The Challenges of Designing Metro Maps

Michael Burch

, Robin Woods

, Rudolf Netzel

and Daniel Weiskopf

VISUS, University of Stuttgart, Stuttgart, Germany

Communicarta Ltd., High Wycombe, Bucks, England

Keywords:

Cartography, Schematic Mapping, Transit Maps, Map Design.

Abstract:

Metro maps can be regarded as a particular version of information visualization. The goal is to produce

readable and effective map designs. In this paper, we combine the expertise of design experts and visualization

researchers to achieve this goal. The aesthetic design of the maps should play a major role as the intention of

the designer is to make them attractive for the human viewer in order to use the designs in a way that is the

most efﬁcient. The designs should invoke accurate actions by the user—in the case of a metro map, the user

would be making journeys. We provide two views on metro map designs: one from a designer point of view

and one from a visualization expert point of view. The focus of this work is to ﬁnd a combination of both

worlds from which the designer as well as the visualizer can beneﬁt. To reach this goal we ﬁrst describe the

designer’s work when designing metro maps, then we take a look at how a visualizer measures performance

from an end user perspective by tracking people’s eyes when working with the formerly designed maps while

answering a route ﬁnding task.

1 INTRODUCTION

Stop! I want to get off... Not an unreasonable request.

But where?

Cartographers face many challenges, not least of

which is how to represent the physical world in as true

a form as possible. Public transport mapping is no dif-

ferent. Leaving aside the long-standing debate about

schematics, one of the challenges the map designer

encounters is the requirement to provide maps to cus-

tomers in the travel and publishing industry where

page size is everything. Here, schematic maps work

very well indeed.

We ﬁrst go back to the basics, i.e., to good old-

fashioned ﬁeldwork. On surface-based networks,

such as Amsterdam, Budapest, or Pisa, walking the

streets plotting each of the routes and stops had to

be done. And in 1990, when we (second author and

his company) ﬁrst started doing this, there was no

internet, which meant that pre-departure information

was very limited. We would not necessarily have

known how many lines there were and over how much

ground. That meant guessing the time required to do

the job and that in turn meant either a long or short

stay or a short stay with long hours.

There was no technical aid like digital cameras

for recording tram and bus stop names in those early

(a) (b)

Figure 1: (a) One of the charming little stops in the middle

of a park in Soﬁa. A direct line from the city center goes

here. (b) A section of the corresponding map. The stop

labeled 1879/1880 is the one in the photo on the left.

days. Everything was handwritten. The arrival of dig-

ital cameras, at prices making ownership a reasonable

proposition, was certainly a boon. It has made ﬁeld-

work so much easier, particularly where the Latin al-

phabet is not used in signage, i.e., in Athens, Bel-

grade, Soﬁa, and St Petersburg, for example. We now

have a library with more than 48,000 photos of every

single station or stop name on every line in every city

we cover, together with an assortment of other inter-

esting subjects, objects, artifacts, and oddities. Fig-

ure 1 shows a typical example of a photo of a stop

and the associated map.

A decision as to which lines to incorporate and

which to leave out must be made and that always de-

pends on the type of system. Mindful that the end-

Burch, M., Woods, R., Netzel, R. and Weiskopf, D.

The Challenges of Designing Metro Maps.

DOI: 10.5220/0005679601950202

In Proceedings of the 11th Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2016) - Volume 2: IVAPP, pages 197-204

ISBN: 978-989-758-175-5

197

user is a tourist who may not have much time to get

to grips with the complexity of the network, we gen-

erally ﬁrst select the ﬁxed route types, such as tram,

trolleybus, or metro and then bus if there is either

space or the book author speciﬁcally mentions a par-

ticular point of interest only reachable by bus.

Naturally, if the system is predominantly bus, like

Pisa, or in the case of Venice, waterbus, then we

choose a selection of routes that properly cover the

most important points of interests ﬁrst, and then in-

clude others to complement those where space and

time allow.

