Developer platform orientation

From c,mm,n

c,mm,n developer platform
Remmert Stipdonk Delft University of Technology LogicaCMG Stichting Natuur en Milieu
Project setup Main page Exploration Conceptualization Prototyping Testing Evaluation

Introduction

In the orientation phase relevant literature is taken under consideration and summarized into information that can be useful later on in the process.

Wikinomics

The word Wikinomics comes from the book by Don Tapscott and Anthony D. Williams^[1]

Open source software is a community based development process. With Linux (open source operating system) and Apache (open source web server) as two very successful eye catchers, there are already projects that proved themselves worthy. These projects have become a success because of a relatively new phenomenon: community based product development. Both Linuxand Apacheare not developed by an IT company, but by an online community of devoted software engineers, spending their time on a project because of their enthusiasm for the topic. Altogether, they form an alliance that is stronger than any company in the world. Not only because they outnumber companies like Microsoft in the amount of people working on the project, but mainly because they all share their knowledge and findings. This is the key element. Using the knowledge of others and sharing your own in return. This makes it possible to solve problems quickly because of the amount of knowledge available and develop a product in less time with less money that in the traditional way. Or stated in another famous way:

"given enough eyeballs, all bugs are shallow".^[2]

Other examples of successful online community that make use of collective intelligence or power products are Wikipedia, YouTube, Blender3D and Mozilla Firefox.

Of course, communities like these above do not come into being just like that. They have originated at some point, initiated by a person or a group of people. But for starters, there has to be a demand from a (niche) market. In the example of Linux, there were software engineers who wanted to have a free, reliable operating system (OS) that met their specific demands. The quickest way to achieve the realization of this OS was to make it yourself. Thus a lot of talking through bulletin boards was going on until somebody created a solid starting point for others to elaborate on. This was the software kernel, created by the Finnish student Linus Torvalds. After that, all over the world software engineers started to create drivers and other useful add-ons for the kernel and so Linux was born. However, the reason that Linux became a survivor is the involvement of third parties like IBM to make the OS stable and reliable. Community based product development seems to be self-supporting until a certain point, but always seems to need some strict coordination to get to the finish line. Either it starts with a coordinated product (Mozilla Firefox, Blender3D) or a certain amount of top-down management is implemented later on (Linux)^[3]. This tells us that an online community does not have to be unorganized. It can become an organized and structured group of people depending on its' goal.

Another important factor of an open-source product is that it consists of a “frozen” part and a “developing” part. To implement and develop a product simultaneously, one needs an Alfa and a Beta version. The Alfa version has been closed for development by the community, while the Beta version is being developed and is open for all new initiatives. In most of the time, the Beta version of an open-source product consists of many subdivisions. Everybody can work on his or her favorite subject and make use of their knowledge in the most efficient way. For example, with Linux you can work on the kernel, a driver for a certain kind of graphical card, or on the user interface. When you create and coordinate the subdivisions in the right way, you can create an efficient whole. To make an open-source product reliable, it needs testing. Because there is no company that is specifically responsible, this has to be taken care of by the community. The quality of the product comes with the amount of testing it has undergone and how thorough this has been done. The product has to be tested in every possible suitable environment. To make this a little easier, the product can be judged on beforehand by experts in specific fields. The criticism that Wikipedia has undergone all its life is that its quality does not meet that of official encyclopedias, because it is not checked by certified historians and other scientists. This is of course true, but in the case of Wikipedia the articles are probably checked by hundreds of historians in the community, instead of the tens of historians that an official encyclopedia can dispose of. To summarize, open-source community based product development should keep in mind:

• The demand of the market
• The involvement of active members
• The amount of active coordination it needs
• The structure of subdivisions
• The quality of the end product
• The transition from software to hardware

Web 2.0

The World Wide Web as we know it today is developing at a high rate. Only fifteen years ago it wasn’t even there yet, and today we walk around carrying it in our mobile phones. It plays a large role in our daily life, and email has replaced the physical mail as the main carrier of personal data exchange.

