This document discusses bus headway variability and strategies to achieve regular bus service. It begins by outlining what passengers seek in public transit and how to achieve those attributes. While increasing speed and fleet can help, it is important to also focus on decreasing travel time variability and achieving regular headways for reliability. The impacts of unreliable service with bus bunching are then reviewed, including increased waiting times, occupancy and user dissatisfaction. Finally, the document discusses that solutions do exist, such as dedicated infrastructure, signal priority and real-time control systems, to stabilize service and achieve the goal of reliable transit.

1. The full cost of bus headway variability and our
experience to achieve a regular service
Juan Carlos Munoz
Bus Rapid Transit Centre of Excellence
Centro de Desarrollo Urbano Sustentable, CEDEUS
Pontificia Universidad Católica de Chile
Hong Kong, December 8th, 2018
@JuanCaMunozA

2. Work done with collaborators at PUC

3. Cities need its citizens to use public transport
This is easier if citizens like their public transport,
if they are satisfied with its use

4. What do people seek when travelling
in public transport?
Fast Low waits Comfortable Reliable

5. How can we achieve these attributes?
DECREASE
TRAVEL TIME
INCREASE
FREQUENCY
INCREASE FLEET
OR
INCREASE SPEED
INCREASE
TRANSPORT CAPACITY
DECREASE
TRAVEL TIME VARIABILITY
Fast Low waits Comfortable Reliable
INCREASE SPEED
INCREASE FLEET, VEHICLE SIZE
OR
INCREASE SPEED
REGULAR HEADWAYS

6. To achieve these attributes….
Fast Low waits Comfortable Reliable
INCREASE SPEED!
AND THEN MAYBE IF WE
HAVE SPARE TIME FOCUS
ON REGULAR HEADWAYS….

7. BRT : Bus Rapid Transit

8. But… is this enough?

10. What about reliability?
What are the impacts of an unreliable service?

11. To achieve these attributes…
Fast Low waits Comfortable Reliable
I N C R E A S E S P E E D
A N D
R E G U L A R H E A D W A Y S !

12. Vehicles in public transport systems
behave like magnets….
Any perturbation and the equilibrium is lost… all around the world.
Regular intervals is an unstable equilibrium

13. Bus bunching… observed in London

14. Bus bunching… observed in Beijing

15. Bus bunching… observed in Boston

16. Bus bunching… observed in Santiago

17. Today we will address three questions
Why does bus bunching happen?
What are the impacts?
1
2
Does it have a solution?3

18. Why does bus bunching happen?
• Dispatch strategy
• Speed variability
• Travel time variability
• Dwell time variability

19. Why does bus bunching happen?
Time (minutes)
Position (km)

20. Why does bus bunching happen?
• Dispatch strategy
• Speed variability
• Travel time variability
• Dwell time variability

21. Why does bus bunching happen?
Congestion Traffic lights Driver heterogeneity
Travel time variability

22. Why does bus bunching happen?
Stochastic passenger arrivals
at each stop
Dwell time variability

23. Today we will address three questions
Why does bus bunching happen?
What are the impacts?
1
2
Does it have a solution?3

24. What are the impacts?
Impacts are various, affecting:
Users Operators

25. What are the impacts?
• Consider the following bus service
• 5 min average headway
• 80-pax capacity buses
• An average demand of 50 pax/bus

26. What are the impacts?
8:00 8:05 8:10 8:15 8:20
8:00 8:02 8:10 8:12 8:20

27. What are the impacts?
• Increase in average occupancy
• Longer waiting times correlate with high
occupancy
• More users suffer high occupancy
• Increase in occupancy variability
• Generates uncertainty in the level of service

28. What are the impacts?
• Increase in waiting times
• One minute of waiting is perceived as
two or three minutes of in-vehicle time
• Increase in the variability of waiting
times
• Leads to earlier departures for users

29. What are the impacts?
• Increase in dwell times
• Boarding and alighting is slow
and creates friction between passengers
• Increase in the variability of dwell time
• Damages the reliability of the service
• Operators need extra fleet and drivers due
to variable cycle times

30. Cities need its citizens to use public transport
This is easier if citizens like their public transport,
if they are satisfied with its use

31. Cities need its citizens to use public transport
This is easier if citizens like their public transport,
if they are satisfied with its use
How does reliability affect user satisfaction?

