Ethical issues in engineering design
Trang 2Ethical issues in engineering design; Safety and sustainability,
Proeféehrift
terverkriiging van de graad van doctor aan de Technische Universiteit Delft,
cop gezag van de Rector Magnificus prof dr ir .T Fokkema voorzitter van het College voor Promoties,
in het openbaar te verdedigen op maandag 14 november 2005 om 10:30 uur door Anke Christine van GORP
Trang 3Dit procfschriftis goedgckeurd door de promotoren: Prof dei PLA, Kroes
Prof de Mj van den Hoven Samenstelling promoticcommissie
Rector Magnificus, voorzitter
Prof dr ir P.A, Kroes, Technische Universitet Delft, Promotor Prof dr Mj van den Hoven, Technische Universiteit Delft, Promotor Prof de A, Grunwald, Forzungszentrum Karleuhe GmbH
Prof đc B.A.W Musschenga, Vie Universitet Amsterdam Prof dei P Kruit, Technische Universiteit Delft
Prof ir A Reukers, Technische Universiteit Delf Dr H, Zandvoort, Technische Universitet Delft
Dr ir LR van de Poe! heeft als begeleider in belangrjke mate aan de totstandkoming, san het proefschrift bijgedeagen
Trang 5‘Simon Stevin Series in the Philosophy of Technology
Delft University of Technology & Eindhoven University of Technology Editors: Peter Kroes and Anthonie Meijers
Volume 2
{© Anke van Gorp, 2005
All sights reserved, No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without prior permission in writing ofthe publisher
mak anke_van_gorp@yahoo com ISBN-e: 0o-getgge7-t
Trang 6Contents
1 Introduction 3
1 Research question
spans 2.11 Ethics indesign processes 2 a
ava Design process 25
sa Delign problems 2.23 The design process as agcial process 29 2
2.2.4 Organisation ofthe design process 30
2.4 Characteristic 12.1 Design toe and design hierarchy of design processes in relation to ethical soues a 2 3 52a Normative frameworks 2.3.3 Moral responsibly and the tus eaonship between 6
engineers and society 2
‘a4 Summary ra
troduction tothe cas
dies 4B
saWorking hypotheses gs
5 Acquisition of empirical data +6
4 DutchEVO, safe or sustainable? 49
1A light family car °
Trang 7Ethical issues in engineering design lation PE ene ee) mm ze 6 45 Suramar ofthe cee and We aod STR > 4.6: Ethical issues ra 4.64 Decision snaking on Aa] nna 3 4.6 Regulative framework 74 4:7 Acknowledgements z6 4 Piping and Equipment 2
‘5.1 The design of a (petrojchemtical plant 7
5.2 Regulation regarding pressuze 5.21 Regulation
5.2.2 Codes regarding pressure vessels and piping $2 Sands negarding posure waned Sd pipe 8z Br 5.3 Clear responsibiliies and tasks 8 $a Ethical 84 5.5 Summary g1 EHhiolisgoek ofthe case and the regulative framework se ot 5.5.2 Decision making on ethical issues 2 $.5.2 Regulative framework 2 3 6 Acknowledgements s 6 Designing a Bridge 9!
6.1The design problem = 95 6.2 Trying o reconcile all requirements and suakeholders 98 6 Legislation and codes 6.3.1 Safety during construction se» 102 Gp Sittin 3e 6.33 Sustainability tre 6.4 Responsibility nd lability ma 43 (65.1 Bhi iues ig 6.5.2 Decision making on ethical i n6 6.5 Requlative frameworks “ry 6.6 Acknowledgements nt 2 Digna tightest đi ‘21 A lightweight truck trailer tr
Trang 8Contents
17.3 Safe in what sense? s
1734 Structural reliability aa
75.2 Misuse and overloading 41
74 Aseribing responsibilities 143
12.5 Summary ofthe case and the regulative framework Hà
17.544 Ethical issues way
76 Acknowledgements 151
8 Conclusions of the empirical study 25;
sts an ndhirarchy ng 53a resolve ethical issues and design type and
hierarchy $8.51 Decision makers and design type and hierarchy 260 162 8.4 Regulstive fameworks 165 85 Design problem formulation 167 8.6 Generalisstion ofthe conclusions 169
8 Summary ofthe te
‘9 Towards warranted trust in engineers 1
‘94 Normal design 1
‘9.11 Required competence of engineers in normal design 176 9.1.2 Grunvwald’s requirements reconsidered Hy
9.2 Radical design a
‘9.2: Mentifying what affected actors value 182
19.3 Further research 186
9.4 Recommendations for engineering education 387
Trang 10Introduction
On March the 6", 1987, the rll-on/rolloff (ro/r0) passenger and freight ferry the Herald of Free Enterprise capsized just outside the Zeebrugge harbour.’ Water rapidly filled the ship, leading to the death of 150 passengers and 38 ‘crewmembers The main cause of the disaster was that the inner and outer bow doors were open when the ship left port
‘The assistant bosun should have closed the doors, but he had fallen asleep “The absence of warning lights made it impossible to see from the bridge whether
the bow doors were closed On at least two previous occasions, similar negligence ‘with sister ships owned by the same company had led to the shipe leaving port ‘with the bow doors open These incidents however passed without disastrous
results [London Crown, 1987]
Pressure to depart quickly and poor communication had contributed to leaving port with the bow doors still open in the case of the Herald As is often ‘the case, it was human error that preceded the disaster, but it was the design of
the ferry that made the occurrence of such a disaster possible in the first place It ‘was the inherent instability that ro/ro ferries encounter when water enters a deck that played an important role in the disaster This is an aspect of ship design It might be expected that while designing the Herald and her sister ships, the ‘engineers were aware that if water were to flood on the decks the ship might quickly become unstable Following the Herald disaster there was a similar disaster with another ro/ro ferry, the Estonia, Water filled one of the decks and the ship capsized killing nearly 800 people This happened despite the fact that Estonia's owners had complied with the proposed new regulation concerning ro/ro ferries formulated after the Herald disaster [Van Poortliet, 1999]
In the following, more detailed description of the Herald disaster I will focus ‘on decisions made in the design process that made the ship very vulnerable to ‘water flooding the car decks, This example demonstrates the possible ethical
impact of decisions made in the design process
‘One ethical question that arises in relation to the design of the Herald of Free Enterprise, and other ro/ra ferries, is whether it should have been designed to be riore safe given the fact that it was known that water entering the deck might result in rapid capsizing This is a moral problem because passengers, crew and
"This description othe Herald of free enterprise disasters based on ‘nsineeing design processes” Van Gorp and Vande Poel, 2003],
Trang 11
Ethical issues in engineering design
their families are harmed when a ship capizes There were, and age, simple technical solutions if one wants to prevent rapid capsizing when water enters a deck Bulkhieads created on the decks can easily prevent water from owing freely over a deck and prevent rapid capsizing [wwwsafetyline wa gov, 2005] Bulkhheads on decks, however, give rise to longer loading / unloading times and take up space on the decks, hence this costs money
‘When we look at ethical problems in relation to the design of the Herald and comparable ships, ethical issues become relevant at different stages of the design process and during the use of the product Ethical issues are relevant during the formulation of criteria and requitements for the design and in the acceptance of tradeoffs between requirements, 1 will focus on the formulation of safety requirements for a ro/ro passenger or freight ferry, and on the trade-offs that exist between safety and economic requirements, This description will explain why ro/ro ferries were not designed in a way that would prevent rapid capsizing if water floods a deck
When it comes to formulating legal safety requirements, the International Maritime Organisation (IMO) has an important past to play This international organisation is responsible for adopting legislation for ships IMO's safety legislation deals with the ship and passengers The SOLAS (Safety of Life at Sea) convention is especially concerned with passenger safety and with lifesaving equipment on passenger ships IMO officials knew as early as 198r that if water centered the car decks of a ro/ro ferry, the ship could be lost in a rapid capsize [Van Poortliet, 1999, 52], Water entering the car deck will flow to the lowest point leading to a greater inclination, resulting, ifthe inclination exceeds a certain, angle, in a rapid capsize This fact has been regarded as common knowledge in the maritime world, t least since 1981 The IMO did not adjust its regulations to solve this problem, even though simple technical solutions, e.g, bulkheads, were available