Web2.0 mind-map^[4]

The internet at this moment is more or less one-way traffic. Mostly, it has readable content in the form of text, images and videos. Most of us use it in a passive way and do not actively attribute to the content. The most important language on the net is HTML, the language in which all websites are written. There are many additional languages such as Flash to create more specific applications like the YouTube flash movies, or internet flash games. And there is JavaScript, PHP and XML to make websites more interactive. This brings us at an important moment in evolution. It seems that we are right at the turning point in internet evolution, changing from passive users to active participation. In the Netherlands, one of the biggest websites is Hyves, a social networking site like Facebook and MySpace. People create a personal profile, upload photos and make new friends trough the network. Then there is the rise of image content websites such as Flickr and video sharing website YouTube. The big difference with conventional websites is that users of the internet are becoming real users now; the make use of the website not only by looking at it but also by actively participate in it. Maybe most of the visitors of the YouTube website do not even have an account and just watch movies, they do watch movies that are added by some user. This new trend in webbing is called Web 2.0. It is a development phase of the Internet into more active participatory. Where we are now is just the beginning of this trend.

Maybe the best example to go into a little deeper is Wikipedia. Wikipedia cannot be named a website anymore. It is a network of content, a web of data, created by the world. It is everything that the ancient Alexandrians wanted to achieve when they started to build their library, except for one important aspect^[1]. It is not created by an elite group of experts, but by the common mass of people. Nonetheless, it was not initially created this way. Founders Jimmy Wales and Larry Sanger, wanted to create an open encyclopedia in the internet. This initiative was called Nupedia, where all articles would be written by experts and reviewed through a formal process. This initial idea was not very successful, having about fifteen articles after one year. The next step was to create a wiki-based website with the same goal, but now the articles could be written by anyone. Off course right from the start this triggered hesitation amongst many people because it is hard to check the quality of the content. Either way, the amount of articles grew exponentially. As we know today, Wikipedia is used by millions of people, in hundreds of languages and the quality of it is almost as good as official encyclopedias. The important key issue is that each article is viewed by many people every second, which means that mistakes, bugs, are discovered and corrected extremely fast. This makes Wikipedia a reliable source of information.

Next to active participation of the users of websites, Web 2.0 also represents the rise of ‘active’ internet languages, also called rich internet application languages. They make it possible to make the websites’ content rich; it provides more than just a plain layer of information. In general, these techniques focus on the relationship between different information sources, the relation between users, and the enhancement of the website interaction.

Social Community Laws

Community based product development all depends on the involvement of the members. The more they interact, share knowledge and contribute to the community the quicker and better the community as a whole can innovate. Just like any other community, there are certain laws one has to keep in mind when participating in a community. Even more important, these rules have to be kept in mind when starting up a community.

First of all, a community needs members. For the startup, a group of members has to be found that have a common goal, and are willing to cooperate and share knowledge. They have to be able to create something useful together. As soon as there is something to show to the outside world, the community can start to widen its circle with new members. The new members can all zero in on existing subdivision or start new ones. A common rule for starting up a new subdivision is to be with at least four or five members who are going to work on this subdivision actively.

This takes us immediately to another important law. There has to be some kind of active coordination. There has to be a person or group of people that guards the integration of the subdivisions and makes sure that all subdivisions are useful for the whole community. This “controller” is most definitely not the boss of the community, as there is none. But to create a good working environment for the members, somebody has to keep the overview. When the community is mature enough this controller can lay back a little more, since the basis of the product is stable and the organizational structure of the system is clear to everybody.

Cooperation Mechanisms

Hierarchical clustering of cooperation inducing mechanisms (colums) in 21 state-of-the-art websites (rows), by Fokker et al.^[5]From right to left the clusters of "user profiling", "social visibility", social distance" and "power of collectivity".