Fixed routes are good to map for tourists—it is

easy for them to locate lines visually at street-level—

either overhead catenary in the case of trolleybus, or

tracks and catenary, in the case of trams. Getting lost

is no problem; exploring the city you will eventually

come across a route somewhere. It is a simple case of

following the line in either direction until you come

across a stop, note the name (or number), look at your

schematic map, and off you go, or on you get!

In fact, perhaps to the wrath of other cartogra-

phers, we would argue that getting lost and exploring

somewhere without a street map is absolutely ﬁne and

very enjoyable—armed with your transport map you

will always get back to where you want.

In this paper, we illustrate some challenges when

designing metro maps from a designer’s point of view

but also from a visualizer’s point of view. Both have

different goals in mind, but in the end they also share

some common goals, i.e., generating readable and

useful metro maps helping travelers ﬁnd their ways

in a foreign city.

2 RELATED WORK

From a visualization perspective, metro maps fall into

the category of relational data, i.e., they should be

an efﬁcient means to display graphs (Battista et al.,

1999). The difference to abstract graph data comes

from the fact that the vertices already have some kind

of spatial information attached which stems from the

inherent topographic dimension in the data. There-

fore, we are not allowed to freely choose a stan-

dard graph layout algorithm from a given reper-

toire (Fruchterman and Reingold, 1991; Kamada and

Kawai, 1989; Sugiyama et al., 1981) following aes-

thetic graph drawing criteria (Ware et al., 2002).

‘Metro map’ is a commonly used term aiming to

describe a schematic map, i.e., a diagrammatic repre-

sentation of some kind of geospatial data which was

epitomized by Henry Charles Beck (Garland, 1994).

There is an ongoing debate by commentators and his-

torians about the real origin of the metro map as it

is known today. One version is to suppose that Beck

himself had a signiﬁcant inﬂuence by George Dow,

who designed, among other ideas, a carriage diagram

for the London & North Eastern Railway company

(LNER). In Andrew Dow’s book “Telling the Passen-

ger Where to Get Off” (Dow, 2005), we can ﬁnd a fas-

cinating perspective on the evolution of diagrammatic

railway maps and the work of someone who, although

nowhere near as famous as Beck, made a number of

pioneering designs preceding Beck. These map de-

signs are already some kind of information visualiza-

tion since their common goal was to graphically de-

pict data in a sense to make them understandable and

readable in a rapid manner. The difference to today’s

map design is that the map creators in these early days

had to do map research without the help of computers

and the pre-knowledge acquired on the internet.

Today the term metro map is widely used but

maybe the term public transport map may be the bet-

ter choice since there is a variety of transportation

modes and systems combined in a single map like

tubes, metro lines, buses, water buses, and so on. Al-

though such combined maps are beneﬁcial, they eas-

ily produce an information overload resulting in vi-

sual clutter (Rosenholtz et al., 2005). This high de-

gree of complexity may lead to the assumption that

some commentators and aﬁcionados would argue that

they are not even maps at all and should be referred to

as diagrams (Horne, 2012).

For example, the London Underground map has

been known, at different times during its history, to

refer to itself exactly as a diagram of lines. In any

case, it no longer portrays purely a metro system ei-

ther now. The most recent editions include suburban

surface rail (the so-called London Overground), light

rail (Docklands), and even a cable car (so-called Emi-

rates Air Line) across the Thames river. In an exten-

sive collection of maps, Ovenden brings together ex-

amples from around the world (Ovenden, 2005).

To really understand the usability and readability

aspects in metro map systems, i.e., graphical depic-

tions of transportation system data, user studies have

been conducted. Eye tracking can be used as one tool

(Burch et al., 2014b; Yarbus, 1967) to analyze where

and when visual attention is paid (Kurzhals et al.,

2014) to a map and which visual task solution strate-

gies (Burch et al., 2013; Netzel et al., 2014) are ap-

plied to perform a route ﬁnding task. From a graph

visualization perspective, such route or wayﬁnding

tasks have already been explored (Huang, 2007) with

the main result that people tend to follow the geodesic

path tendency (Huang et al., 2009). This tendency is

also present in metro maps (Burch et al., 2014a) but

IVAPP 2016 - International Conference on Information Visualization Theory and Applications

198

here, the original task is subdivided into subtasks be-

tween the interchange points. Each subtask is per-

formed by following the geodesic path and ﬁnally

cross checking the answer for correctness. Moreover,

it was found out that the map complexity has an inﬂu-

ence on task completion times.

