RAPID PROTOTYPING ON THE EXAMPLE OF SOFTWARE

DEVELOPMENT IN AUTOMOTIVE INDUSTRY

The Importance of their Provision for Software Projects at the Correct Time

Andreas Holzinger, Olivia Waclik, Frank Kappe

Institute of Information Systems and Computer Media, Graz University of Technology

Infeldgasse 16, A-8010 Graz, Austria

Stephan Lenhart, Gerbert Orasche

AVL List GmbH Graz, Hans-List-Platz 1, A-8020 Graz, Austria

Bernhard Peischl

Softnet Austria, Institute of Software Technology, Graz University of Technology, Infeldgasse 16, A-8010 Graz, Austria

Keywords: Software prototyping, Usability, User experience, User-centered design, Prototyping.

Abstract: Software prototyping is a powerful method for the identification of usability problems at the very beginning

of software development. This paper deals with the development of a prototype used for usability testing

and presenting it to stakeholders at the correct time. A low-fidelity (lo-fi) prototype was created for a

software product in the automotive industry, however the usability test was shifted to conduct it with the

latest build of the software application. This paper emphasizes on the effectiveness of prototypes together

with usability studies. It gives an overview about the experiences with usability testing on a high-fidelity

(hi-fi) prototype late in the software development process. The main conclusion is that we assume that

solving the usability findings of a hi-fi prototype is more difficult and expensive than using results from a

lo-fi prototype earlier. In future, we will conduct a lo-fi prototype usability study to confirm this

assumption.

1 INTRODUCTION

AVL List GmbH, an Austrian based enterprise in the

automotive industry, has developed AVL InMotion,

which is a sophisticated three-dimensional software

solution for maneuver and event based testing at the

test bed. It supports key business objectives such as

hybridization and electrification of power trains.

AVL integrated this standalone software into their

existing business software suite. In a first step, the

standalone application has been reviewed based on

usability heuristics in order to find out how the

application can fit into the existing software suite.

For this purpose a lo-fi prototype (mock-up) has

been developed (Brown & Holzinger, 2008),

(Holzinger & Brown, 2008) in order to verify the

graphical user interface (GUI) design proposals and

to identify the end user needs, along with the

weaknesses and strengths of the existing application

Target process Actual process

Third-party

simulation software

Delivery

Adaptation,

modifications

Integration

Lo-fi prototype

AV L I nM ot i on

Adaptation of

improvement

suggestions,

user needs

Main idea

Hi-fi prototype

Development

Usability Test

AV L PU M A Open

software suite

Third-party

software component

AV L I nM ot i on

Figure 1: Usability test - target vs. actual process.

Holzinger A., Waclick O., Kappe F., Lenhart S., Orasche G. and Peischl B..

RAPID PROTOTYPING ON THE EXAMPLE OF SOFTWARE DEVELOPMENT IN AUTOMOTIVE INDUSTRY - The Importance of their Provision for

Software Projects at the Correct Time.

DOI: 10.5220/0003524200570061

In Proceedings of the International Conference on e-Business (ICE-B-2011), pages 57-61

ISBN: 978-989-8425-70-6

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

(Figure 1). This paper emphasizes the importance of

prototyping combined with usability testing before

the implementation of a software product takes place

and to figure out the best moment in which to

present results to the decision makers.

2 BACKGROUND

After gathering the requirements for a software

application, a concept fulfilling these requirements

has to be developed. This concept has to be

implemented into a lo-fi prototype, which shows the

possible layout and structure of the system and

supports developers in getting an idea of the

interaction and design of the software product. It is

known that the exploration of different low fidelity

prototypes foster creativity (Fonseca et al., 2009).

Moreover, software prototyping is an effective

tool used to reduce risks and point out requirements

that should be considered (Hsia, Davis & Kung,

1993). It is an effective method to evaluate early

designs of the user interface (UI) and to get early

feedback from potential end users. Prototypes are

especially used to create an authentic UI for design

and evaluation in the early stages of a software

project (Holzinger, 2004).

The characteristics of prototypes range from lo-fi

to hi-fi prototypes. Lo-fi prototypes are created

quick and simple and are relatively cheap in

development. Both users and stakeholders (decision

makers) are able to get a first idea of how the system

will look (Casaday & Rainis, 1995), (Beyer &

Holtzblatt, 1999).