32. What are the impacts?
User
Satisfaction
Occupancy per bus

33. What are the impacts?
User
Satisfaction
Occupancy per bus50 pax
80%
Imagine a regular service with
a 5 minute interval,
50 passengers per bus and
63% occupancy.

34. What are the impacts?
User
Satisfaction
50 pax
80%
Suppose the service is not
regular, and operates with
intervals of 2 and 8 minutes
alternately (with occupancies
of 25% and 100%
respectively).
90%
20 pax 80 pax
10%
Occupancy per bus

35. What are the impacts?
User
Satisfaction
63%
80%
90%
25% 100%
Average occupancy of buses
will be 63%, with an
apparent satisfaction of 50%.
50%
10%
Occupancy per bus

36. What are the impacts?
User
Satisfaction
63%
80%
90%
25% 100%
Average occupancy of
perceived by users will be
85%, with a satisfaction level
of only 26%!
50%
85%
26%
10%
Occupancy per bus

37. What are the impacts?
User
Satisfaction
63%
80%
90%
25% 100%
50%
85%
26%
10%
This is what we planned for…
This is what we achieved
with no control!
Occupancy per bus

39. To achieve these attributes…
Fast Low waits Comfortable Reliable
I N C R E A S E S P E E D
A N D
R E G U L A R H E A D W A Y S !

40. Today we will address three questions
Why does bus bunching happen?
What are the impacts?
1
2
Does it have a solution?3

41. Does it have a solution?
•Yes!
• Dedicated infrastructure
• Transit signal priority
• Real-time control at dispatching
and along the route
• (developed by )

42. Does it have a solution?

43. Does it have a solution?
Source: mto.gov.on.ca

44. Does it have a solution?
No control
With control
Color indicates
bus load!

45. Time-space trajectories
0 20 40 60 80 100 120
0
1
2
3
4
5
6
7
8
9
10
s2 NETS sc corrida17Distance(Km)
Time(minutes)
Real Time Holding
0 20 40 60 80 100 120
0
1
2
3
4
5
6
7
8
9
10
Scenario 1 threshold run17
Distance(Km)
Time(minutes)
No Control

46. Bus loads
0 5 10 15 20 25 30
0
20
40
60
80
100
120
Scenario 1 HBLRT alpha=05 Beta=05
Load(Pax.)
Stop
0 5 10 15 20 25 30
0
20
40
60
80
100
120
Scenario 1 HBLRT alpha=05 Beta=05
Load(Pax.)
Stop
No Control Real Time Holding

47. Cycle times
25 30 35 40 45
0
50
100
150
200
250
300
350
mean =33.64
Std.Dev. =3.51
No control
Frequency
Cycle Time (Minutes)
25 30 35 40 45
0
50
100
150
200
250
300
350
mean =32.11
Std.Dev. =1.2
HRT 05
Frequency
Cycle Time (Minutes)
No Control Real Time Holding

48. Does it have a solution?
Ok! Hold Go faster
(if possible)
Slow down

49. Does it have a solution?

50. The future of BRT should be written with double R
BRRT
Bus Rapid and Reliable Transit

51. The full cost of bus headway variability and our
experience to achieve a regular service
Juan Carlos Munoz
Bus Rapid Transit Centre of Excellence
Centro de Desarrollo Urbano Sustentable, CEDEUS
Pontificia Universidad Católica de Chile
Hong Kong, December 8th, 2018
@JuanCaMunozA