Legislation adopted by the IMO needs to be implemented by governments, and only governments accepting the IMO convention will implement it, Thus when making 2 convention, it is important to make it acceptable for as many governments as possible, otherwise only a small percentage of all fleets will be obliged to abide by the convention A shipping company can decide to sail under the flag of another country which has not ratified an IMO convention, if, in the opinion of company management, complying with the convention will costa lot of money So there is a certain amount of pressure on the IMO not to issue safety requirements that are considered by some governments to be too tight or too costly to implement
Trang 12Introduction Most IMO conventions affect new ships but do not apply retrospectively to shipe already at sea, This is known as the grandfather clause The grandfather clause protects the pooter states, because for them: it would be too costly to adapt their older fleets to new legislation IMO legislation may thus be said to be weak and 10/ro vessels complying with IMO legislation are still prone to rapid capsize
‘Apart from the IMO, insurance and classification companies also have a part to play in the formulation of safety requirements To obtain hull insurance from, insurance companies such as Lloyd's of London, a ship needs to be classified, Classification organisations ate private organisations that monitor compliance with legislation during construction and certify sea worthiness during a ship's lifetime Only the ship's equipment and construction are taken into account by the classification organisations, they do not deal with passenger safety [Van Poortvliet, 1999),
‘There is litle incentive for shipping companies to ask for, or for shipyards to design, ships that are safer than required by IMO conventions and hull insurance regulations When disasters occur the investigation that follows usually concludes that it was a human error that led to the disaster Little attention is given to the design of the ship as long as, on completion, the ship complied with the current regulations of the time
Six actors are important in the formulation of the safety requirements laid down for rojto ferries, These actors are: the IMO, governments, insurance companies, classification companies, shipyards and shipping companies To understand why these six actors have not formulated tighter safety requirements, its important to realize that when safety requirements are formulated 2 trade-off is made with ‘economic requirements
Economic considerations are important for insurance and classification ‘companies because they depend on shipyards and shipping companies When the safety requirements they impose are more costly than those of competitors they will lose customers, Insurance companies will want the requirements to be tight enough to prevent them from having te pay out too frequently for hull lose and damages However, they usually do not want to impose more or tighter requitements than their competitors as they are aftaid of loosing thei customers
Shipyards do not have loyal customers, To be competitive the price needs to be kept as low as possible or at least lower than that of the competitors Safety measures are usually only built in when there is a legal obligation to do so, Shipyards may not be held liable if, at the time they were built, their ships ‘complied with the relevant legislation
Trang 13Ethical issues in engineering design
loading times In the ease of ro/ro ferties, chipping companies do not want to hawe bulkheads on the decks because it takes time to put them in place while loading the ferry Moreover, fewer cars can be transported because the bulkheads have to be designed in such a way that larger and smaller cars can fit between thems; this requites spacing between the bulkheads that is less efficient if large cars and fieighttrucks are considered So shipping companies also trade off safety against economic considerations
Finally, the IMO and governments of individual countries also trade off safety considerations against economic ones As we saw earlier for IMO conventions to be effective as many countries as possible have to support them For many countries, economic considerations will play an important part when it comes to deciding which safety requirements they consider acceptable This is reinforced by the fact that shipping companies can choose which flag they sail under Governments could forbid ships that do not mieet their stricter national regulations from entering their harbours There are economic reasons not to do this A government’s national harbours, where stricter regulations are enforced, will have a competitive disadvantage compared to harbours in countries that do ‘not impose stricter regulation than the IMO regulation This, in turn, reinforces competition between countries when it comes to devising attractive rules for shipping companies Such competition may well water down safety requirements
‘To summarise, some of the important ethical issues in the case of the Herald of Free Enterprise are the following: the ship's design was inherently unstable once water entered the car deck Is it ethically justifiable to design, produce and use ships that are in certain circumstances inherently unstable? What is the responsibility of engineers in this complex situation? There were no warning lights on the bridge, therefore it was not possible to establish from the bridge whether the bow doors were closed or not, Should engineers attempt to anticipate Jhuman errors during the design process? Is it the responsibilty ofthe engineers to design in a way that prevents human errors as much as possible or even to design idiot proof ships? It is, for example, desirable to design ferries that cannot leave port unless the bow doors are fully closed and secured As we have seen in setting the design requirements, trade-offs are made between safety and economics There is economic pressure to water down safety requirements Are tradeoffe between economics and safety acceptable? Which of the choices regarding this trade-off can be justified? Does following the regulations lead to morally acceptable designs? Ethical issues that come up in design processes like the ones mentioned above will be central inthis thesis,
Trang 14Introduction 41.1 Research question and objective
‘Technology has a profound influence on society New possibilities and new risks arise as a consequence of the employment of new technologies and products, Decisions made during design processes shape the possibilities and risks of products These decisions are ethically relevant Some decisions, for example, can hhave a large influence on the safety of people using the product, Although there is extensive literature on design processes and on engineering ethics, specific attention for ethical issues in design processes is relatively new, A lot of the literature on engineering ethics has been developed from the study of disasters
such as the Herald of Free Enterprise or cases of whistle-blowing but in these studies only litle attention is given to the design process In this research T will focus on daily practice in engineering design It might seem strange to start this chapter with the description of a disaster given that I will look at daily ‘engineering practice, the description of the Herald disaster is only intended to ntoke it clear that engineers make choices regarding ethical issues during design processes These decisions can, but need not, have detrimental consequences, With or without the actual occurrence of the Herald disaster, the design of this
ro/ro ferry may be seen as an example of daily engineering practice My research question is as follows:
‘What kind of ethical issues come up and how do engineers deal with these ethical issues during design processes?
‘This analysie of design practice will contribute to engineering ethics It will provide detailed information on which ethical issues play a part in engineering design and how engineers decide about these issues This information should enrich discussions on the moral responsibility of engineers in design processes, “The objective ofthis research can be summarised as follows:
To provide a contribution to discussions on the moral respon: ‘engineers in engineering design processes
‘The contribution will consist of detailed descriptions of engineering design practices and a normative analysis of these design practices As can be seen in the Herald case there can be regulation pertaining to the design of the product ‘This research should provide information for answering, amongst others, the
following question: do engineers behave in a morally responsible manner if they follow the existing regulations or should responsible engineers do more than just follow the regulations?