3 DESIGNING METRO MAPS

Designing metro maps is a challenging task since the

designer must be aware of two major aspects: the de-

thetically appealing to a traveler and second, the de-

as a means to effectively and safely travel in a foreign

city (Burch, 2015).

3.1 The Designer’s Perspective

Let us take a look at an example after the completion

of ground research. Figure 4 shows a section of the

Communicarta Style45

 map for the city of Soﬁa.

Generally, in transport mapping we must be aware

of the information a traveler needs to make success-

ful journeys. There is a bare minimum and, for some

strange and unknown reason, not all transport author-

ities deem this part important. That perhaps, is for

another discussion. For now, however, we can agree

that a traveler will need at least three pieces of infor-

mation for stress-free wayﬁnding that are at ground

level, on and inside the vehicle, and on the map (see

Figures 2, 3, and 4):

• At Ground Level (Figure 2): If people walk

around in a city, the most important aspect

when using public transportation is the easy

recognition of the metro, tube, and bus stops.

Without them, the metro map is useless even if

it is well-designed.

1. Boarding Point: The name of the stop should

ideally be written on the stop sign and should

be the same as in the metro map to avoid mis-

interpretations (Figure 2 (a), (1)).

2. Interchange Point: If the stop is an inter-

change point, this should be already indicated

on the stop sign. Again, ideally the name of the

stop should be written on the stop sign (Figure 2

(a), (2)).

3. Where to Alight: For passengers, it is very

important to directly see where a metro line is

going to. This gives them the information in

which coarse direction they will be going even

(a) (b)

Figure 2: At ground level: (a) A multi-route nameboard:

Note the red number (1307) above the stop name, which is

unique for this platform. Each platform has its own unique

number. (b) Central Soﬁa: Typical tram stop and name-

board position.

(a) (b)

Figure 3: On and inside a tram: (a) Front of a tram dis-

playing route number (in this example route 14), the service

identiﬁer, and destination names. The destination name of

both ends of the line are shown. Ideally, only the actual di-

rection the vehicle is traveling in should be shown. (b) The

inside signboard of a tram where the red arrow shows the

‘direction’ the tram is going.

if the destination stop is not indicated (Figure 2

(a), (3)).

4. The Lines or Service Identiﬁers: Numbers,

names, letters, or symbols of routes for each of

the intended journey(s) should be clearly indi-

cated. A certain unique color for each metro

line should be used since color can easily be

perceived and mentally mapped to a certain

metro line (Figure 2 (a), (4)).

5. The Direction of Travel: The ‘towards’ or

‘destination’ of the services so as to select (a),

the correct platform, and (b), the right vehicle

traveling in the desired direction is also very

important (Figure 2 (a), (5)).

• On and Inside the Vehicle (Figure 3): Once en-

tered a tram, we are dependent on this ﬁrst de-

cision. Are we right? Is it going to the correct

direction? Do we have to change lines and after

how many stations do we have to do that?

1. Destination: The ‘towards’ or direction of

travel should be clearly indicated on the front

The Challenges of Designing Metro Maps

199

Figure 4: Information about the public transport system

in an easy-to-read and intuitive format: the Communicarta

Style45

 map. Note that all stops have both the names and

individual unique platform numbers.

of a tram, train, or metro (Figure 3 (a), (1)).

2. Service Identiﬁer: Also the service identi-

ﬁer, i.e., which line or route the train operates

should be given to the passenger before enter-

ing the vehicle (Figure 3 (a), (2)).

