Bus services usually tend to be irregular and their level of irregularity depends on various factors, such as crowding level, terminal departure behaviour, passengers’ behaviour, operator behaviour, traffic and weather condition and etc. High-frequency bus routes have shorter headways (usually headway less than 10 to 15 minutes) and higher passenger demand compared with normal routes. Therefore, level of irregularity can be even higher in bus service at high-frequency operation. Running time variability comes from both systematic changes in ridership and traffic levels at different times of the day, which can be accounted for in service planning, and the inherent stochasticity of homogeneous periods, which must be dealt with through real-time operations control. This study evaluated impact of ridership changes and traffic condition through time of the day on running time variability, using Automatic Vehicle Location system (AVL) and Automatic Fare Collection system (AFC). All data extracted and collected from RapidKL Company for route U32, which is a high-frequency route in downtown of Kuala Lumpur. Descriptive analysis on data showed a high variation in running times, especially in morning peak hours. A liner regression model also proved than crowding level (extracted from AFC data), number of stops and congestion zones have relatively high impact on running time variation.
This paper presents a development of an expert system to be used as an advisory in finding the solution to problems which are normally solved by human experts. The E-ACTIVETRANS is developed to help young engineers/planners in designing a new cycle lane in urban areas and also to help in reallocation of an existing roadway space for cycle lanes. This system has three sub-systems: Planning on Strategies to Shift from Passive Transportation to Active Transportation, Design on Bicycle Facilities and Examples of Successful Implementation. This paper focuses on the design of bicycle facilities whereby the prototype was developed based on data acquired from the domain experts who are involved in bicycle facility module design, as well as the initial text analysis obtained during the domain familiarisation stage. The validation of the system was performed through a comparison of knowledge content in E-ACTIVETRANS based on expert opinion. The average level of acceptance is 91 percent which validates the system and knowledge of the experts.