ilties of
Trang 15Ethical issues in engineering design 4.1.1 Ethical issues
Until now I have assumed that the reader knows intuitively what ethical issues are, Iwill now explain in more detail the meaning of the term “ethical” as used here} I will call a problem an ethical or moral problem if moral values are at stake, In characterising moral values I will follow Thomas Nagel According to Nagel, there are different sources of value, special allegiances, general rights, utility, perfectionist ends of selfdevelopment and individual projects, that cannot bbe reduced to each other or to more fundamental values Values based on special allegiances are, according to Nagel a result of a subject's relationships to others and consist of special obligations to other people or institutions General rights are rights that everyone has as a human being, These rights constrain action: actions that violate these rights are morally not permitted According to Nagel, “ulility includes all aspects of benefit and harm to all people (or sentient beings) Nagel, 1979, 129] Perfectionist ends of selfdevelopment refer to the intrinsic value of certain achieventents Nagel provides examples of the intrinsic value of scientific discovery or artistic creation The fifth type of value derives from individual projects, Nagel says that ‘this is value in addition to whatever reasons: ‘may have led to them in the first place’ [Nagel, 1979, 130] An example Nagel gives is that ifyou have set out and started to climb to the top of Mount Everest then this project gains importance Ethical theories usually focus on one of the sources of value Kantianism focuses on universal rights, Utilitarianism only accounts for utility Virtue ethics concentrates on perfectionist end of sell development I do not want to limit myself to one source of value by choosing a definition of ethical issues that refers only to utility or virtues or universal rights In this thesis, issues that are related to one of the sources of moral values identified by Nagel are called ethical issues and decisions concerning ethical
issues are called “ethically relevant" decisions For example, issues concerning safety ate related to utility but also to universal rights, therefore safety is an ethical issue The term: “ethical issue" only indicates that the way engineers deal with an issue can be evaluated ftom an ethical point of view
Trang 16Introduction iseuee will not be considered ethical ietes in this thesis An example of thie is that some industrial designers conflate aesthetic and moral values
Some of the ethical issues are also legal issues, for example safety iseues ‘There is 2 lot of legislation, standards and codes pertaining to safety and design ‘This makes decisions regarding safety no less ethically elevant, it only provides ‘engineers with rules they should follow from a legal point of view when making decisions In these cases the way engineers deal with these issues can be ‘evaluated both from an ethical and a legal point of view Decisions about the safety of a product might then be morally right or wrong and legal or illegal A question that can be raised in such instances is whether a design that is sale enough according to legislation is also ethically acceptable and vice versa
Legislation, codes and standards regarding safety can also be evaluated ethically 41.1.2 Ethical issues in engineering design
‘To take into account all ethical issues connected in one way or another toa design process would be impossible It is not that difficult to point out the ethical relevance of what seems to be a very trivial choice, like which tea to drink during, meetings of a design team Tea can be produced organically or with the use of herbicides and under good or bad working conditions The choice of what tea to drink is therefore related to utility and universal rights Lots of ethical issues might pay a role in the design contest, for example some pasts might be produced in countries where child labour is usual and therefore it might be assumed that these parts are made by children, Although all these issues like child labour, exploitation of underdeveloped countries, use of herbicides and pesticides are indeed ethical issues, these issues will not form the main focus of interest in this research I will concentrate on ethical estes that have a direct n of a product and the way itis used In particular, I will focus on ethical issues concerning safety and sustainability The reason for the focus on safety and sustainability is that these play 2 dominant role in many design processes Given the conception of ethical issues it is clear that safety and sustainability may give rise to ethical issues Decisions made about these issues are related to utility and general rights, Decisions regarding safety and sustainability are made in almost every design process, although the importance of these subjects may differ In some cases, sustainability or safety will not be regarded or discussed by the engineers, but this does not mean that there are no choices made regarding sustainability or safety
In the following two examples I will show that the impact of decisions made influence on the desi
Trang 17Ethical issues in engineering design
by an individual user about safety and sustainability are usually of 2 much smaller magnitude than those for decisions made during a design process
‘When designing a printer/copier, a choice needs to be made as to whether the printer/eopier will be able to print two sided or not Once a choice is made for two sided printing and eopying, an additional choice needs to be made about the default properties Iftwo sided printing isthe default option, users have to make an explicit choice to print one sided, Usually the prints and copies coming of the ‘machine will be printed two sided Only in exceptional eases, where the two sided copies and prints option is switched off by the user, will papers be printed one sided This default option will probably save a lot of paper compared with a printer/copier that can only print one side The environmental effects of saving paper are not that big ifa single printer/copier is regarded but when the total ‘umber of printers/copiers in use is considered the amount of paper saved by printing two sided copies and prints is enormous As paper is produced from ‘wood, 2 reduction in paper use will also reduce the amount of wood used The production of paper, the transportation of wood and the transportation of paper all require energy The amount of energy used will also be reduced and the total reduction in the resources used will be significant on a global scale This example shows that decisions made during a design phase of a product, and that seem ‘rival during that phase, can have large environmental effects
‘Another example ofthe ethical impact of design decisions is the following A person may decide not to drive too fast as this is usually dangerous and not covironment friendly The government of a country might decide to regulate the speed of cars by imposing speed limits, If there are speed limits imposed drivers, can still drive as fast as they wish, and is possible in their ear, but they will run the risk of being fined when exceeding the speed limits, Car engineers might decide to design a ear in which iis impossible to exceed the speed limits Trucks for example in the Netherlands are equipped witha speed regulator that makes it impossible for the driver to drive faster than 90 kmh This example illustrates the influence engineers may have; they can promote or prevent speeding Independently of what regulation requires or what speed limits are legally enforced, engineers can design cars with lower top speeds.* Cars with top speeds of 300 km/h make speeding possible and might perhaps invite drivers to test the top speed while installing a speed regulator or designing a car with 2 less powerful engine would make such speeding impossible Designing cars with
“nthe Netherlands only tucks are legally required tobe equipped with speed regulators Ensinees mish however, decide to equip cas with speed regulators even hss na eally required, Theresa gentlemen's agreement between German car producers omit the speed of ‘carto2504km/h, examples are the Mercedes Benz CLK 55 ANG cabriolet, the BMW 1S (ith ‘peed limiter 250 mn without 338 km/n), Audi Ad sporback 3.2 Quatro Although MG isnot = German eat producer, the MS 21/27-T260 i ls limited to 250 kh [Caro 2008,
Trang 18
Introduction lower top speeds would also save a lot of fuel as the fuel consumption is higher at hhigher speeds, Lower fuel consumption also decreases CO, production Smialler speed differences, for example between trucks and cars may possibly decrease the number of accidents occurring on roads and thereby the number of people injured and killed on roads So by choosing to design a car with lower top speeds, be it by actively limiting the top speed of the car or designing a less powerful ‘engine, engineers can reduce fuel consumption, CO, production and the amount
and severity of accidents on highways 1.2 Research approach
Descriptions of design practices have to be obtained to answer the research question, This will be done through case-studies [Yin, 1984/2989] In case
studies, different ways of obtaining data can be used In my ease-studies, I have interviewed engineers, observed design teams at work and read official and informal design documents,
‘Observing of design meetings allowed me to collect information about the ‘way the decisions are made by engineers Observing design meetings is also a ‘way to get information about what engineers perceived to be the difficulties and challenges of design processes The meetings used for the case-studies were taped and the tapes transeribed See appendix x for a list of the meetings that ‘were observed
Design documents, especially the official ones meant for customers, give a kkind of reconstruction of the decisions made in design meetings, The design documents can sometimes provide additional information, for example, in some design meetings arguments for or against certain choices were given but the actual decision was not taken during the meeting although the decision was documented in the design documents The more informal design documents ‘often gave information on specific aspects of the design process This information can be used for later official design documents for customers Sometimes the informal documents were more detailed than the official design documents,
