Discussion Papers
Dec 30, 2025
6
Minutes read

Assessing Bus Performance in Greater Kuala Lumpur

Key Takeaways
Data Sets Overview
  • Most services cluster near the on-time corner, with Rapid KL recording 86% and MRT Feeders 91% of departure within the acceptable window. However, variability differs sharply. Rapid KL routes were more scattered, while MRT feeder routes are tightly clustered, pointing to more consistent performance in MRT feeders.
  • Early departures (earlier than 1 minute ahead of schedule) were observed in 11-12% of trips. These typically appeared along major roads with low service frequency.
  • The Median Bus Performance Index (BPI) is higher for MRT feeders (~0.88) compared to Rapid KL (0.72-0.78), and Rapid KL exhibits a much longer tail. For commuters, this means that while most routes are serviceable, certain Rapid KL routes remain unreliable
assessing-bus-performance-in-greater-kuala-lumpur
Discussion Papers
Peer-reviewed publications designed to share emerging research or early analysis on key issues. Discussion Papers aim to invite feedback, encourage dialogue, and inspire further study or collaboration.
Disclaimer
As we transition to a digital-first communication and continue building our knowledge hub, publications released before October 2025 are preserved in their original format. Publications released from October 2025 onward adopt a new, digitally friendly format for easier online reading. The official versions of earlier publications, including their original language and formatting, remain available in the downloadable PDF.

Summary

• Bus reliability in Greater Kuala Lumpur remains uneven, contributing to commuter distrust in bus services. Using high-frequency GTFS static and real-time data, we show that while many routes achieve acceptable punctuality scores on average, a non-trivial subset of Rapid KL bus routes exhibit large and unpredictable deviations. These deviations undermine commuters’ ability to plan daily travel and erode confidence in busses as a dependable mode of transport.

• We develop a Bus Performance Index (BPI) that integrates punctuality with the severity of service deviations. By combining on-time performance with a normalized measure of deviation magnitude, the BPI distinguishes between routes that are occasionally late, and those that fail severely when they happen to be late. This allows the index to more closely reflect commuter experience.

• MRT Feeder services consistently outperform Rapid KL bus routes both in reliability and predictability. Across the study period, MRT Feeder routes record higher BPI scores, tighter performance distributions, and near-zero collapse rates. In contrast, while the specific Rapid KL bus routes with very low scores vary from day to day and week to week, the proportion of such low-performing routes remain relatively stable over time. This pattern points to a systemic reliability issue embedded in the network.

• Meaningful near-term improvements in reliability can be achieved through targeted operational reforms, even within existing infrastructural constraints. The results point to practical interventions such as timetable recalibration using real-time data, improving real-time passenger information, and bus control strategies such as conditional transit signal priority. These measures directly address reliability failures that commuters experience today, while also strengthening the effectiveness and resilience of future infrastructural investments when capacity expansion becomes necessary.

Read Full Publication

Article highlight

featured report

Conclusion

Attributes
Footnotes
References
Photography Credit

Related to this Publication

No results found for this selection
You can  try another search to see more

Want more stories like these in your inbox?

Stay ahead with KRI, sign up for research updates, events, and more

Thanks for subscribing. Your first KRI newsletter will arrive soon—filled with fresh insights and research you can trust.

Oops! Something went wrong while submitting the form.
Follow Us On Our Socials