As more people are becoming members of the community, we need a system that keeps them involved into the project. Members can only be of real use if they keep coming back and keep interacting with each other. In the 2007 paper about the psychological backgrounds for inducing cooperation in peer-to-peer television^[6] relevant theories that could explain altruistic behavior are listed. Users of community websites sometimes act altruistic, and for a community to become successful it is important to know and understand how these mechanisms work. These mechanisms can be used to keep community members satisfied with the community and keep them coming back for more. Fokker et al^[5] found four clusters of mechanisms for cooperation that occur in existing websites like Last.fm and Wikipedia.org:

1. User profiling
   Creating trust among users by showing personal information
2. Social visibility
   Visualizing the individual behavior of users to others
3. Social distance
   Influence of the social distance of users on their individual behavior
4. Power of collectivity
   The mass decides what is popular and what is not

These mechanisms are all frequently used in several state-of-the-art websites. Some are very frequently used in specific websites, such as the social distance mechanism. This mechanism is used in social networking websites, but not very relevant for for instance Wikipedia. The usability of cooperation mechanisms depends on the type of community that needs to be developed. The power of collectivity on the other hand, is an important factor in all reviewed websites. Most of the times this mechanism is not consciously applied by the members of a community. The example of tagging, frequently used in many websites, is most of the times used by members to structure their own content. These tags can then be used to find similarities between members and in this way show the popularity or other metadata about the content of the website.

An important finding in this research is the importance of rewarding community members for altruistic behavior (what could mean that it is not altruistic behavior at all anymore). The mechanism of social visibility is based mainly of making visual what an individual has done for the community. This is not only a common mechanism in modern websites, but also in open-source software development communities, as found by van Wendel de Joode.^[3] In the section about "Names attached to improvement" in the chapter below you can read more about this. A community can only grow if the members attribute actively to it and if they interact with each other. This means they have to come back again and again to the community website to keep up to date and they have to stay interested. What we learn from the research is that community members find it rewarding enough to receive credit according to their contributions to the community. As long as you can grow into the community others will pay you with respect and honor you for your contributions. Like in the Linux community, as we will read further on, community members will find this rewarding enough and stay satisfied with the community.

Psychological Backgrounds

The interaction found in communities is in fact not so different than interaction in physical communities and thus we could look at them in the same way. The mechanisms discussed before, are social interaction mechanisms that are most of the times not consciously applied by the community members but on the other hand, they apply them in masses. It is an interesting question to find out which social psychological theories can explain this behavior. This is interesting because it might help in finding the right solution for the common interaction challenges of an online community. In the 2007 paper by Fokker et al^[6] about psychological backgrounds for inducing cooperation in P2P television, thirteen theories are found. In a following 2007 paper by Fokker et al^[5] these theories are linked to the cooperation inducing mechanisms. The thirteen theories are shortly explained first and in the scheme you can find the clustering of the theories to the mechanisms. These theories are taken from social psychology and evolutionary biology, and try to explain altruism.

Costly signaling theory

Hierarchical clustering of 13 theories from (social) psychology (columns) and cooperation inducing mechanisms (rows), by Fokker et al.^[6]

People sometimes donate money or other benefits to others. The donated money will not be coming back to the good person, so this can be considered an altruistic act. The person therefore will receive respect and credit by others, who know about the donation. Thus, public display of altruism gives people status and reputation making the person an interesting interaction partner.

Social impact theory

Characteristics of the group a person is in, can influence the individual behavior of group members. This can be subdivided into three characteristics:

• Strength of the group in terms of power and status
• Immediacy of the group; how close or far is the group to the individual member in space and time
• Size of the group; the larger the group is, the less responsible members will feel

Social exchange theory

When people give something, they expect something to be given back to them. This can be product for product, service to service or a mixed version. Next to that, esteem can also be regarded an exchangeable good.

Social learning theory

People tend to copy each others behavior and learn new behavior for themselves in this way. They observe others and use their behavior to function in future situations. People tend to copy behavior of others if these others are similar to them or if they enjoy a high status.

Social balance theory

People try to find balanced states in their relationships to be satisfied with the situation. Balance can be either liking a group you are in, or disliking a group you are not in. This is a balanced state you can be satisfied with. If the situation is not balanced, people will try to change it and act.

Group selection theory

Groups of individuals can compete with each other, just like individuals can. For individual members of the group, the group is the greater good that comes before the individual. This means that individual behavior should benefit the group, even to the cost of the individual member.

Social identity theory

The social value of the group an individual is in, influences the self esteem of the individual group member. If the individual group member is a member of multiple groups with a high social value, the member itself will have a high self esteem. People tend to keep having a positive judgment on the groups they are involved with and a negative judgment on the ones they are not in.