Interviews were used to get further information on what role specific cengincers had in the design process and whether they encountered ethical issues In the interviews engineers were asked what they considered to be the cthical iseues in their design process Most interviews were held at or near the ‘end of the observation period, soit was possible to ask engineers about anything that war not yet clear to me after having read the design documents and observed the design meetings All the interviews were transcribed and the
transcriptions were approved by the interviewees, see appendix
Trang 19Ethical issues in engineering design
After the interview and observation periods I gave a presentation of my results to each of the design teams This presentation was followed by 2 discussion, Engineers could indicate whether I had made some incorzeet factual statements about the design process and whether or not they recognised the results, The presentation was also a last opportunity for me to ask about some details that were not yet clear to me The reason for giving a presentation at the end of the observation and interviewing period was that the engineers were curious about, and interested in, my research and resulls [engineers asked me about the results, I told them that I would present my results, and give them the opportunity to react to these results, later on This gave me the opportunity to postpone discussions on safety and sustainability until the presentation in order to influence the design process as little ax possible
T have chosen to change the names of the participating engineers in the descriptions of the design processes The engineers did not ask me to do this but Thhave chosen to protect their privacy in the niain text, see appendis 1 for more concrete information, The exact identity of the persons involved does not matter for the case descriptions, his or her arguments, decisions and formal position in the design team are relevant for this thesis
In doing the case-studies, I made the choice to present myself as an engineer among engineers A large advantage of this choice was that the design team, members knew that I could understand the “language” of engineers; and although my participation was kept to a minimum during the observation and interview period, the members of the design teams knew that I was qualified cengincer; this made communication easier The design team members did not feel compelled to explain everything they were doing in a simplified way
In line with keeping my participation at a minimum, I did not contribute to the solution of the design problem, This choice was also made because participating in a design process would require a lot from me as I had no previous design experience, The design task would probably completely absorb zy attention and time, making it difficult also to observe what was going on in the team Therefore, I was involved but not as a memiber of the design team 1 hhad some input in the design process at the end of the data gathering period when I presented my results
Trang 20Introduction could not imagine humon rights issues in their company Second, others thought
that ethics was only concemed with what kind of life a person should live and wondered what designing hod to do with that Third, some people, Bachelor students especially, thought that I would study the etiquettes in their design team Fourth, some engineers shared the interpretation of ethical issues used in this study and expected me to look at decisions concerning the safety of the design or the prevention of disasters I deliberately did not correct the engineers ‘who thought that I was interested in etiquette the good life or human rights My presence might have influenced the design team but as most ofthe engineers did not know exactly what I was looking at it is not likely that they placed more ‘emphasis on decisions concerning safety and sustainability I taped whole design mieetings and made notes throughout design meetings and interviews, not only when safely or sustainability issues were under discussion If engineers asked {questions about my research results I usually referred to the presentation that Ï ‘would give later on
‘As litte is known about the way engineers deal with ethical issues in daily ‘engineering design, this was an exploratory research project Based on ideas taken from the litersture on design processes, and to be presented in chapter 2, ‘working hypotheses were formulated These working hypotheses and the selected cases are introduced in chapter 3 The cases are described in chapters 4 to.7, Conclusions are drawn from the cases and an effort is made to generalise the results of the case-studies in chapter 8 The results fiom this research are ‘used to make a start with defining conditions for warranted trust in designing ‘engineers, in chapter 9 These conditions lead to a preliminary delineation of the
ioral responsibilities that engineers have during a design process
Trang 222 Engineering ethics and design processes
“The research question and objective formulated in chapter 1 indicate that this thesis contributes to engineering ethics This research should lead to deseriptions of the ethical issues engineers encounter and how they deal with these issues in design processes, The focus on ethical issues in engineering design processes is relatively new As yet in engineering ethics there has not been a lot of systematic attention for design processes, as will be indicated in section 2.1 An overview of the literature on the nature of design processes is presented in section 2.2, This overview is relevant because the ideas about the nature of design processes are used to guide the gathering of data in the case- studies Ideas about design processes that are particularly relevant for this thesis, because these ideas explain and inform the formulations of working hypotheses presented in the nest chapter are introduced in section 2.3
2.1 Engineering ethics
Research into ethics and design is part ofthe research field of engineering ethics In this section I will not give a complete overview of engineering ethics literature, Twill restrict myself to a description of the main issues that are focussed on in ‘engineering ethice and how this research ie positioned with regards to these Engineering ethics is the field of study that focuses on the ethical aspects of the actions and decisions of engineers, both individually and collectively A rather broad range of (ethical) issues are discussed in engineering ethics: profecsional codes of conduct, whistle blowing, dealing with safety and risks, liability issues, conflicts of interests, multinational corporations, privacy etc (eee for example [Harris et al,, 1995] {Davis, 1998] and (Bird, r998)) A substantial amount of literature on the teaching of engincering ethics to engincering students has been developed since the beginning of the 1980's, (cf Baum, 1980] [Unger, 1982]
[Martin and Schinzinger, 1989] [Harris etal, 1995], [Birsch and Fielder, 1994) ‘A salient feature of engineering ethies literature is that a lot of it has been developed based on studies of disasters like the Challenger disaster ([Vaughan, 1996] and [Davis, 1998) Another feature of engineering ethics is that, especially in the United States, there are a lot of proponents who regard engineering ethics as a kind of professional ethies (cf [Schaub et al, x983} [Davis, 2001] and [Harrs, 2004) The idea is that the engineer as a professional has obligations not only to bis or her employer but also to the general public, as for example doctors or
Trang 23Ethical issues in engineering design
lawyers also have obligations Engineers should adhere to professional codes of conduct that state, for example that engineers shall hold the safety and welfare of the public paramount Based on descriptions of the Challenger disaster, Davis emphasizes that there is a difference between engineers and managers Engineers should adhere to their professional norms and hold safety paramount and managers do not do this [Davis, 1998), This tendency to regard engineering ethics ae a kind of professional ethies has led to s focus on the individual cengineer and his or her responsibilities in his or her job and profession in most (American) engineering ethics textbooks This can also explain the focus on whistle blowing that can be found in some of the engineering ethics literature
The individual engineer should in certain cases take his or her moral and professional responsibilities seriously and blow the whistle,
According to Zandvoort et al [Zandvoort et al, 2000] and Devon et al Devon, etal, 2001] engineering ethics should focus on more than the individual engineer They argue that the ethical problems that engineers encounter are partly due to the contest they work in Some of the ethical problems cannot be solved by individual enginects or the profession
Trang 24Engineering ethics and design processes persons who clearly cooperate in the design process just because they have another job description or another educational background,
2.1.1 Ethies in design processes
Engineering design is an interesting topic for research from the point of view of engineering ethics because design is one of the core activites of engineers
Moreover, technology has social and ethical implications because of the kinds of products produced, as the outcomes of design processes.” Only recently has more attention been given to ethics and engineering design (Lloyd and Busby, 2003], [Devon et al, 2001) [Van de Poel, 2001] Interesting developments in ethies and design can be found in the field of software design and computer ethics Efforts to incorporate values into the design of software have been labelled "value sensitive design’
Lloyd and Busby use empirical data to deseribe how engineers deal with ethical issues in design [Lloyd and Busby, 2003], They use three main ethical ‘theories to see whether reasoning and argumentation during the design process it within these theories They refer to the thtee ethical theories as “consequentialism", "deontology” and “virtue ethics” They looked at all reasoning, not just at reasoning about issues that are clearly ethical like safety [Lloyd and Busby, 2003, 514) They relate, for example, reasoning about making a better product to consequentialist reasoning, They conclude that, contrary to their ‘expectations, consequentialist reasoning is not prevalent in engineering design
Engineers also use deontological reasoning and engineers identify what Lloyd and Busby call virtues of engineers like collectivity, consistency and emphasising evidence, Lloyd and Busby have considered nornial day-to-day situations in which design decisions ate made According to Lloyd and Busby s great number of small design decisions that each seem to be ethically neutral, can add up to cthically relevant consequences:
