-
1
Term Project Report, Fall 2012 MIE440F: Mechanical Design Theory
and Methodology
Dept. of Mechanical/Industrial Engineering, University of
Toronto
REDUCING UNEVEN PASSENGER DENSITY ON TTC SUBWAY PLATFORMS
Sissi (Haiyun) Wang (33%) Jennifer Wong (33%)
Song Yang (33%)
ABSTRACT
We describe the design process, as well as the final outcome,
for the development of a product which promotes even passenger
density on Toronto Transit Commission (TTC) subway platforms. We
identify lead users and their needs in order to gain a greater
understanding of the problem. We review competitive products,
identifying their shortcoming, as well as their benefits. We then
generate possible designs using a variety of methods, including
6-3-5, TRIZ, SCAMPER, and biomimetic design. Finally, we evaluate
our concepts using both a weighted decision matrix, as well as a
PUGH chart, to arrive at the final solution. The solution we
propose consists of a series of barriers along the platform, which
will visually segment the platform into sections. We believe that
this will aid passengers in identifying less crowded sections,
thereby promoting movement to those areas. The barriers are
designed to have adjustable length, and can be easily installed
without altering existing TTC infrastructure.
INTRODUCTION Overcrowding reduces the efficiency of transit
systems in many countries. Although crowding itself is difficult to
address without changing existing policy and infrastructure, some
of its negative effects can be mitigated using product solutions.
One such effect is the uneven distribution of passengers on loading
platforms. Jaiswal et al. (2008) demonstrated a tendency for
passengers to congregate in preferred areas of loading platforms
when waiting for public transit. Moreover, these regions tend to be
near platform entrances and exits. The most direct impact of uneven
crowd density is an increase in train boarding times. Fang et al.
(2003) showed an inverse exponential relationship between crowd
density and crowd movement velocity. An increase in train boarding
time results in less trains being able to service a station during
any given time period, and further aggravates the crowding. Figure
1
shows the relationship between crowd density and walking
velocity.
Fig. 1: Graph from Fang et al. (2003) showing the
relationship between crowd density and walking velocity. Note
that this relationship holds even during emergency
situations. Overcrowding can also cause negative psychological,
health, and financial effects. Psychological effects include higher
stress levels, more aggressive behavior, and lower cognitive
performance due to sensory overload (Sammons). Negative health
effects include an increase in blood pressure and a depressed
immune system (Sammons). Overcrowded conditions also facilitate the
spread of infectious diseases. A major safety issue caused by
crowding is possible injury and the loss of life if one were to get
pushed off the platform.
From a financial standpoint, it is beneficial to solve platform
crowding as getting one or more additional trains into the subway
station within an hours time can buy an extra 3- 5%
-
2
capacity and possibly an increase in ridership according to Adam
Giambrone, the TTC chairman. According to New Yorks subway transit
system, a 2% increase in ridership led to 1.3% increase in revenue
in year 2012. It also saves passengers commute time allowing them
to make more productive use of their time.
LEAD USERS
The lead users for the subway platform crowding problem are
interpreted as people and professionals that have to deal with very
large crowds on a regular basis, and who encounter crowds more
frequently than the general population.
Who The lead users identified are: law enforcement, military,
and daily TTC commuters. Law enforcement and military have to
manage large crowds in chaotic situations such as riots through
extreme measures of crowd control. Daily TTC commuters encounter
the subway crowding problem more often than non-regular riders.
How A focus group was conducted to identify the daily TTC
commuters needs. Six students from different disciplines at the
University of Toronto were gathered to answer a series of questions
with regard to their TTC commuting experience. The commuters came
from various stations, each of which served a different purpose
(e.g. interchange station, terminal station). The project group
first set out to verify whether a crowding issue exists around
subway platform staircases, and the commuters verified that
crowding was an issue during high-traffic times. They then
proceeded to identify the obstacles commuters faced that prevented
them from dispersing themselves on the subway platform such as
laziness, ineffective TTC signs and broadcasts, and following the
crowd due to unfamiliarity with station. Finally, they asked
improvement questions on how to improve the commuters TTC
experience in order to extract the needs of the commuters to work
into the product solution for reducing crowding.
What
The focus group participants provided suggested improvements
that revealed underlying needs. Although the ideas proposed were
improvements in infrastructure, we were able to extract from these
three basic user needs: effectiveness, comfort, and accessibility.
Table 1 categorizes all suggested improvements from the focus group
participants into the three basic user needs.