Theory of reciprocal altruism

Cooperation is only possible under one condition; people will keep helping an unrelated individual, unless they refuse to reciprocate.

Equity theory

People want to contribute to a group if they receive an appropriate share of the groups' resources.

Kin selection theory

People tend to display altruistic behavior to closely related people easier than to people that are less close related. This comes from the theory that people benefit from helping people with genes that are similar to theirs. If this person succeeds, their genes do and with their genes, yours survive too.

Theory of group cohesiveness

Groups of people tend to exert pressure on the individual group members to uphold social norms of the group. Cohesiveness has three determinants:

• Attractiveness of the group to its members
• Mutual attraction of those members to each other as individuals
• Mediation of individual goals by the group

With all three determinants on a satisfactory level for the individual group member, the group has a bigger chance to stay cohesive.

Content Management

The management of shared content is a key element of open-source product development. In the research that has been done on this subject, several mechanisms have been found to be used in the existing open-source software development communities. Many people, including the community members themselves, think that the community is self-organizing or non-organized at all. They think that the members just do as they feel like, and in the end everything automatically comes together. Though it surely is a very bottom-up organization, these open-source communities are not entirely without a given structure. In this chapter I will explain some of the mechanisms that were found by van Wendel de Joode^[3].

Elegance and Modularity

To start with, there are mechanisms that are implemented in the very core of the organization only to ‘relieve the need for coordination’ for the community members. This is important for the following reason. Community members do not know each other most of the times, meaning that they do not communicate often and especially not about other things than the open-source topic. Thus communication should only be used when necessary, to streamline the total process. Elegance is necessary so that the written code in open-source software is clear and unambiguous. Others should be able to understand it and work with it later on. Elegance thus means simplicity, making it possible for developers to work on each others code without asking personally for an explanation and helping each other out every time. If a code developer writes an inelegant piece of code, then either nobody uses it, or he has to explain what he meant every time somebody tries to work with his code. Modularity means simply dividing the software into smaller and independent parts, so that they can be worked on independently by different members and member groups. Members can develop pieces of code for one part of the software program, and at the same time other members can work on totally other parts, not having to wait for each other. The relations between the different parts are important though, and have to be stated out clearly. This will be discussed in another topic later on, called “Coding style guides’.

These two mechanisms are somewhat preventive mechanisms. They are meant to create a streamlined system providing the best possible structure for the members. Next, I will discuss some mechanisms that are used to coordinate the communication between different members. The mechanisms here are the ones that could be relevant in a car development platform too.

Concurrent Version System

The development of a product goes through many stages. It has some kind of starting point and hopefully an end point, but in between there are many versions that are semi-finished or parts of the product that need revising, etc. In software development, these versions are stored as data in a database. This is a kind of digital library. In open-source development, it is important that different members can work on the project simultaneously, and that they all work with the latest version. But, also older versions should be available in case some new patch (piece of software) proves to be unsuccessful. At the same time the source code should always be available but never to be damaged. To make this all possible open-source communities make use of a Concurrent Version System (CVS). This is a client-server repository, which makes this kind of tree-structure possible. To software developers this is a logical way of structuring their process, but it is a key element of open-source development. If the car development platform wants to be successful, a certain CVS has to be in it.

Mailing Lists

To keep all different members involved and to make it easy to follow and join discussions, open-source communities intensively make use of mailing lists. Most of the time there are many different mailing list, hosting discussing about specific topics. If a member is part of the development group that is related to this topic, he is part of the mailing list. Different members can ask questions through these lists and others quickly and honestly can give answers.

Manuals and Coding style guides

In open-source software development the unity of the end product depends on the unity of the written code. Everyone should speak the same language here. You could compare this with writing a book with others, were everybody writes a chapter. It would be very unclear and unreadable if every chapter has its own writing style, one being very childish and the other using lots of difficult words. It would be even more horrible when the chapters are written in different languages. To regulate this problem in open-source software development, there are always several manuals, all describing how to write code, and how to make it easy to make their packages ready to install. Next to that, they tell the members what kind of information they should add to their packages so the relation to other packages is stated out. As mentioned before, this is important because it makes it possible for the members to work independently, but at the same time develop an integrated end product.