“Although it is simply a fact that not much of engineering designing is specifically about what one might normally consider to be ethical issues, the products of engineering design and particularly the use of those products- undoubtedly ate [Lloyd and Busby, 2003, 4]
In many design processes, ethical problems are indeed difficult to recagnise and less specific than some of the examples given in the literature on disasters 1 agree with Lloyd and Busby that in every design process “smaller’ ethically relevant decisions are made I think however that itis problematic to regard all, decisions as being possibly ethically relevant Some values like, for example,
efficiency are not moral values (see section 1.11) Efficiency is therefore not * Ofeauree leo the extent and the way in hich product ate used is important
Trang 25Ethical issues in engineering design
necessarily an ethical issue Decisions regarding efficiency can, however, be cthically relevant if they ate related to, for example, sustainability Making a more energy efficient product is ethically relevant because it is a part of designing a more sustainable product The same holds for trying to design a product as simply as possible Simplicity is a normative term but it need not be a moral term Decisions concerning simplicity are only sometimes related to moral values Simplicity might be an ethical issue if it is related to ease of operation A simple product ean probably prevent accidents related to unintended misuse, If operation of a machine requires a complex procedure, there is a chance that operators will make a mistake when carrying out the procedure Another issue related to simplicity might be that a simple product can be used by everyone, unlike some video recorders or microwave avens that people find too difficult to use, Simplicity can therefore sometimes be related to moral values but this need not be the care
In contrast to Lloyd and Busby, who studied the (ethical) reasoning that engineers use in design processes, Van de Poel distinguished five actions during the design process that may be ethically relevant
1) The formulation of goals, design criteria and requitements and their operationalistion,
2} The choice of alternatives to be investigated during 2 design process ad the selection among those alternatives ata ater stage inthe process 3) The assessment of trade-offs between design criteria and decisions about the acceptability of particular trade-offs,
4) The assessment of risks and secondary effects and decisions about the acceptability of these
5) The assessment of scripts and political and social visions that are (implicitly) inherent in a design and decisions about the desirability of| these scripts [Van de Poel, 2000, 3}
Van de Poels approsch would imply that, for example, the formulation of requirements is an action that can be expected to be done during design processes, Formulating requirements can be ethically relevant, for example, if safety requirements are formulated ‘These requirements need to be eperationalised and this operationalisation is also ethically relevant Different altematives that score differently with respect to different requirements and different operationalisations of requirements may have to be assessed, Tradeoffs between different requirements may have to be made, In accordance with Van de Poel's approach these actions can all be ethically relevant if related to moral values, and therefore these are included in this research
“The concept of value sensitive design has been developed within computer ethics and human computer interface design According to Friedman and others
Trang 26Engineering ethics and design processes “Value Sensitive Design i a theoretically grounded approach tothe design
of technology that accounts for human values ina principled and comprehensive manner throughout the design process [Friedman eta 2005,
‘This definition does not imply that value sensitive design is only applicable Within software and computer design Yet, the concept has until now been mainly used within these fields and not with regard to the design of other kinds of technology [www.nyu.edu/projects/valuesindesign/indexhtml In using a philosophical analysis of values and sociological research into the use and development of technology, value sensitive design is an attempt to make software <designs that account for moral values like privacy and autonomy
‘The research under taken here may be considered research into the way ‘engineers deal with moral values during engineering design processes and therefore research into value sensitive design There is, however, a difference: ‘whereas researchers into values sensitive design are trying to develop a method for dealing with moral values, I will concentrate on describing how engineers deal with ethical issues like safety and sustainability Another difference is that some ethical issues that are very important in software design like, for example, privacy and identity are not that important for my case-studies of engineering design
2.2 Design
“The features of design problems and design processes relevant to the topic of this thesis are presented in this section At the end of this section | will present
the conception of design processes used in this work 2.2.1 Design process
According to Cross, the design process can be seen as a process in which products or tools ate created to suit human purposes (Cross, 2000, 3|, The starting point of a design process is usually some stated or perceived customer's needs, A material structure that meets these functional requirements is designed The design process is usually constrained by economic and time restrictions, A design should be finished by a certain date and the costs of the whole design process should not exceed a certain amount of money
In the literature on design methodology a lot of different models of design processes can be found ([Cross, 1989], [Roozenburg and Cross, 1991] and [Baxter,
Trang 27Ethical issues in engineering design
1996): Crots presents a model of the design proceas that consists of three phases: generation, evaluation and communication, A concept is generated in the fist phase of the design process A designer needs to understand the design problem and to find possible solutions for it; this usually happens simultaneously Possible solutions help the designer to get a better tnderttanding of the design problem The concept ia evalutind in the second phase During the evaluation, a decision is made as to whether the possible solution meets the requirements The concept is adapted in sn iterative proce Often, more than one iterative step is necessary because adaptation ofa part of the design can lead to problems in other parts of the design The design is communicated to the people who are responsible for production in the third ‘have Drawings, enmputer drawings al descrotions ofthe design ane used in
‘this communication [Cross, 1989],
Another more detailed model is proposed by French [cited in Cros, 1989, 21 22), French divides the design process into four activites: + analysis ofthe problem
‘conceptual design + embodiment of schemes + detailing
‘An analysis of the design problem should lead to clear statement of the problem, The requirements and constraints are formulated in this phase The designer searches for different possible solutions and makes schemes of them in the conceptual design phase, In the next phase, embodiment of schemes, z choice is made between the schemes The scheme is further detailed in the detailing phase
Although there are different models that can be used to divide the design procen intn diffrent phates and tute diferent terme to nar the phases, there
are similarities between the models, (see also [Roozenburg and Cross, 991] and
{VDI azar: Systemic Appronch tothe Design of Technica! Systems and Products ited in Cross, 2000, 39) The design process can groso modo be described a follows The goal, requirements and constraints are defined atthe beginning ofthe design process, This i sometimes done by the customers or in co-operation between customer and engineers After this creaise part fllows in which concepts are generated and evaluated In the next phase, one concept is chosen and that concept is further detailed Finally, drawings and descriptions ofthe design are rade for the production of the product, The design process i not 2 linear process; iti trative t may always be necessary to go back one or moe steps and then move forward gain
Trang 28Engineering ethics and design processes 2.2.2 Design problems
If design problems are problems in which the requirements alone determine the solution then engineers can say that they are not responsible for ethical issues because the requirements determine everything and the customers define the requirentents, Some authors maintain that engineers are not, and should not be, involved in the formulation of design requirements, criteria or goals [Florman, 1983], According to Florman, the formulation of requirements and goals is cthically relevant, but this should not be done by engineers Managers
politicians, customers ete should formulate the requirements In this line of thinking, the task of engineers is to discover what is technologically the best solution given certain requirements, This task is seen as ethically neutral Ethical {questions may arise in the user phate when technologies ate used for certain purposes and produce certain (socal) effects According to Florman these ethical {questions concerning use are also oulside the scope of the engineers and should be solved by the user (see figure 2.1) In this model, the sole responsibility of ‘engineers is to carry out a task formulated by others in a competent way
boa, reaurerert| an constamns Engineering exon (oni soca ofc) Use of product Poticane pringpals
manages, Designing Users
(anticipated) ene nil
Figure a: Division of labour with respect to engincering design if design problems were well-structured problems in which the requirements fully Aetermine the solution, afer [Van Gorp and Van de Poel, 2001 Design problems are, however, usually not problems where a clear set of requirements is available that completely determines the solution, Design problems are more or less ill-structured problems ([Simon, 1973] and [Cross, 1989) Simon states that in the ill-structured problem of designing a house:
