Peer comparisons allow transit decision-makers to objectively analyze their operations and to identify and prioritize problem areas for management actions. An important requirement of this comparative analysis is to identify peer systems that have similar characteristics across select peer variables such as vehicle fleet, service population, and so on. This paper presents a fully automated system that can identify peer systems and peer states using data from the National Transit Database (NTD). Known formerly as the Section 15 data, NTD has been used extensively to derive values for transit performance measures. The automated system described in this paper allows a comparative analysis to be performed in a matter of minutes. Following the user specifications, the system searches for the most comparable systems using the nearest-neighbor method. Users can select the desired number of peers, transit modes, service types, and geographic areas. Users can also select the desired peer variables to determine the degree of similarity among potential peer systems. The peer variables can be weighted for different levels of importance. The analysis can be performed at the individual mode, systemwide, or statewide level. In addition to finding the peer systems for a specific system, the system also includes a procedure based on the K-mean clustering method to allow the user to divide transit systems into peer groups. Selected peer groups can then be used in peer analysis using various analysis tools that come with the system.