To-do lists

Because there are always parts of the product that need development, many open-source communities make use of specific To-do lists. These lists come from members inside the community and create a very democratic mechanism. To-do lists are a way to state out where there are holes in the program which you cannot fix yourself for whatever reason. On the other hand it is a hands-on assignment for members that are eager to start with writing something. When a task is finished it is clear for everyone that this has been done, and that progress has been made. It gives a good overview of the remaining tasks.

Orphanage

As said before, most open-source projects are divided into smaller bits and pieces to make it chewable. If some member or group of members quits with the development of a package, or when the coordinator of this package quits, it becomes a so called ‘orphan’. It needs a new maintainer, which is stated out clearly. The content that has been created so far is still available to a new maintainer so that it is easy to adopt an orphan. Like with the to-do lists, there is a list of orphans that needs adoption.

Names attached to improvements

An important reason for community members to contribute to the community is the credits that they earn for their work. In many cases there is something like a credits-list, which contains the names of developers who have contributed to the development of the packages. This list also states which specific package has been made by which developer. For future developers, this makes it easy to find the person that is responsible for a certain change in the code and the reasons behind it. The responsibility for the created packages therefore lay with a person, and not an anonymous community. This mechanism makes it possible to fix problems with a part of the code quickly and improves communication greatly.

Involvement of manufacturers in user-centered innovation

In the world of product development there are several examples of manufacturers involved in the user-centered innovation. It has been kind of acknowledged that for user-centered innovation of physical products, professional manufacturers are needed for the realization of products^[7]. The main question in this case is how these manufacturers of products, services or processes can make profits out of their involvement. For many companies this question will remain unanswered, either because they do not put effort in finding the answer or because there core-business is not suitable for user-centered innovation. Generally there are three possibilities for manufacturers to get involved in user-centered innovation that have been explored at the moment.

OpenMoko mobile phone

• The production of user-developed innovations (1)
• The development of toolkits for user-centered product development (2)
• The selling of products or services complementary to user-developed innovations (3)

For every possibility I will show a short example.

1) Google presented their open-source, Linux based, mobile operating system Android in November 2007. The operating system is a basic platform for which users can create additional patches with the provided toolbox. This way, Google hopes to create user input and innovation/ At the same time the OpenMoko is being developed; an open source mobile phone developed by its users and strongly supported and initiated by the electronics company FIC.

2) Epic Games is the producer of the popular Unreal game series. This game is a shoot ‘em up up game, where players walk around in levels, performing assignments in teams. The amount of different levels partly determines when you are finished with the game. To increase the life-span of this game, Epic decided to add design-tools for users to create their own levels. This became a great success because the created levels could be uploaded to the site of Epic and downloaded by other users. For users, their game stayed interesting for a longer period of time and for Epic this was a way to keep selling the basic software of the game.

3) The American software company Red Hat, is a commercial distributor of a version of Linux. Clearly Linux is open-source and you can download user-friendly version such as Ubuntu online for free. Nevertheless these versions are unsupported and thus not so easy to install and maintain for non-software developers. Red Hat offers a commercial version that is easy to install and offers additional support and services for end-users.

What we see here is that it is possible for manufacturers to get commercially involved into an open-source project, if the product suits their capabilities. The intellectual property of the developed products remains under license of the community, but that does not mean that others cannot make profit out of it in their own way.

Automotive Development Process

The automotive development process is not so much researched in scientific ways. The process has not changed very much in the last years, except for the upcoming of digital CAD modeling. It seems that the automotive development process is not a process that is approached in a scientific way. Car designers are highly skillful people, better to be treated as craftsmen than as scientists. The initial styling phase of automotive development is a non-verbal process based on visual communication^[8]. This phase can be compared to that of an artist; the designer thinks of the design problem in an holistic way, and does not think in partial solutions on itself. On the other hand, the initial design problem is presented to the designers through a package presented by the engineering department, containing all technical restrictions.