“There i initially no definite criterion to test a proposed solution, much less 2 mochanizeable process to apply the criterion The problem space is not defined in any meaningful way, [Simon, 1975, 31]
Trang 29Ethical issues in engineering design
For Simon the main characteristics of an ill-structused problem are that the solution space is not well-defined and that there is no criterion to test different solutions and decide which is best Cross gives the following characteristics of ill- structured problems:
“1 There is no definite formulation ofthe problem, _ Any problem formulation may embody inconsistencies, 3, Formulations of the problem are salution-dependent
‘4 Proposing solutions isa means to understanding the problem, 5 There is no definitive solution t the problem [Cross, 1989, 143)
Some design methods require that engineers formulate the requirements and solutions separately and independently, but this is impossible if design problems are ill-structured, In a redesign of an existing design it might be possible to forniulate most of the requirements atthe start ofthe design procese but this is not a definition of the requirements independent of the solution, The solution space is, in these cases, limited because 9 redesign is made; certain features of the product will remain the same, Other design problems aiming at designing a completely new product are very illstructured and only some vague requirements can be formulated at the start of the design process So design problems can be more or less illstructured
‘An example ofan illstructured problem isthe following In the mid nineteen nineties substitutes were sought for replacing CFCs as coolants in refrigerators, because CFCs damage the ozone layer [Van de Poel, 1998 and 2001], Two altematives were considered: HEC 1343 and hydrocarbons, both have their advantages and their disadvantages Hydrocarbons are for example flammable and existing reftigerator design needed to be changed if hydrocarbons were used, HFCi54a has a long atmospheric lifetime and if released would therefore still damage the environment, although to a lesser extent than CFCs, There were different operationalisations available for the environmental, health and safety criteria Both proposed solutions scored differently under different operationalisations of the criteria There was no solution that was best under all, ‘operationalisations No definite criterion was available to say which solution was the better one This example shows that even for the seemingly simple case of looking for a substitute coolant in existing refrigerator design, there are features of the problem that make it il-structured
In cases where a design problem is an illstructured problem, there may be ‘more than one solution: each of these solutions can be valid Engineers, in this cate, have to make a choice: it is not the case that the requirements will lead to just one solution, At the start of design process, there may not even be a clear and unambiguous set of requirements During the design process it may be proved that there is no solution to the illstructured problem, In some cases it
Trang 30Engineering ethics and design processes right prove to be necessary to adjust or drop some requirements because n0 solution meeting all the requirements can be found So a design problem can be under, or over, determined by the requirements, Either way, engineers need to mmiake choices during the design process for example regarding which xequiremente can be dropped or which of the possible solutions to the design problem is the best
2.2.3 The design process as a social process
Most designs are made by ø team of engineers Designing is in these cases a social process Choices are miade in, and by groups of people During the design, process, communication, negotiation, argumentation, {mis)trust between ‘engineers and power differences between engineers influence the design This has consequences for design research as the design process should be ‘conceptualised a8 a social process There is some research into actual design processes with design teams [Bucciatelli, 1994], [Lloyd and Busby, 2001}, [Lloyd, 2000] and [Baird et al, 2000} Bucciatelli describes the design process asa social process in which negotiation is necessary:
‘Contemporary design is, in most instances, 2 complex affair in which participants with diferent responsibilities and interests must bring their stories in coherence [Buciatell, 1994, 83]
‘The different engineers, with their different educational backgrounds and ‘experiences, will all conceive the design task differently Take for example the ‘age construction and bodywork ofa car A mechanical engineer looks at stresses and strains within the cage construction and bodywork of a car He or she tries to design them in such a way that stresses and strains remain low dusing normal se and absorb energy during a crash, An aerodynamics engineer might look at the same bodywork and sees 4 body that needs to have a low frontal area and a low drag coefficient Although both the mechanical and the aerodynamics engineer look at the same parts they see something different and think of <ifferent requirements the parts should meet All these different views have to be
‘brought in coherence’ (Bucciareli, 1994), just like all the parts have to fit and fanction together This “bringing into coherence’ is done as a process of ‘communication and negotiation,
Other authors also recognise the importance of social processes during the design process and stress the importance of communication, Lloyd stresses the importance of storytelling within the design process [Lloyd, 2000] Engineers construct stories during the design process These stories are used to come to a ‘conimon understanding:
Trang 31Ethical issues in engineering design
Engineering design as a social activity consists in the construction of secial agreements We have observed storytelling to be a mechanism that ids this construction’ Lloyd, 2000, 376]
‘The stories can be used within the entire company or within single departments ‘The whole company or design team might know the story of a previous design failure Stories of specific difficult customers might only be known in the sales department Knowing the stories is part of being part of the design team or department, Stories can therefore be inclusive or exclusive in their use (Lloyd, 2000}
Baird and others conducted an ethnographic study at Rolls-Royce Aerospace [Baird et al, 2000) They conclude that personal interaction between engineers is crucial for information to be disseminated throughout an orgenisation Engineers ‘who know each other from other projects tend to ask each other for advice when working on new projects At the beginning of the design process more experienced senior engineers are very important, They help younger engineers and point them ‘to the sequences and sources of expert opinion they should seek’ [Baird et al, 2000, 350) According to Baird etal, this helps with structuring the design problem
2.2.4 Organisation of the design process
A division of labour exists in most design processes For example, the design of car can be divided into the design of the drive shaft, the engine, the seats, the clectronic systems, the suspension and the styling of the car ete The partitioning of a design team into smaller design groups responsible for a part of the design is, fiom an ethical point of view, noteworthy because it may lead to the problem of many hands [Bovens, 1998] and [Thompson, 980] This problem presents itself with regard to active and passive responsibility Passive responsiblity is responsibility after something has happened: being held accountable Active responsibility refers to being or feeling responsible for something or some task [Bovens, 1998}
With regard to passive responsibility, the problem of many hands is the following, It might seem to be quite clear who is officially responsible for what as this depends on formal job descriptions in organisations, but in practice it is very difficult to point out the people responsible for acts of organisations that have caused damage, Organisations are often opaque to people outside of the organisation It is not clear who is responsible for what and who was able to influence a certain decision, Above this, when an organisation is organised hnjerarchically, people lower in the hierarchy can indicate that those higher in the hierarchy are responsible while those higher in the hierarchy claim to have no knowledge of the situation,
Trang 32Engineering ethics and design processes With regard to active responsibility, the problem of many hands can be seen ‘when no one feels or thinks that he or she is responsible for certain issues If issues are not specifically part of comeone’s task description, everyone can avoid taking responsibility for them These issues may then be neglected in the design process
In a paper on the relationship between how companies are organised and harm they cause other people, Darley studied a case of the design and testing of landing gear for a military aircraft that failed during landing after test flights [Darley, 1996] People in the company knew that there were calculation errors and because of these errors there was a large chance the landing gear would fal Certain social mechanisms made the people in the organisation actively conceal
the calculation errors for their customers and tinker with data in certifying documents In this case people felt either forced by their superiors or felt they ‘were already too involved and caught up with the tinkering of the data to stop the concealing of caleulation errors
Darley also points to the way a decision is framed Stopping 2 production process or changing a design requires action, while going on is often seen as not requiring action Action has to be defended to other people while in the case of doing or changing nothing, such defence is not requited Action is only taken when there is proof of harm This frames the decision in an “innocent until proven guilty" way This is different from using a precautionary frame for the decision, where no harm has to be proven, A suspicion of harm can be enough to ‘warrant acting against it Framing a similar decision differently can lead to different actions [Darley, 2996] This might be relevant for design processes, ‘especially when decisions have to be made to stop or go on with a design process, “The aspects of design processes mentioned above are all relevant from an ethical