Table 1: Identified user needs
Identified needs Participant responses 1. Effectiveness Conveyor
belts, board trains faster 2. Comfort More benches, leaning
against
walls, more personal space 3. Accessibility Wider platform,
bigger doors, more
frequent trains, more staircases, better control of entrance vs.
exit flow, open path across platform
Effectiveness is measured by the time it takes for
passengers
to board the train. We assume that a more even distribution of
passengers on the platform leads to a reduction in boarding
time;
Comfort was measured by the amount of sensory discomfort a
commuter experiences (e.g. standing instead of sitting while
waiting); and
Accessibility is measured by the amount of clearance available
for the commuter to travel from one point to another on the
platform.
COMPETITIVE ASSESSMENT Our assessment of competitive products
covered solutions currently employed by the TTC, as well as
equipment used by the military and law enforcement in riot
control.
Physical Barriers They can be seen at Bloor & Yonge station
during high traffic times. They are effective at controlling the
crowd movement, but have poor accessibility, making it difficult
for people to travel from one point to another on the subway
platform. Figure 2 shows the configuration of physical barriers at
Bloor & Yonge station during rush hour.
Fig. 2: Barriers at Bloor & Yonge
-
3
Signage & PA Broadcasts Signage and PA broadcasts are
currently used at TTC subway stations to instruct passengers to
spread out on the platform. However, they are not effective since
they are a very passive method to direct crowds.
Law Enforcement Equipment An example of such equipment are the
Mosquito alarm which emits sound at a very high frequency to
disperse crowd and the Active Denial System which shoots focused
electro-magnetic wave energy to disperse crowd. While they are
effective at controlling the crowd, they are very harmful, cause
discomfort for people, and are illegal for public use. Figure 3
describes what each component of the Active Denial System does
during implementation. Table 2 provides a summary of the benefits
and shortcomings of existing solutions.
Fig. 3: Conceptual diagram for the Active Denial System, which
uses directed heat to disperse crowds
Table 2: Summary of the pros and cons of existing solutions
Existing solution Gap
1.Physical barrier Effective crowd control, but severe reduction
in accessibility
2. Signage and announcements
Passive solution, ineffective
3. Law enforcement equipment
Effective crowd control, but harmful and illegal in public
CONCEPT GENERATION
Biomimetic Design Two concepts were generated using biomimetic
design:
1. Using the word open, we found the following excerpt: The Na+
concentration is much higher outside the axon than inside, so when
the sodium channels open, Na+ ions from the outside rush into the
axon. This led to the One-way Turnstile, as shown in Figure 4:
Fig. 4: One-Way Turnstile turnstiles are installed halfway
between the platform entrance and platform end. Passengers are
permitted to move freely from the entrance towards the end, but
movement in the reverse direction is
limited.
2. Using spread, we found the following excerpt: Plant viruses
may infect cells... then spread rapidly through... microscopic
channels which traverse the cell walls of plant cells. This led to
the Discretized Platform, as shown in Figure 5:
Fig. 5: Discretized Platform the platform is divided into
several distinct segments using barriers. Passengers can
more easily identify crowded areas, and avoid them accordingly.
Each segment is thought of as a cell, and the
open walkway on the platform is the channel.
-
4
TRIZ We found it difficult to identify appropriate improvement
and conflicting parameters. Therefore, we browsed through the list
of TRIZ principles, and were inspired by No. 15 the Principle of
Dynamism: Divide an object into parts capable of movement relative
to each other. Using this, we developed the Seesaw Platform, as
shown in Figure 6:
Fig. 6: Seesaw Platform several large plates of metal are
installed onto the platform. Each plate is balanced on a fulcrum
located in the center. The plate tilts when one
side has more passengers, resulting in a feeling of discomfort
due to lack of balance. This motivates
passengers to relocate to the other side of the plate
SCAMPER In using SCAMPER, we selected as our base case the signs
being used by the TTC. For S (substitute), we substituted the sense
of vision with other senses, such as tactile or aural. From this we
developed ideas such as PA broadcasts, also currently in use by the
TTC, and placing heating units at the ends of the platform during
the winter to encourage movement away from the center.
However, we felt these solutions were ineffective. By using M
(magnify), we intensified them into a group of products we call
Aggressive Solutions. These solutions aim to move users away from
platform entrances through negative reinforcement. Figure 7 shows
an example of an aural aggressive solution:
Fig. 7: An example of what an aggressive solution might look
like.