The process

Developing a new car takes lots of time for a professional company. If a new model has to be developed, it has a time-to-market of about five years. The illustration shows the different design phases according to their moment in the process. On top, the design and styling part of the exterior and interior can be seen. Underneath, you can see how the development process of mechanical parts shapes itself. On the bottom you can find that for early testing, many prototypes are made. This can be prototypes of specific parts of the car, or multiple integrated parts together.^[9]

Due to the implementation of CAD systems, the development process can be finished in less time nowadays. With CAD systems, it is possible to do more virtual testing in an early stage. The schematic representation as seen here, represents the time-to-market for American production line cars. In Japan the time-to-market is about three years.^[10]

Actors in the present automotive design process

Exterior Designer

The initial exterior styling is designed by a single or small group of designers. They create several styling drawings based on the package material provided by marketing and technological specifications. Further in the process the exterior designer adjusts the design so that it fits changes that come up in the engineering phase.

Interior Designer

Almost parallel with the exterior styling phase, the interior designer starts with designing the interior based on the package material provided by marketing and technological specifications. The interior designer is mostly a styling job, but he works together with ergonomic experts and engineers so that every part stays integrated properly.

Engineer

The engineering department is a large department consisting of many different engineering disciplines. Vehicle dynamics, mechanical engineers, aerodynamics engineers, electrical engineers and software engineers are the most important. They are responsible for the design of all the different parts and the integration of them into each other.

Prototype Builder

For both the engineering department as well as the exterior and interior design department prototypes need to be made to make a well considered judgment. For the exterior these will be scale clay models, for the interior full scale models and for the engineering it will most of the times by full scale mechanical testing prototypes.

Expert

To prepare the testing phase thoroughly and to predict errors in the designs, experts are called in to analyze the designs in an early phase. Experts can be in: materials, vehicle dynamics, ergonomics, safety, manufacturing, legal, sustainability etc.

Stylist

The stylist is responsible for the styling the interior in terms of materials and coloring. The stylist will create a range of material and color combinations for the marketing department which will be communicated to the customer. They will be based on the specific target group the car is manufactured for.

Tester

Before the manufacturing process starts, several full scale, full running prototypes are made for extensive testing. Several testers will run specific test tracks to come up with recommendations for specific parts or the car in total. After the testing phase, engineering has to improve the car where needed. Testers are specifically educated people with a large responsibility due to safety issues.

Marketer

The marketing department is responsible for creating the automotive package; the requirements list with which the automotive design process starts. The base their findings on former car models and competitors’ developments and create a complete image of what the future car model has to become. When the car is ready to manufacture, they are responsible for marketing it and to create enthusiasm under the sales dealers.

Manager

At several key moments in the automotive design process managers are called in to make decisions. They are responsible for making the right choice and test it on the automotive package. Most of the times managers come from specific disciplines within the design process and are experts themselves.

Management intervention points

From sketch to clay model[1]

After the initial styling phase, many phases follow until the car is ready. Thus the styling phase is only a fraction of the total car development process. An important conclusion that has to be made concerning the automotive industry is that it is based on evolution rather than on revolution; every model follows another model directly, and is based upon the competitors’ models. The structural part of the automotive development process comes from the management intervention points. These are moments in the development phase where the design department has to present results to the management, and decisions are made. ^[8]

Full-size clay model of a corvette[2]

1. Issue of brief and product specification / Issue of package
2. Review of competition and influences
3. Informal selection of concept sketches
4. Management review of concept proposals
5. Tape drawing presentation
6. Scale model presentation
7. Presentation of reworked tape drawings or scale models
8. Presentation of full-size clay model
9. Re-presentation as required and approval of 3D model

Issue of brief and product specification / Issue of package

Example of a package drawing; passenger hight and other dimensions are specifically stated

The design process starts with a pile of information provided by the engineering and planning departments, stating the framework of the new car. The package consists of a “full-size three view layout giving the mechanical, ergonomic and legal parameters of the proposed vehicle”^[8]. This gives the designers handles to start with the design phase.