point of view and therefore they were included in the conception of design processes used in this thesis In view of the foregoing, I regard design processes as organised social processes which aim at solving more or less illstructured design problems in this work, All of these aspects mentioned above were used to support data-collection in the case-studies For example, in the case-studies attention was paid to the organisation of the design team and the social processes within the design team because this helped me to determine who was involved in what decisions when dealing with ethical issues
Trang 33Ethical issues in engineering design
2.3 Characteristics of design processes in relation to ethical issues*
So far, some very general characteristics of design processes have been discussed When studying ethical issues in engineering design it may be useful to make further distinctions between different kinds of design processes, It might be expected for instance that during the design process for a bolt for a car wheel different ethical decisions have to be made than those that have to be made during the design process for a completely new personal transportation device Ideas drawn from the literature are used to characterise the different kinds of design processes Working hypothetes are formulated in the next chapter based fon these ideas I conclude this chapter with a discussion of some preliminary ideas regarding what this research, and its results, might contribute to discussions about the moral responsibilty of engineers during design processes ‘These ideas will be further elaborated in the last chapter of this thesis
2.3.1 Design type and design hierarchy
Let us go back to the example of a design for a wheel bolt om a car as opposed ta the design of a new personal transportation device The bolt design has to comply with dimensional constraints, safely norms, standards, financial constraints ete It is a small part of a known product, Moreover, most designs of bolts are redesigns of existing bolts Norms, standards or dimensional constraints are absent for a new personal transportation device, or it is questionable whether existing norms, standards or dimensional constraints can or should be used The design problem for a new personal transportation device is more ill-structured than that of designing a bolt, The reason why the design problem for the bolt is better structured than that for the new transportation device is that there are ‘more extemal constraints pertaining to the design of the bolt? 1 use the term “external constraints" here for all constraints that ate taken for granted during 2 design process Some of these constraints may be set by the engineers at the start of the design process, for example by the engineers deciding to redesign an existing bolt instead of designing a new one, Other constraints are set by other stakeholders and not the engineers, such se the customer's requirements, governmental regulations or codes and standards These external constraints are ‘usually already operationalised into specific and clear requirements Ina redesign, it is also usually obvious how these requirements can be implemented
“the ideas in setion 2.1 and ection 2.2.2 are based onthe paper‘The need for thea reflection in enginceting design; the elevnce of typeof design and design erarchy’ Van de Poel and Van Gop, 2006)
Trang 34Engineering ethics and design processes [As the example of the bolt and the new transportation device suggests, ‘engineers are confronted with different degrees of external constraint in different design processes According to Vincenti Vincenti, 1992], this degree of external constraint depends mainly on two dimensions: the level of design hierarchy and
the type of design (normal versus radical)" Design hierarchy
Most modern products consist of several parts, subassemblies and subsystems” In many cases these subsystems and parts are more or less independently designed Depending on how the design process is organised, different teams
and engineers work on different parts of the product There is communication and co-operation between the teams or at least there usually is These design teams can be from the same or from different companies The parts, subassemblies and subsystems are ordered hierarchically The complete product is designed at the highest levels of the design hierarchy: subsystems and parts are designed at lower levels Vincenti divides the design hierarchy of the design
process of an airplane in the following levels
1, Project definition: translation of some usually ildefined military or commercial requirement into a concrete technical problem for level 2 2, Overall design: layout of arrangement and proportions ofthe airplane to ‘meet the project definition,
3 Majorcamponent design: division of project into wing design, Fuselage design, landing-gear design, clectricalsystem design et
4 Subdivision of areas of component design from level 3 according to engincering discipline required (eg, aerodynamic wing design, structural wing design, mechanical wing design)
5, Further division of categories in level into highly specific problems o, scrodynamic wing design into problenis of platform, arfll section and high-ift devices) [Vincent 1996, 9}
‘There are similarities between Vincenti's ideas of design hierarchy and the design hierarchy levels defined by Disco et al [Diseo et al, 1992} Disco et al distinguish the following levels of hierarchy:
+ systems, like a plant, electricity or cable networks + functional artefacts, like cars, ete
+ devices like pumps, motors ete
+ components, ike materials, nuts, condensers ete
slsleadlon tor in ny vcw alnce some the constralts menlowed by Vineet are not {ehnieal but ather social in nature Vincent seems to use the tr techies constant forall onsrants thatthe engineer sofa design team canna change Vincent 19921 "Watton, forexample, estimates that there are abou 0.000 pars inthe For Taurus Waton, 1997]
Trang 35Ethical issues in engineering design
Vincenti’ divisions of the design hierarchy are more fine grained at the lower levels Component design is not the lowest level but middle level Disco etal add the systems level atthe highest level
‘According to Vincenti, the degree of external constraint is larger if the design process is lower in the design hierarchy The higher levels of the design process pose external constraints for the lower levels Examples of these constraints at, amongst others, dimensional constraints, a part needs to fit in the whole product, bbut also constraints concerning the function of the part
‘The ideas of Vincenti and Disco et al, about design hierarchy seem to resemble the phases in the design process presented in section 2.2.1, but there is a difference The difference between design hierarchy and the phase ofthe design process is that a design process is done for every product, subassembly or part
This means that at every level of the hierarchy all phases of the design process are gone through One could think for example of the generation of concepts design phase of a part of a product This would be the generation of concepts design phase of a middle level design The relative importance of design phases can differ in the design hierarchy, for example in high level design the phase of the design process in which detailed drawings are made can be relatively ‘unimportant compared with the generation of concepts
Type of design
Besides the notion of design hierarchy Vincenti also introduced the notion of design type ranging from normal to radical design Vincenti uses the terms
“operational principle” and “normal configuration" to indicate what normal design as opposed to radical design is [Vincenti, 1990) “Operational principle” is + tem introduced by Polanyi [Polanyi, 1962], It refers to how a device works, The ‘normal configuration is described by Vincenti as
the general shape and arrangement that are commonly agreed to best embody the operational principle Vincenti, 199, 209]
Examples of different working principles of car engines are a combustion engine and a fuel cell electrical engine Both engines power cars but they have different working principles In a combustion engine fuel and air are let into a cylinder and ignited The expanding volume of the ignited gates is used to get a rotational ‘movement In fuel cell cazs an electrochemical reaction between hydzogen and oxygen produces electricity This electricity is used to drive the car So although Doth engines power the car they do so in a different way In normal design, both operational principle and normal configuration are kept the same as in previous designs In radical design, the operations principle and/or normal configuration, are unknown or it is decided that the conventional operational principle and normal configuration will not be used in the design
Trang 36Engineering ethics and design processes Vincenti’s description of radical design focuses on the structure and physical aspects of the design For my purpose, it is useful to introduce 2 somewhat broader definition of radical design In design, function and structure are joined in an artefact (ef, [Kroes, 2002] and fwww.dualnature.tudelft nll) This means that a design can also be radical with regard to its fumetion or design eriteria An ‘explicit choice can be made at the beginning of the design process to change the usual idea of a good product of this product type This means setting different criteria or changing the relative importance of criteria For example, speed is ‘often accorded some importance but it is usually not the most important criterion in the design ofa car The standard idea of a good cat is a safe, reliable and perhaps fast car If the aim of a design process is to design a car that can break the sound barter, this is a radical design process Radical designing in this functional way can make reconsideration of the operational principle and the normal configuration necessary Reconsidering may, but does not have to, lead to changes in the operational principle or normal configuration Thus a radical design process with regard to the function may lead to a radical design of the physical structure, but this is not necessarily so, It is also possible that a new ‘operational principle leads to new criteria,
Regulative framew ork