Prototypes that are developed with design

software (GUI Builder, cf. with (Spinellis, 2002) or

GUI Designers, cf. with (Gunderloy, 2005)) provide

GUIs with a slightly higher fidelity and are closer to

the final design. Furthermore, HTML, Flash or

Silverlight can be used for the creation of such

design concepts. Especially the latter is well suited

for 3D simulation software (Maiti, 2009).

In contrast, hi-fi prototypes are already very

close to the target system. These prototypes contain

functions that are integrated into the software

application in order to enable users to get a more

realistic impression of the software and to make it

easier for them to make decisions on the software

design (Rettig, 1994).

3 MATERIALS AND METHODS

3.1 Prototype Development

3.1.1 The Target System

The target system is represented by the software

component AVL InMotion that supports hardware-

in-the-loop (HIL) tests of hybrid vehicles at early

stages of the development process. The application

simulates virtual routes and maneuver based testing

at the test bed, as shown in Figure 2.

Figure 2: Maneuver and event based testing via AVL

InMotion.

3.1.2 Customization and Design

The integration of an external component into an

existing, complex software system usually comes

with problems. The reasons for this can be different,

for example, incompatible system architectures and

frameworks, or missing interfaces that prevent the

communication between two applications. However,

the problems need not be primarily technical in

nature. There can be communication problems

between developers or different goals pursued by

stakeholders (Larsson, Crnkovic & Ekdahl, 2004).

“The target is to integrate components into a

product and to ensure that the product works

appropriately so that it can be delivered to

customers” (Larsson et al., 2004).

In most cases, it is impossible to use a third-party

component without making adaptations to suit the

particular needs. (Szyperski 1997) discusses the

problems and implications regarding the

specification of functionality and quality attributes

as well as the challenges in the integration of third

party components. There is also a need for AVL List

GmbH to modify the purchased software application

for the integration into their proprietary existing

software system, especially considering that the

corporate design of the system must be consistent.

(Holzinger, 2005) emphasizes that a modification

and reengineering of the UI can be more expensive

ICE-B 2011 - International Conference on e-Business

and complex to set up in its implementation. This

can be avoided by considering user-centered design

(UCD) at the very beginning of software

development.

3.1.3 Prototype of AVL InMotion

The primary objective for developing a prototype for

the application AVL InMotion was the creation of a

user friendly GUI. In this context, the documented

requirements had to be analyzed to determine the

required parameters, which were necessary in order

to create a model for the design of the application.

The acquired, externally developed software

application provided many individual windows as

well as confusing menu entries, so that the goal was

to create a less complex UI for AVL InMotion by

considering the existing interaction design of the

AVL software suite and by following the AVL

corporate design guidelines. It was challenging to

cover the essential and most important functions of

the standalone application as well as to inspect each

window in order to create the new GUI for the

software component AVL InMotion, which should

present all key parameters and collect all functions

in one common frame.

InMotion Vehicle

ISAC Vehicle

InMotion Driver

ISAC Driver

Select Driver and Vehicle

waclik, 2009

Simulation

Help OK Cancel Apply

Segments Movie InterfaceGeneral Settings

Segments

Maneuver

Road

Driver

Override selected Segment Attributes

Track Width

Margin Width

left right Friction

Friction Stripe 1

Friction Stripe 2

from [m] to [m] Friction

Segment Deﬁnition

Info (optional)

Type of Segment

Straight

0.0

410.5

Length [m]

Grad [%]

Slope [%]

Camber [m/m]

Bird's Eye View of Road

Object Deﬁnition

RECHA40 = 10

Start Offset [m]

PUMA AO Type

Name

Typ e of O bje ct

Activation Object

+ 10

System Channel

Set Value

AO Ed i t o r

Segments Overview

Straight 0.0 200.0

Nr.

Type St a r t Length

1 69° Right 200.0 81.9

2 Straight 281.9 220.0

3 47° Right 501.9 80.1

4 10° Left 582.0 91.6

5 Straight 673.6 101.0

6 17° Left 774.6 24.4

7 Straight 779.0 410.5

Stop -5

PUMA_AO +10

8 20° Left 1189.5 10.0

==== END ==== ====

AVL InMotion - Road Deﬁnition [Edit]

Figure 3: Screenshot of the lo-fi prototype.

The final lo-fi prototype of AVL InMotion (see

Figure 3) was developed with HTML, in

combination with a GUI-Builder tool.

This represents a quick method to prepare a

presentation of the first scenarios of the planned

software design.