3. Inside the Vehicle: Clear announcements be-

fore each stop would be great, in addition, next

stop indicators in visual format would be fabu-

lous. Without these last two items, it is not im-

possible to get around it is just rather annoying

(Figure 3 (b)).

• On the Map (Figure 4): Having a metro map

in hand, a traveler is well-equipped to get not

lost in a foreign city—assumed the map is well-

designed and the visual encodings and designs

inside the map correspond to those given at

ground level and also on the vehicle.

1. Linking between the Stations: An easy-to-

read format has to be used for the visual encod-

ing of the linking of individual stations. From

a graph visualization perspective, a node-link

visual metaphor is suitable. Also laymen are

able to directly understand the principle of con-

nectedness when links in the form of straight

colored lines are used.

2. Topography: In many situations—even if trav-

elers are not familiar with a city—they already

know some points of interest like sights, the

airport, the main station, or a river. This to-

pographical information is important to rapidly

ﬁnd stations in a map.

3. Individual Lines: Color coding is typically

used to visually encode individual metro lines.

Keeping the same color coding for the lines is

important since map users build some kind of

mental map which helps them directly manage

the many map details.

4. Interchange Points: Traveling long routes in

a city oftentimes demands for changing lines,

i.e., passengers have to identify suitable sta-

tions where they could change trains without

wasting time or making the journey unneces-

sarily unpleasant.

Having pre-purchased a street map, printed it in

black and white for easy mark-up, we can commence

ﬁeldwork; to begin we need to establish the types of

transport and if possible highlight the routes on our

street map. Visiting the transport information ofﬁce,

we are sometimes able to obtain some information

about the routes. Now at least we may know the types

and extent.

3.2 The Visualizer’s Perspective

From a visualization expert’s perspective, metro maps

can be considered as node-link diagrams representing

stations as nodes and metro lines between stations as

links. There is additional data attached to the stations

and metro lines, e.g., categorical data expressing the

direction of travel under certain conditions or the time

periods a link (a metro line) is active. Additional in-

formation is placed on a metro map to guide a spec-

tator through the city or serve as a mental aid, i.e.,

the mental map. This is important in graph visual-

ization but also generally in all visual depictions of

data, in order to ﬁnd and observe visual patterns, re-

locate them, and ﬁnally compare them with others to

derive meaning and knowledge by remapping them to

the underlying data, i.e., the metro map system.

Interpreting a schematic map reliably and effort-

lessly is important to make it a useful guide for a

traveler in a foreign city. If there is only the metro

map with its lines and stations, it may be readable and

understandable, but the information it provides is not

enough to be really useful for the traveler. The map

designer has to outweigh the beneﬁts and drawbacks

of the design being sure to not end up at either end

of the design space, i.e., information overload is as

worse as information sparsity.

To fully understand how the end user of schematic

maps behaves we need user studies comparing differ-

ent map designs with each other. Although we can

measure accuracies and response times of the study

participants, we also investigate the applicability of

eye tracking techniques to explore the visual task so-

lution strategies, visual attention, and visual behavior

of people when working with the map. Consequently,

the map designer generates questions that the visual-

ization expert tries to answer.

From a visualization perspective, comparative

user studies are highly appreciated (Lam et al., 2012).

In the best case, recording the time-varying visual at-

tention paid to a metro map stimulus is very inter-

esting but such spatio-temporal eye movement data is

difﬁcult to analyze. The inherent space and time di-

IVAPP 2016 - International Conference on Information Visualization Theory and Applications

200

mensions in the data build another dependent variable

in a user experiment in addition to the traditional ones

like error rates and response times.

By inspecting eye movement data visually, e.g., by

visual analytics techniques (Andrienko et al., 2012),

we might be able to analyze the map designs for de-

sign ﬂaws, reading difﬁculties, or missing visual sup-

port like labels or legends. These evaluations can sup-

port the map designer in improving the maps or cho-

sen design.

