principles that removes many of the disadvantages. This process involves the construction of a floating band system, which takes into account all previous segment values when determining segmentation tolerances. As stated, the traditional application of Dynamic Segmentation is in the field of GIS. This very often limited the segmentation to one attribute of the network. However, treatment practices for assets are generally based on multiple attributes. Therefore it is necessary to use a system that combines multiple attributes into one item. This has lead to the development of category codes, which are a combination of attribute codes. These attribute codes are transformations of each attribute necessary for categorizing a road into a numerical value. These values can then be combined to form the category code. There are several ways to combine segments into homogenous sections. The most common of these is continuous collation. This method combines adjacent segments of the network where all segmentation criteria are met. The second method for merging segments is network wide collation. This allows all segments that are homogenous to be considered as one single section. The main benefit is that the number of sections fed into the analysis engine is greatly reduced. However it does require that all attributes of the network are dynamically segmented by the set banding method. The major benefits of dynamic segmentation can be seen in its application to Asset Management analysis software. The data set-up and processing times are significantly influenced by the population size of the network. This relationship may be linear or exponential. Therefore, with reduction in network size of between 10% and 90% achievable through dynamic segmentation, there is certainly scope for huge savings by adopting the philosophy. The successful future of dynamic segmentation relies on the ability to reduce the number of network sections, while maintaining detail in the data required for decision making. Therefore it is recommended that a focus on developing an optimum segmentation process, for a range of applications, is required. This would mainly focus around producing bandwidths and tolerances used in the segmentation process that produce the best possible combination of network size and data relevance.


  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00964428
  • Record Type: Publication
  • ISBN: 087659229X
  • Files: TRIS, ATRI
  • Created Date: Oct 22 2003 12:00AM