Typically, these mockups are called “click

dummies”, which represent a prototype that looks

like the final software, but does not provide

specified functionality. The idea is to achieve a

perception test of the user interface by providing

little interaction opportunities for the test users. By

using this prototype, the test users were able to

navigate through the different main menus and

submenus represented by tabs and they were able to

interact with the main features provided by AVL

InMotion.

3.2 Experimental Settings

The goal of this usability test was to evaluate the

ease of use and simplicity of the software

component AVL InMotion and to get feedback

about the look and feel of the prototype. In addition

the planned usability test should verify the UI design

proposals. All findings and possible

recommendations for usability improvements should

directly influence the development of the application

AVL InMotion afterwards.

In practice this approach could not be followed

as supposed in advance. As the prototype was

presented to responsible stakeholders, they

immediately started implementing the proposed UI

design without testing it with representative users

before. Instead of using the lo-fi prototype for the

usability study, the fully implemented software

system of AVL InMotion was finally used for

testing.

3.2.1 Participants

During the testing phase, it seemed that five test

participants were sufficient to gain meaningful

results. All participants were AVL List employees

who had no experience with using the integrated

software component AVL InMotion, but already had

experience with parts of the existing software suite.

It turned out that these participants and their

experiences perfectly matched the ideal

representative user profile.

3.2.2 Test Procedure

Each test participant was asked to complete a

questionnaire before and after the test, in order to

acquire the necessary background knowledge for the

further course of the analysis.

After completing all the formalities, the test

participants had to solve 19 tasks in a feasible period

of time. The interaction of the test participants with

the UI was observed, and at the same time the test

RAPID PROTOTYPING ON THE EXAMPLE OF SOFTWARE DEVELOPMENT IN AUTOMOTIVE INDUSTRY - The

Importance of their Provision for Software Projects at the Correct Time

participants were asked to express all their thoughts

and actions during the tasks aloud. According to this

it should be mentioned that thinking aloud (THA) is

probably the most common technique in testing

usability (Holzinger, 2006).

The advantage of this

method is the documentation and traceability of

individual steps via the verbalization of the thoughts

of the test participants and so problems can be

systematically located and identified.

For further evaluation of the task completion, the

time the test participants needed to finish each task

was logged. For rating each specified scenario the

so-called “Time-to-Task” had to be analyzed. Figure

4 shows the test environment of the usability study.

Figure 4: A test person is using the application within the

usability lab.

4 RESULTS

A closer look is given at the number of main issues

that were found by the participants who tested the

software component AVL InMotion during the

usability test.

4.1 Identified Problems

The identified problems were grouped into

manageable categories and they were prioritized

according to their severity. The ranking of the

severity (Levi & Conrad, 1996) is listed in Table 1.

Overall, 35 problems were identified, whereby the

weighting of the severities is as follows: Four

problems were assigned to severity level 4, eleven

problems to severity level 3, nine problems to

severity level 2 and eleven problems to severity

level 1. No problem was classified as level 0.

4.2 Recommendations

A list of recommendations was prepared on the basis

of the results provided by the usability test and its

identified problems. In total, 16 recommendations

Table 1: Severity rating scale for usability problems.

Severity Description

0 I don’t agree that this is a problem at all.

Cosmetic problem only - need not be fixed

unless extra time is available on project.

Minor usability problem - fixing this should

be given low priority.

Major usability problem - important to fix, so

should be given high priority.

Usability catastrophe - imperative to fix this

before product can be released.

were identified and their categorization was as

follows: [R1] use of contextual menu, [R2] meaning

of labels, [R3] parameterizing of segments, [R4]

redesign of drop-down lists, [R5] behavior of input

fields, [R6] composition of driver input mask, [R7]

modification of the sequence of segments/

maneuvers, [R8] adding new segments/ maneuvers,

[R9] structure and layout of the entire dialog, [R10]

functionality and layout of toolbars, [R11] bird’s eye

view of road, [R12] functionality of the bird’s eye

view of road, [R13] maneuver settings, [R14] three-

dimensional movie, [R15] simulation of the 3D

movie in PUMA and [R16] online editing mode of

test run. For each of the previously identified

categories, a problem description and a possible

solution were presented to AVL.

5 LESSONS LEARNED

It is important to provide the developed prototype at

the correct time; in principal prototyping is an

attractive approach in gaining a quick and cost-

efficient impression of the possible design of the

user interface. However as the preliminary version

of the UI of AVL InMotion was created and

demonstrated to AVL, the implementation and

technical development followed immediately and it

was decided to conduct the usability test with the

latest build of the software application - without any

evaluation of its UI design based on various

usability criteria. After the completion of the

usability test, several issues were identified

concerning the usage and design of AVL InMotion.