Review of competition and influences

Concept sketch of an Alfa Romeo [3]

Informal discussion of concept sketches

To get the design process really started, the design department has to present its’ initial concept sketches to the management. Together with this, there will be a presentation about the competitors’ car models. This way, the designers are aware of the present status of competitors and know what they have to beat in terms of design. The management team use this moment to state their view on the new car, by discussing the initial sketches.

Informal selection of concept sketches

This is another moment for the management to take a look at the developed sketches, and for the designers to receive some feedback. The management has an opportunity to see where things are heading and if needed they can adjust the course of the design team. The concept starts to develop an identity.

Management review of concept proposals

In a viewing room with several displays, management and possibly the designers themselves are reviewing the concept proposals, developed by the design department. This is an important decision point in the development process, because decisions are taken about the design theme, the visual identity and the overall character of the design, used in the rest of the process.

Tape drawing presentation

A tapedrawing of a car

From the chosen concept proposals the design department develops three different tape drawings. Tape drawings are full scale side view drawings, showing the contour of the car. From these alternatives, the management can pick the one they want to develop further. Most of the time, the chosen concept incorporates aspects from the other concepts. With the tape drawing, the management can also judge whether or not the concept fits the initial package.

Scale model presentation

Clay model of the Shelby GR1, covered with metal sheet

The same decisions can be made as with the tape drawing presentation, only now with 3D scale models. Presentation of reworked tape drawings or scale models After this presentation, the design department will go on making a full scale clay model of the car. Before this is possible, everybody has to agree on the design decision taken. Therefore a full scale tape drawing of the three views of the car are presented, sometimes together with a 3D scale model. This is the result of the modifications made after the first tape drawing presentation. In this phase the management can either be satisfied and let the design team go on to the next phase, or they can have them work again on this model and plan a new tape drawing presentation until they are satisfied.

Presentation of full-size clay model

Highly detailed clay model of the Shelby GR1

In a viewing room, a full-size clay model of the car is shown to the management, together with its main competitor. The management will be judging important things in this phase. They will judge whether the design will fit the initial package requirements and the concept drawings they have seen before, and make recommendations for further development. Most importantly, the management will judge if the design is attractive, admirable, if it will stand out in its future market. To do this, the clay model looks almost like a real car, imitating wind screens and painted in a way that it looks metallic. When the model is approved here, it will continue in the process for detailed engineering and productionizing.

Representation as required and approval of 3D model

The recommendations made after the presentation of the clay model are incorporated in a revised model and presented for approval. After this phase, a model will be made for presentation purposes to communicate the concept to other departments. From here on it will be developed into a real prototype, ready for testing.

CAD modeling in Automotive Design

In the world of automotive design, the last fifteen to twenty years CAD modeling has become an essential part of the design and delivery phase. CAD (Computer Aided Design) is literally the help of computers in design. In practice, it is the translation from sketch on paper, 2D, into a computerized, geometrically precise, 3D model. Over the years, the importance of CAD in design has grown due to improvements in the software industry. The CAD modeling programs have become more user-friendly and are more sophisticated nowadays. For designers CAD modeling now comes very early in the design phase, sometimes as early as the sketching phase.

In CAD modeling, every individual part of a design is modeled separately. Afterwards, the individual parts are connected together into an assembly. In this phase, the specific relations between the individual parts become critically important. Another large benefit of CAD modeling is that even in the digital phase of product development, the parts and assemblies can be tested for real-life characteristics. Parts cannot stick into each other and they can be tested on strength and dynamics and so on.

After the design has been modeled successfully in CAD, technical drawings can be extracted from it for manufacturers. Even digital models from CAD can be imported directly into milling machines for building the parts.

CAD software

CAD software is used in virtually all branches of design; aerospace engineering, automotive, architecture and industrial design are examples of disciplines where products first come to life in a digital world and after its success there, it is built in real life. To provide the design departments with the right tools, there are different forms of CAD modeling and many different CAD software providers. The most important difference in this case is on one hand solid modeling and on the other hand free-form modeling (surface modeling). I will explain both types with an example.

Solid Modeling

For the creation of parts that need mass properties, so that they act in the digital world as they would in the real world, solid modeling is needed. Mass properties are needed to make the model as realistic as possible. Without mass properties, a vehicle dynamics expert cannot perform calculations on the model for instance and manufactures cannot predict what the costs will be to manufacture the part.