4 SOFIA: A METRO MAP

DESIGN—REAL-WORLD USE

In public transport mapping, we must be aware of the

information a traveler needs to navigate around a sys-

tem. Here, in these illustrations, we can see how the

designed map links to the real-world situation, i.e.,

what the map user will see in the city. There is a bare

minimum and, we can agree that a traveler will need

these pieces of information for wayﬁnding (Figure 2

illustrates such a scenario at ground level): boarding

point, interchange point, where to alight, the line or

service identiﬁers, and the direction of travel. Note

that the last two points are particularly important for

stops or stations where multiple routes operate from

the same platform, as in this example here in Soﬁa.

Soﬁa has good coverage by trams with the major-

ity of platforms having signboards with both a name

and unique number. Those facts, combined with our

digital camera, proves very useful in saving time an-

notating the street map markup because, as all the

names are in Cyrillic, we do not actually have to write

the names down merely the numbers.

The signboards also show service identiﬁers to-

gether with the destination stations of each route. Not

only that, there are also timetables for each route, a

line diagram with each of the stop names in end-to-

end order and, highlighted in red, the name of the ac-

tual stop where the sign is located. This is a perfect

scenario. All very good information for the user and

for us is present to build a great map.

The service identiﬁers and destination names also

appear on the front and sides of most of the trams.

Even if they were not, so long as you have the map

and have noted the name at the end of the route in the

direction you wish to travel, you merely need stand

on the correct platform to make the journey.

Given all this information and in possession of a

well-designed and thoroughly researched Style45



transport map, it is truly a synch to get around

and enjoy the city, its suburbs and the surrounding

countryside—even if you do not read Cyrillic.

Strangely, despite providing all this really useful

navigational information on the ground, it is amaz-

ing why some transport authorities around the world

(actually quite a lot) choose to issue the most useless

map information to their customers.

Back in the ofﬁce after 5 full days of research and

having obtained all the photographic evidence, plot-

ted the stops and lines, it is time to complete the job.

We import the marked-up street map (Figure 5), over-

draw the geographic layout of each of the routes on a

‘trace’ layer, then draw each of the individual routes,

typeset the stop names, straighten the lines, stylize

with 45 degree angles, add station symbols and ser-

vice identiﬁer numbers.

In Figure 5, we can see all the individual tram

routes overlaying the street map we used for the re-

search, i.e., the network of interconnected tunnels,

lines, and stations. This ‘hardware’ is essentially in-

visible to passengers’ eyes. What we mean is that,

without good and clear map information there can be

no perception or indeed, comprehension on the part of

the customer as to where the network goes and how to

navigate their way around.

What brings the invisible to light? What makes

the wonderful network—all the hardware of overhead

wires, rails, tunnels, stations and vehicles—the bril-

liant engineering and the careful attention to safety

and good materials, actually of use to the traveling

public? The answer is a brilliant map. The kind of de-

sign that brings (carto)graphic reality to the user. The

map brings conﬁdence and connection to the user. It

is this graphic representation with all the careful skill

of design and attention to detail that makes the differ-

ence.

In Figure 6, all that was effectively hidden is now

revealed in all its glory. This is one of the key ele-

ments in customer information and whether it is de-

livered to customers on a website or as a paper give-

away, or indeed on digital devices like a smart phone,

the net effect is that the network makes sense, is visi-

ble, and gives customers the opportunity to travel with

knowledge and conﬁdence.

Maps are constantly being referred to. We have all

seen customers looking at maps and trying to make

sense of what they see. And indeed, the most im-

portant consideration when commissioning a map de-

signer is to make sure they have an understanding of

what works and what does not.

Working with visualization experts, the map de-

signer can gain deeper knowledge and insight of how

the end-user interprets the information and so can de-

sign more effective maps leading to better use of the

transport systems.