6-3-5 The most interesting solution which resulted from our
6-3-5 brainstorming session was the Textured Floor, as shown in
Figure 8:
Fig. 8: Textured Floor a rubber floor mat with protruding rubber
spikes is rolled out at platform entrances/exits. This
discourages passengers from staying near staircases
CONCEPT EVALUATION/SELECTION
Seven criteria were used to evaluate the concepts. Three of
these (accessibility, comfort, effectiveness) resulted from the
focus group we conducted with frequent TTC commuters. The rest were
factors identified during team discussions. The criteria were:
Accessibility: the amount of clearance available on the
platform (m2). This determines how easily a passenger can move
from one location on the platform to another;
Adaptability: the number of stations at which the product can be
deployed without being remanufactured with different specifications
(e.g. size);
Comfort: the amount of sensory discomfort a commuter experiences
while waiting;
Cost: the amount of money ($) required to manufacture the
product, as well as the raw material cost;
Effectiveness: the time it takes for all passengers to board the
train (seconds). We assume that increased evenness in platform
density leads to reduced boarding time;
Durability: the amount of time before the product is expected to
require repair (weeks).
We evaluated our design concepts using two methods: the weighted
decision matrix and the PUGH chart. For the weighted decision
matrix, we first evaluated the relative importance of our criteria
using pairwise comparison:
-
5
Table 3: Evaluation of criteria using pairwise comparison
Criteria A B C D E F G Weights
A) Accessibility - 0 1 0 1 0 0 10% B) Effectiveness 1 - 1 0 1 1
0 19% C) Comfort 0 0 - 0 1 1 0 10% D) Cost 1 1 1 - 1 1 0 24% E)
Durability 0 0 0 0 - 1 0 5% F) Adaptability 1 0 0 0 0 - 0 5% G)
Safety 1 1 1 1 1 1 - 29%
Total 100% We subsequently applied this to a weighted decision
matrix, the result of which is found in Appendix A. Furthermore, we
also used a PUGH chart to evaluate our concepts. The datum used for
the PUGH chart was signs and PA broadcasts as currently employed by
the TTC. The results of the evaluation are summarized below:
Table 4: Summary of results obtained using weighted decision
matrix and PUGH chart
Concept Weighted Matrix PUGH Chart
Textured floor 2.10 -2 Aggressive solutions 1.62 -2 See-saw
platform 1.38 -3 Turnstiles 1.86 1 Discretized platform 2.38 0
An interesting observation from the PUGH chart is that the
majority of our concepts seem to perform worse than the TTC
signage/PA broadcasts. This is due to the lack of weighted criteria
in the PUGH chart. Because the current solutions employed by the
TTC naturally score very high in accessibility (signs and
broadcasts do not occupy platform space), comfort (signs and
broadcasts do not directly impact passengers), and cost (signs and
broadcast are likely less expensive than new products).
Furthermore, due to the nature of our design concepts, the TTC
solution also performs well in safety. The full PUGH chart, as well
as the full weighted decision matrix, can be found in Appendix A:
Concept Evaluation. Based on the results of the weighted decision
matrix and PUGH chart, we selected the Discretized Platform as the
final design concept. This result seems reasonable, since the
discretized platform is the safest design out of all concepts. The
aggressive solutions and seesaw platform pose obvious hazards to
the passenger's safety. The textured floor poses a safety risk
if
passengers fall, especially during winter, and the turnstiles
are an obstacle in an emergency evacuation.
FINAL DESIGN
Focus was placed on designing the barriers which would be used
to separate the platform into sections. Through discussion, we
arrived at three constraints for the barrier:
The length of the barrier must be adjustable, since platform
width is not uniform across all TTC stations. Variable length
would allow barriers to be deployed at multiple stations without
remanufacture.
The barrier should be easy to install without modifying existing
TTC infrastructure. The installation must be very secure, as people
tend to lean on barriers (an affordance). It should also be easy to
take down for maintenance.
The barrier must not lead to significant visual obstruction on
platforms. This is a safety consideration.
Overall Structure
The final design is a barrier comprised of 4 panels nested
within each other, with the innermost panel being solid. The
innermost panel can be extended from second innermost panel, and
the second innermost panel can be extended from the second
outermost panel, etc., allowing for adjustable barrier length. The
height of the barrier is 1m, and each panel has a length of 0.75m,
giving the fully extended barrier a total length of 3m.
Fig. 9: Arrows depict the location of barriers relative to
train doors The number of barriers installed along a platform is
chosen to be 5 since each subway train is comprised of 6 cars
linked together. Each car has a set of doors at either end,
allowing adequate space for the barriers to fall in between the
doors, despite how the final locations of the train doors tend to
be inconsistent on the platform. Figure 9 shows the relative
locations of the barriers to the train doors. Considering safety
aspects, the barrier was designed to minimize visual obstruction,
therefore, a translucent plastic material is used for the main body
of the panels. Each panel will have a surrounding frame with holes
to secure panels with respect to each other. Figure 10 shows a
possible configuration of the barrier with the innermost panel
extended.