Example: This model of a car brake light is an assembly of multiple solid parts with mass properties, thickness and material properties.

Freeform Modeling

If the designer does not need mass properties but does have to digitalize a design with difficult form properties, freeform modeling is commonly used. As the name says, it gives the designer the opportunity to shape the form freely, creating difficult, three-dimensionally curved forms. It is also called surface-modeling, because the designer works with surfaces with an infinitely thin thickness. This is not realistic, but it does make it possible to experiment a lot more with the form. In design worlds such as computer game design and computer animated movies, this is a commonly used technology. On the other hand it is also used in some phases of industrial design.

Example: With freeform modeling the user can edit a surface with control points that have specific relations to each other. The computer calculates what the changes in relations do to the shape of the surface. The designer can connect different surfaces to each other, where the connection surface is computer generated. Rendering these kinds of shapes can take a lot of time.

Automotive Industry

Both solid modeling and freeform modeling exist in the automotive design industry. The eventual CAD model that is needed to manufacture the car parts is a solid model with all the required characteristics. Freeform modeling is used in the design phase, for example for styling in the design of the exterior. The most commonly used software program in the automotive industry is CATIA from Dassault systems. This program combines different forms of CAD and can provide software that is adjusted to the specific field of expertise the designers need. It is also commonly used in aerospace engineering. The advantage of using one CAD software system in different fields of expertise is that it creates better communication possibilities due to the given standards.

Sharing CAD models

As mentioned above, communication is an important issue in the CAD world. There are many parties involved in on part of the design model. There is the designer, the engineer, the materials expert, the OEM, the marketer and so on. It is very important that all parties can use the digital CAD model in its original format. If every party uses the same software program (for instance CATIA) this is not a problem. If not, it is a large problem to overcome.

For CATIA, IBM has developed a web-based collaboration platform called 3DLive. This platform gives insight in all individual parts of an assembly, as it does in the assembly and sub-assemblies itself. All external data such as test data can be integrated, so that the complete digital product can be experienced web-based by all involved parties. Nonetheless this is only suitable for digital models developed in CATIA.

In the summer of 2007 Adobe introduced a new PDF format: Acrobat 3D. All CAD formats can be exported to this PDF format and it has two large benefits: it reduces file size dramatically (up to 150 times less than the original CAD file) and it is readable by anyone with Adobe Acrobat 8 or higher on his or her computer. The PDF has all the properties the original CAD models have except the design history; you can measure it, you can make cross sections, you can export to STEP and IGES models and you can isolate parts or sub-assemblies from assemblies. This makes it ideal as a sharing format for anyone who does not directly has the need to edit the model, but does have to review its characteristics.

References

1. ↑ ^{1.0 1.1} Don Tapscott and Anthony D. Williams, Wikinomics, ISBN 1591841380
2. ↑ Linus' Law, stated by Eric S. Raymond in The Cathedral and the Bazaar, ISBN 1-56592-724-9
3. ↑ ^{3.0 3.1 3.2} Ruben van Wendel de Joode, Understanding open source communities, 2005, ISBN 9056381385
4. ↑ Markus Angermeier : Web 2.0 Mindmap Translated versions
5. ↑ ^{5.0 5.1 5.2} J. Fokker et al, Inducing Cooperation in Peer-to-Peer Television Systems, 2007
6. ↑ ^{6.0 6.1 6.2} J. Fokker et al, Psychological backgrounds for inducing cooperation in peer-to-peer television, 2006
7. ↑ Eric von Hippel, Democratizing innovation, The MIT press, 2006
8. ↑ ^{8.0 8.1 8.2} Michael Tovey, Intuitive and objective processes in automotive design, Design studies vol.13 no.1 January 1992
9. ↑ P. Bertholon, L. Gaillartdot, G. Olivier, Les Metiers du Style Automobile, 1988, ISBN 2-7268-8061-4
10. ↑ Elmer D. van Grondelle MBA, Prof. Ir. Matthijs B. van Dijk, Automotive Design, 2006