Aesthetics plays a major role in the design of

The Challenges of Designing Metro Maps

201

Figure 5: Research markup for Soﬁa showing the post re-

search ‘trace’ of the individual lines. The background map

is artiﬁcially blurred for illustration purposes.

metro maps. Without careful attention to detail and

careful design, the maps would be useless. Commu-

nicarta employs many of the ‘Good Design Princi-

ples’ (Ovenden, 2008), and more. Chief among those

being the horizontal placement of station names, en-

suring names do not cross over lines, smooth curves

as opposed to right-angles which aid, what we like to

call, the ‘ﬂow’ of the designs, and services separated

out into their individual routes as required.

As we also now know from our experience in

wayﬁnding and research, the direction of travel-

‘towards’ names and line or service identiﬁers play a

major role in the users’ ability to navigate the system.

Emphasis on these aspects are important elements in

the design.

And of course, famously, we use the 45 degree an-

gle for the diagonal lines which we have found work

very well for guidebook size pages which has been

the majority of the use of the maps.

Space is everything. This helps both the ﬂow and

the way that users are able to read the maps. Not only

does the correct spacing help the user the designs also

look and feel better.

5 CHALLENGES AND OPEN

ISSUES

There are many problems, limitations, challenges,

and open issues which could not be addressed in the

context of this paper. These aspects can be found on

both sides, i.e., either for the map designer but also

for the visualization expert.

Figure 6: The ﬁnal transport map of the city of Soﬁa in the

Style45 as designed for Thomas Cook City Spot guidebook.

5.1 Interactive Maps

Interactive maps are much more difﬁcult to design

since an infrastructure is required in the form of soft-

ware and data bases updating the displayed informa-

tion on user’s demand. However, interactive maps—

those on digital devices or computers—are a topic

worthy of note. They build the basis for a separate

subject altogether in the sense that the type of eye-

tracking study would need to be different from that of

a static kind of map.

That begins to open the topic up to a far

greater study realm since it would involve other ar-

eas such as user interface and user interpretation of

information—actually something we should be very

interested in, speciﬁcally in regards to public trans-

port, i.e., not just maps but wayﬁnding in general, also

in the areas of timetables, ticketing, vehicle design

(for example, where and what types of information

are placed on-board and outside) and even subjects

such as ‘how to use the system’.

5.2 Scalability and Visual Clutter

A real problem for both—the map designer and the

visualization expert—are growing datasets. For the

map designer much more work on recording data,

taking pictures, or traveling has to be done until the

ﬁnal map is ready for print. For the visualization

expert, larger datasets cause visual, perceptual,

and also algorithmic scalability problems, i.e., the

question comes up if the visual design can still keep

pace with the growing number of metro stations,

lines, and additional information.

IVAPP 2016 - International Conference on Information Visualization Theory and Applications

202

5.3 Progress of Transportation Systems

Public transportation systems are typically growing

with the increasing population in a city, requiring

adapted infrastructures to the novel situation. Stations

have to be added or new lines have to be designed

making a journey through a city more efﬁcient and

more pleasant to the passengers. The map designers

have to keep pace with the steady growing or change

of the transportation system in order to provide in-

formative and up-to-date traveling information to the

tourists. The map designer should be aware of possi-

ble changes in a transportation system, otherwise the

maps soon get obsolete.

5.4 Evaluation of Real-World Scenarios

In real life, passengers sometimes do not have enough

time to read a map very carefully. Typically, they are

in a hurry to catch the next train or metro which brings

them to the airport for example. This more realistic

scenario demands for another kind of user evaluation.

The question arises if maps in a certain design are still

useful in such situations where the traveler is under

time pressure.

5.5 Hazards of Map Research

As a public transport map researcher, we have to carry

a rucksack with all the equipment and reference ma-

terial. Spare cameras, spare batteries, spare memory

chips for the cameras, at least two sets of clip boards,

markup pens in a multitude of different colors, high-

lighters, torch for when it gets dark, all the street maps

that have been printed out in large A2 and A1 sizes of

various sections of the city to actually do the markups

on, wet weather gear where necessary, sun blocker as

necessary, drinking water, hand hygiene and a space

to store hats, scarves and gloves as necessary. Basi-

cally anything you would need for an expedition.