! ! ! ! ! ! ! ! !
-
6
Fig. 10: Barrier with one panel extended In order to secure the
barrier onto the platform, we propose to fix it in place at both
ends. A vice will be used to clamp the barrier onto the edge of the
platform, taking advantage of the existing overhang (see Figure
11).
Fig. 11: Cross-section of a TTC platform. Note the overhang at
the edge of the platform.
The other end of the barrier will be secured onto the platform
floor using a vacuum lock. When secured in place, the suction cup
creates a strong vacuum, making it difficult to move the barrier.
Figure 12 shows a vacuum lock when secured:
Fig. 12: An example of a vacuum lock
An additional challenge is securing the panels with respect to
each other so passengers cannot modify the barrier length. We
achieve this by placing holes in each panel, allowing them to be
secured with Torx screws once in the desired configuration. Figure
13 shows the Torx security screw head pattern.
Fig. 13: Torx security screw head. We assume that the typical
TTC passenger would not be in possession of a
screwdriver capable of removing such screws.
Material: The main body of each plate is made of Lexan, a brand
of polycarbonate resin thermoplastic. Lexan is a resistant
material, with a tensile strength of 60-70 MPa (Sabic 2008). Each
plate is surrounded by an aluminum frame, which further improves
the structural integrity of the barrier.
Table 4: Volume, weight, and cost for one barrier, broken down
by material type
Volume (cm3) Weight (kg) Cost (US$)
Lexan 42750 51.3 $99.32 Aluminum 4709 12.71 $24.61 Total 47459
64.01 $123.93
DISCUSSION
Limitations of methods In terms of concept generation, we
(surprisingly) found that biomimetic design generated the most
interesting concepts. The greatest difficulty came in finding the
correct keywords to use. Furthermore, our source was limited to one
natural-language text. On the other hand, TRIZ was particularly
difficult to use, and only produced one result, because finding
relevant improvement and conflicting parameters was problematic. In
terms of concept evaluation, the PUGH Chart method of evaluation
assumes that every objective is equally important which led to many
solutions being worse to the existing TTC solution of PA systems.
The weighted decision matrix method of evaluation scored the
solutions on a 3-point scale. A finer granularity used in scoring
would give more information, which would lead to a more accurate
ranking of the solutions.
-
7
Limitations of final design
The proposed solution has many advantages over existing
solutions. It provides more accessibility than the current physical
barriers implemented at Bloor & Yonge station, since its
purpose is to segment the platform rather than restrict the path
users can take from the platform entrance to the train. Compared to
law enforcement equipment such as heat-ray guns, the proposed
solution is also less harmful to the user and less costly to
implement.
However, one limitation of the final design is that users can
fail to identify where empty sections are if their vision is
obstructed by crowds. A way to mitigate this issue is to use the
barriers in conjunction with LCD screens which notify users of
empty sections. Another improvement is to include grooves or
rollers at the base of each panel to ease implementation.
The solution is likely to succeed in the marketplace as it has
applications beyond the TTC subway platforms due to its ability to
be secured onto any protruding edge and adjustable length. It also
has a low material cost and is easy to implement compared to
existing solutions. The barriers can also be used to generate
revenue by using a portion of the panels on either side as
advertising space, while maintaining non-visually obstructive.
ACKNOWLEDGMENTS The authors would like to thank Jayesh
Srivastava and Hyunmin Cheong for their insightful comments and
clarifications during the design process.
REFERENCES Fang, Z. et al., 2003, On the relationship between
crowd density and movement velocity, Fire Safety Journal, Vol. 38,
No. 3, pp. 271-283. Jaiswal, S. et al., 2008, Relating bus dwell
time and platform crowding at a busway station, Proceedings from
the 31st Australasian Transport Research Forum, pp. 239-249.
Sammons, Aidan. "Effects of Crowding." N.p., n.d. Web. 2 Dec. 2012.
"Higher Subway Ridership Brings Revenue and Crowding." Allvoices.
N.p., n.d. Web. 07 Dec. 2012. Sabic Innovative Plastics. "Lexan*
9030 and Lexan* 9030 TG Sheet." N.p., Jan. 2008. Web. 7 Dec.
2012.
-
8
APPENDIX A: Concept evaluation methods
Weighted Decision Matrix
Pugh Chart
-
9
APPENDIX B: 2-D drawing of final design