This led to the conclusion that it is not always

beneficial to present a prototype to responsible

parties without including a user perspective.

Although solutions have to be developed in time and

within budget constraints, it is far more complex and

takes much more effort to implement the usability

issues too late in the software development process

ICE-B 2011 - International Conference on e-Business

when the development of the software applications

is nearly finished.

6 CONCLUSIONS

This usability study represents an essential input for

all further activities of the AVL InMotion

development. It shows the current weaknesses and

strengths of the application and even of the

companies’ software development process.

It shows that it is essential to integrate usability

engineering at a very early stage of development to

ensure cost-effectiveness and to even have the

possibility to integrate most findings into the final

product. This will of course involve a change from

an engineering-centered corporate culture to a more

user-centered one, but will further enhance the long-

term growth-perspective and customer satisfaction

of a product.

The responsible development departments were

extremely pleased with the results and plan to

integrate the recommendations into the next product

version. Increased user efficiency and reduced

development costs will ensure the long-term success

of AVL InMotion.

ACKNOWLEDGEMENTS

Special thanks to the profound support of the

software engineering team of AVL.

REFERENCES

Beyer, H. & Holtzblatt, K. (1999) Contextual design.

interactions, 6, 1, 32-42.

Brown, S. & Holzinger, A. (2008) Low cost prototyping:

Part 1, or how to produce better ideas faster by getting

user reactions early and often. In: Abuelmaatti, O. &

England, D. (Eds.) Proceedings of HCI 2008.

Liverpool: John Moores University (UK), British

Computer Society.

Casaday, G. & Rainis, C. (1995) Models, prototypes, and

evaluations for HCI design: making the structured

approach practical. Conference companion on Human

factors in computing systems. Denver, Colorado,

United States, ACM.

Fonseca, M. J., Jorge, J. A., Gomes, M. R., Goncalves, D.

& Vala, M. (2009) Conceptual Design and Prototyping

to Explore Creativity. In: Kotze, P., Wong, W., Jorge,

J., Dix, A. & Silva, P. A. (Eds.) Creativity and Hci:

From Experience to Design in Education. New York,

Springer, 203-217.

Gunderloy, M. (2005) Developer to Designer: GUI

Design for the Busy Developer. New York, Wiley.

Holzinger, A. (2004) Application of Rapid Prototyping to

the User Interface Development for a Virtual Medical

Campus. IEEE Software, 21, 1, 92-99.

Holzinger, A. (2005) Usability engineering methods for

software developers. Communications of the ACM, 48,

1, 71-74.

Holzinger, A. (2006) Thinking-aloud - eine

Königsmethode im Usability Engineering. OCG

Journal, 31, 1, 4-5.

Holzinger, A. & Brown, S. (2008) Low cost prototyping:

Part 2, or how to apply the thinking-aloud method

efficiently. In: Abuelmaatti, O. & England, D. (Eds.)

Proceedings of HCI 2008. Liverpool: John Moores

University (UK), British Computer Society.

Hsia, P., Davis, A. & Kung, D. (1993) Status Report -

Requirements Engineering. IEEE Software, 10, 6, 75-

79.

Larsson, S., Crnkovic, I. & Ekdahl, F. (2004). On the

Expected Synergies between Component-Based

Software Engineering and Best Practices in Product

Integration. EUROMICRO Conference, 430-436.

Maiti, A. (2009) Flight simulation in 3D Web Map

Services using Silverlight (TM) 3 to Reduce

Simulation Costs in Pilot Training Institutes. In: Ao, S.

I., Douglas, C., Grundfest, W. S. & Burgstone, J.

(Eds.) World Congress on Engineering and Computer

Science. Hong Kong, Int Assoc Engineers-Iaeng, 591-

594.

Rettig, M. (1994) Prototyping for tiny fingers. Communi-

cations of the ACM, 37, 4, 21-27.

Spinellis, D. (2002) Unix tools as visual programming

components in a GUI-builder environment. Software-

Practice & Experience, 32,

1, 57-71.

RAPID PROTOTYPING ON THE EXAMPLE OF SOFTWARE DEVELOPMENT IN AUTOMOTIVE INDUSTRY - The

Importance of their Provision for Software Projects at the Correct Time