The life of a public transport map researcher is

not always plain sailing either. No, from freezing cli-

mates of -17

◦

C on one occasion in Berlin and Oslo to

the searing sun and humid highs of +50

◦

C in places

like Dubai to torrential rain and ﬂash ﬂoods of Tunis

and the fog of Milan, the researcher has to deal with

it all and still get the job done in the time allocated.

Sometimes the seemingly ridiculous (but true) oc-

casion of getting stuck in the Castelletto-Portello lift

in Genoa and having to be rescued, amazingly by an-

other lift in the opposite lift shaft—in case you were

wondering, are set alongside the more serious situ-

ations of being threatened with arrest in Rome (for

(a) (b)

Figure 7: A couple of amusing photo’s that aptly illustrate

what some of the hazards of map research are in some cities.

taking pictures in the metro) or being chased by hood-

lums in San Francisco and losing a day’s work in the

process, yes our researchers go through it all.

In other cases, it is just plain frustrating whilst

trying to get the job done. Like having to avoid

peak hours in cities like Beijing for example when re-

searching the trolleybuses, or being asked to delete

photo’s from the camera by security guards in Madrid

whilst working on line 12. We know there is a risk of

this, so we exchange the memory chip for a fresh one

every 2–3 hours so there is only that much loss if this

happens.

Weather, of course, plays a major part in our frus-

trations. Rain being one of the biggest threats to pro-

ductivity. In these situations if we have both metro

and surface systems to research we can choose to do

the metro ﬁrst and then go on to the surface system

when it has cleared up. Otherwise, it is a case of

ducking for cover and marking up the map without it

getting wet. All of this prolongs the time to complete.

In some instances, we wonder if we would ever

be able to complete the job before the system crum-

bles apart in front of us. St Petersburg tram system is

an example of this where you can see the state of the

tracks. In some cities, the state of the infrastructure

means we have to keep a look down instead of up.

Soﬁa is a great example of where not looking down

could cause more than just an embarrassment (see il-

lustrations in Figure 7). At other times, we have had

to contend with strikes and vehicle breakdowns.

6 CONCLUSION AND FUTURE

WORK

In this paper, we discussed and explained the art of

designing metro maps. We looked at this challenge

The Challenges of Designing Metro Maps

203

from two different perspectives, i.e., the one from the

map designer and the one from a visualization expert.

We identiﬁed different subgoals such as aesthetics,

understandability, intuitiveness, and readability in the

sense of user performance. A combination of these

goals can result in advanced metro maps serving as

good means to convey data but also to make metro

maps attractive to the eye and aesthetically appealing,

two major goals typically standing in a trade-off be-

havior. We illustrated the design of metro maps in a

real use case showing its design for the city of Soﬁa.

From a visualization perspective, user studies are per-

formed to measure the performance of people trying

to answer route ﬁnding tasks which depends on the

complexities of the maps that can lead to tremendous

design problems for the map designer, in particular

when many transportation systems like bus, metro, or

tube have to be displayed together.

For future work, we plan to collaborate, i.e., the

map designer raises questions that he is not able to

answer where the visualization expert designs user

studies with the goal to get hints about user perfor-

mance problems or even reading problems which can

be analyzed by applying eye tracking techniques for

recording and measuring time-varying visual atten-

tion paid to metro map stimuli. This is challenging

for static maps printed on a sheet of paper but be-

comes even more challenging for dynamic and inter-

active metro maps—maybe also on small displays like

smart phones.

ACKNOWLEDGEMENTS

We would like to thank the German Research Foun-

dation (DFG) for ﬁnancial support within project B01

of SFB/Transregio 161.

REFERENCES

Andrienko, G. L., Andrienko, N. V., Burch, M., and

Weiskopf, D. (2012). Visual analytics methodology

for eye movement studies. IEEE Transactions on Visu-

alization and Computer Graphics, 18(12):2889–2898.