Vincenti claims that there are more external constraints in normal as opposed to radical design [Vincenti, 1992] Normal design is a form of standard design practice guided by existing formal and informal rules A system of regulations and formal niles concerning a product exists in normal design I will efer to this system as 2 regulative framework A regulative framework for a certain product consists of al relevant regulation, national and international legislation, technical codes and standards and rules for controlling and certifying products, A regulative framework is socially sanctioned, for example by a national or the European parliament or by organisations thal approve technical codes Besides the technical codes and legislation, interpretation of legislation and technical codes are part of the regulative framework Interpretations of codes and legislation can be provided by the controlling and certifying organizations and also by engineering societies for example in the courses they organise for ‘engineers on state of the art design practices Engineering societies can also
Trang 37Ethical issues in engineering design
regulative framework Company specific norms and standards can therefore only impose stricter requirements than the regulative framework
“The notion ofa regulative framework differs from notions like technological regime of technological paradigm in that the regulative framework includes only formal norms, rules and their interpretations that pertain to the design, certification and construction of product Technological regimes are defined by Rip and Kemp as:
the ruleset or grammar embedded in a complex of engineering practices, production process technologies, product characteristics, kills and procedures, ways of handling relevant artefacts and persons, ways of defining problems: all of them imbedded in institutions and infrastructures” [Rip and Kemp, 1998, 358]
A technological regime defined in this way includes much more than the regulative framework as it also includes skills and ways of handling artefacts and persons, Other authors have used other definitions of technological regimes Van de Poel for example focuses on defining technological regimes for design processes, Although Van de Poel restricts his definition of technological regimes to technical regimes for design processes, he includes more in a technical regime for the design of a product than is included in a regulative framework, such as promises and expectations of a product [Van de Poel, 1998 and 20003),
2.3.2 Normative frameworks
‘The idea of normal design can be related to Grunwald’s idea that in “business-2= usual’ technology development there is no need for engineers to reflect ethically (Grunwald, 2001], Grunwald suggests that there are situations in which on engineer should ethically reflect on the development of technology and that there are situations which he classifies as business-ar-usual in which there is no such need for ethical reflection (Grunwald, 2000] and (Grunwald, aoorl) Grunwald indicates that in “business-ae-usual" a normative framework existe that governs all ethically relevant decisions that are made during a design process He thinks that a lot, if not most, decisions in technology development are covered by normative frameworks According to Grunwald engineers have to apply the rules, from the normative framework without further ethical rellection if this normative framework meets the following requirements:*
* Grunwald đocs roi cplclly la that engineers have an obligation ease the normative framework btthe equrement observed” seems to imply ths nan article fom 2005, Grunwald has changed the requirement of bserved to ‘omplance: the normative framework leo heo be complied within the fed concerned [Granwal, 2005, 383
Trang 38Engineering ethics and design processes “rapmatial complete: the normative framework has to comprehend adequately the decision to be made, and should leave out no essential ‘aspects fram consideration Lecally consistent there has tobe a “sulicent” degree of freedom from contradiction among the various elements of the normative Framework, Unambiguous: beyond the normative Framework, there has to be a sufficient common understanding among the actors in the context of the decision under discussion, Accepted: the normative framework has ta be accepted as the basis for the decision by those concerned,
Observed: the normative framework has to be in fact abserved; lip service, for instance, in environmental concerns, is not enough [Granvald, 2001, 419]
‘With regard to the acceptance of the normative framework Grunwald states that acceptance needs not be universal, but neither should it be restricted to the very narrow sphere of engineering, Instead it ‘must include further groups or individuals such as the assumed users, but also people possibly affected by the side-effects or other impacts’ (Grunwald, 2001, 419-426] Grunwald considers the normative framework to be a “morale provisoire" itis relative to ‘the actual state of the relation between culture, society and technology, relative to the moral convictions of society and to the knowledge about consequences and impact of
technology.’ [Grunwald, 2000, 191], The normative framework is therefore dynamic
According to Grunwald, normative framework consists of all obligations given by political regulation and all obligations resulting from other societal regulation like technical codes and standards, and codes of ethies [Grunwald, 2000], The normative framework that Grunwald has formulated comprises therefore the same clements as the regulative framework, In my analyses of the different design processes the question whether regulative frameworks exist and
fulfil the above requirements plays an important role
2.3.3 Moral responsibility and the trust relationship between engineers and society
Engineers have specific knowledge and experience and play an important part in the design of products, Engineers are given power to decide in design processes, ‘This power is limited by the regulative framework In this thesis, I will assume
Trang 39Ethical issues in engineering design
codes of ethics formulated by engineering societies all state that engineers should display integrity and honesty in their work, if they are to be trusted by customers and sociely The Institution of Engineers, Australia states the following in the second of their tenets in their code of ethics www ieaust a
‘Members shall act with honour, integrity and dignity in order to merit the trust ofthe community and the profession; [www eaustau, 2]
Its this trust relation and its ethical relevance that I will further analyse here, Much more can be said about trust than I can do in this section and the last chapter I will use a specific notion of trust based on ideas from Annette Baier and Bart Nooteboom as a basis for claims about the moral responsibility of engineers”
In her paper “Trust and Antitrust’, Annette Baier uses the examples of plumbers and surgeons to illustrate that we trust them to do what is necessary to fix what is wrong, We trust them and we do not prescribe what they should do exactly to fix, for example a leak [Baier, 1986, p 250] Baier claims that it is not possible to prescribe precisely what plumbers and surgeons must do because we do not have that knoviledge If we would be able to preseribe precisely step by step what the plumber has to do and what he should not do then we could probably repair the leak ourselves, My claim is that in a similar vein we trust cengincers to design safe products without prescribing precisely what they should do and refrain from doing
Baier claims that trust is a special sort of reliance: in trust we rely on the goodwill of someone else, Trust can be seen as ‘a three: place predicate (A trusts B with valued thing C)’ (Baier, 1986, 236] Discretionary power is given to the ‘rusted person, This means that the trusted person is allowed some discretion but not allowed to do everything he or she thinks is a way to take care of the valued thing There are limits as to what the entrusted person should be or is allowed to do Baier uses the example ofa babysitter:
‘babysitter who decides thatthe nursery would be improved if painted purple and sets to work to transform it, wil have acted, asa babysitter in an untrustworthy way, however good his good will [Bair, 1996, 236]
In most everyday situations the limite of this discretionary power are not negotiated or expressed explicitly, People are expected to know the limits of the discretionary power they get if they are trusted IF Task my neighbour to take care of my plants and mail while I am away on holiday, we both know that I do " Annette Baer has wten am ifuentil paper on trust within thes Bart Nooteboom tena book on trust that includes insights from economies, rational decision making, behav has wt
Trang 40Engineering ethics and design processes not expect or want her to read my mail, pay my bills or reply to my mail I trust hher to water my plants and place my mail on the table without reading or
answering it Therefore, trust gives power and responsibility but within limits ‘According to Baier trust can be morally decent or not Having trust in someone who is cleverly concealing his or her untrustworthiness might be morally wrong according to Baier, especially if that untrustworthy person is using his or her diseretionary power to gain more power over the person who trusts him or her In cates of sects, leaders ate often trusted and abuse this trust to gain power over their followers and harm thent This is of course quite 9 far fetched example but it shows that not all rust leads to the protection of what people value, and that trust can be morally wrong Trust in engineers might be ntieplaced and even morally wrong if engineers are trying to harm people with their work,
An interest har also been shown in trust between people and trust in institutions within economics and management theory [Nooteboom, 2002) ‘According to Nooteboon: trust is a fosr-place predicate:
‘Someone (1) trusts someone (or something) {2} in some respect (3) depending on conditions such as cantet of action (4) [Nooteboom, 2003, 3)
For example, if 1am ill then I trust my doctor to take cate of me and cure me If however, I have a terminal disease then the doctor has not behaved in an
untrustworthy manner because he has not cured me, A doctor miay behave in an tuntrustworthy manner in other respects but the fact that he cannot cure my terminal disease does not make him untrustworthy Nooteboom's notion of trust differs from the one Baier proposes in that the context of action ie made explicit in his notion You trust someone to do things to the degree to which he or she can influence the sittation and has the power to change certain situations for the good, Another difference is that Nooteboom includes the possibilty of trusting an organisation
‘OF course an onganization itself does not have an intention, but it has interests and can ty to regulate the intentions ofits workers to seve those Interests’ [Nooteboom, 2002.75)