Battista, G. D., Eades, P., Tamassia, R., and Tollis, I. G.

(1999). Graph Drawing: Algorithms for the Visual-

ization of Graphs. Prentice-Hall.

Burch, M. (2015). The aesthetics of diagrams. In Proceed-

ings of International Conference on Information Visu-

alization Theory and Applications.

Burch, M., Andrienko, G., Andrienko, N., H

oferlin, M.,

Raschke, M., and Weiskopf, D. (2013). Visual task

solution strategies in tree diagrams. In Proceedings

of the IEEE Paciﬁc Visualization Symposium, pages

169–176.

Burch, M., Kurzhals, K., and Weiskopf, D. (2014a). Vi-

sual task solution strategies in public transport maps.

In Proceedings of the 2nd International Workshop on

Eye Tracking for Spatial Research, pages 32–36.

Burch, M., Raschke, M., Blascheck, T., Kurzhals, K., and

Weiskopf, D. (2014b). How do people read metro

maps? An eye tracking study. In Proceedings of 1st

International Workshop on Schematic Mapping.

Dow, A. (2005). Telling the Passenger where to Get Off:

George Dow and the Evolution of the Railway Dia-

grammatic Map. Capital Transport Publishing.

Fruchterman, T. M. J. and Reingold, E. M. (1991). Graph

drawing by force-directed placement. Software, Prac-

tice, and Experience, 21(11):1129–1164.

Garland, K. (1994). Mr. Beck’s Underground Map. Capital

Transport Publishing.

Horne, M.A.C.(2012). Information design as-

pects of the London underground map. http://

www.metadyne.co.uk/UndMap.html.

Huang, W. (2007). Using eye tracking to investigate graph

layout effects. In Proceedings of the Asia-Paciﬁc Sym-

posium on Visualization, pages 97–100.

Huang, W., Eades, P., and Hong, S.-H. (2009). A

graph reading behavior: Geodesic-path tendency. In

Proceedings of the Paciﬁc Visualization Symposium,

pages 137–144.

Kamada, T. and Kawai, S. (1989). An algorithm for draw-

ing general undirected graphs. Information Process-

ing Letters, 31(1):7–15.

Kurzhals, K., Fisher, B. D., Burch, M., and Weiskopf, D.

(2014). Evaluating visual analytics with eye tracking.

In Proceedings of the Fifth Workshop on Beyond Time

and Errors: Novel Evaluation Methods for Visualiza-

tion, pages 61–69.

Lam, H., Bertini, E., Isenberg, P., Plaisant, C., and Carpen-

dale, S. (2012). Empirical studies in information vi-

sualization: Seven scenarios. IEEE Transactions on

Visualization and Computer Graphics, 18(9):1520–

1536.

Netzel, R., Burch, M., and Weiskopf, D. (2014). Compara-

tive eye tracking study on node-link visualizations of

trajectories. IEEE Transactions on Visualization and

Computer Graphics, 20(12):2221–2230.

Ovenden, M. (2005). Metro Maps of the World. Capital

Transport Publishing, 2nd edition.

Ovenden, M. (2008). Paris Metro Style: In Map and Station

Design. Capital Transport Publishing.

Rosenholtz, R., Li, Y., Mansﬁeld, J., and Jin, Z. (2005).

Feature congestion: a measure of display clutter. In

Proceedings of the Conference on Human Factors in

Computing Systems, pages 761–770.

Sugiyama, K., Tagawa, S., and Toda, M. (1981). Methods

for visual understanding of hierarchical system struc-

tures. IEEE Transactions on Systems, Man, and Cy-

bernetics, 11(2):109–125.

Ware, C., Purchase, H. C., Colpoys, L., and McGill, M.

(2002). Cognitive measurements of graph aesthetics.

Information Visualization, 1(2):103–110.

Yarbus, A. L. (1967). Eye Movements and Vision. New

York: Plenum Press.

IVAPP 2016 - International Conference on Information Visualization Theory and Applications

204