James, where did your passion for this issue originate and how can the resources available for information literacy be put to use both by faculty and students?
We live in a time marked by an unprecedented deluge of information, where distinguishing reliable and valuable content has become increasingly difficult. My concern was to help engineering educators meet the critical challenge of fostering ethical behaviour in their students in this complex world. Students are in real need of an ethical compass to navigate this information overload, and the digital landscape in particular. They need to acquire what we call ‘information and digital literacy’, specifically, learning how to research, select and critically assess reliable data. This is both a skill and a practice.
For students, how does this skill relate to the engineering workplace?
From observing professional engineers, it’s clear they require comprehensive insights and data to resolve problems, complete projects, and foster innovation. This necessitates extensive research, encompassing case studies, standards, best practices, and examples to validate or refute their strategies. Engineering is a profession deeply rooted in the analysis of failures in order to prevent avoidable mistakes. As a result, critical and unbiased thinking is essential and all the more so in the current state of the information landscape. This is something Knovel specifically strives to improve for the communities we serve.
Knovel – a reference platform I’ve significantly contributed to – was initially built for practising engineers. Our early realisation was that the biggest obstacle for engineers in accessing the best available information wasn’t a lack of resources, but barriers such as insufficient digitalisation, technological hurdles, and ambiguous usage rights. Nowadays, the challenge has evolved: there’s an overload of online information, emerging yet unreliable sources like certain chatbots, and a persistently fragmented information landscape.
How is Knovel used in engineering education? Can you share some insights on how to make the most of it?
Knovel is distinguished by its extensive network of over 165 content partners worldwide, offering a breadth of trusted perspectives to meet the needs of a range of engineering information challenges. It’s an invaluable tool for students, especially those in project-based learning programs during their Undergraduate and Master’s studies. These students are on the cusp of facing real-world engineering challenges, and Knovel exposes them to the information practices of professional engineers.
The platform is adept at introducing students to the research methodologies and information sources that a practising engineer would utilise. It helps them understand how professionals in their field gather insights, evaluate information, and engage in the creative process of problem-solving. While Knovel includes accessible introductory content, it progressively delves into more advanced topics, helping students grasp the complexities of decision-making in engineering. This approach makes Knovel an ideal companion for students transitioning from academic study to professional engineering practice.
How is the tool used by educators?
For educators, the tool offers support starting in the foundational years of teaching, covering all aspects of project-based learning and beyond. It is also an efficient way for faculty to remain up-to-date with the latest information and data on key issues. Ultimately, it is educators who have the challenge of guiding students towards reputable, suitable, traceable information. In doing so, educators are helping students to understand that where they gather information, and how they use it, is in itself an ethical issue.
Knovel for Higher Education is an Elsevier product. As a publisher-neutral platform, Knovel helps engineering students explore foundational literature with interactive tools and data.
46% of EPC members already have access to Knovel. To brainstorm how you can make the best use of Knovel in your classroom, please contact: Susan Watson, susan.watson@elsevier.com.
Faculty and students can check their access to Knovel using their university email address at the following link: Account Verification – Knovel
Get Knovel to accelerate R&D, validate designs and prepare technical professionals. Innovate in record time with multidisciplinary knowledge you can trust: Knovel: Engineering innovation in record time
Any views, thoughts, and opinions expressed herein are solely that of the author(s) and do not necessarily reflect the views, opinions, policies, or position of the Engineering Professors’ Council or the Toolkit sponsors and supporters.
Authors: Ahmet Omurtag (Nottingham Trent University); Andrei Dragomir (National University of Singapore / University of Houston).
Professional situations: Communication; Honesty; Transparency; Informed consent; Misuse of data.
Educational level: Advanced.
Educational aim: Practising Ethical Analysis: engaging in a process by which ethical issues are defined, affected parties and consequences are identified, so that relevant moral principles can be applied to a situation in order to determine possible courses of action.
Learning and teaching notes:
This case involves Aziza, a biomedical engineer working for Neuraltrix, a hypothetical company that develops Brain-computer interfaces (BCI) for specialised applications. Aziza has always been curious about the brain and enthusiastic about using cutting-edge technologies to help people in their daily lives. Her team has designed a BCI that can measure brain activity non-invasively and, by applying machine learning algorithms, assess the job-related proficiency and expertise level of a person. She is leading the deployment of the new system in hospitals and medical schools, to be used in evaluating candidates being considered for consultant positions. In doing so, and to respond to requests to extend and use the BCI-based system in unforeseen ways, she finds herself compelled to weigh various ethical, legal and professional responsibilities.
This case study addresses two of AHEP 4’s themes: The Engineer and Society (acknowledging that engineering activity can have a significant societal impact) and Engineering Practice (the practical application of engineering concepts, tools and professional skills). To map this case study to AHEP outcomes specific to a programme under these themes, access AHEP 4 here and navigate to pages 30-31 and 35-37.
The dilemma in this case is presented in three parts. If desired, a teacher can use the Summary and Part one in isolation, but Parts two and three develop and complicate the concepts presented in the Summary and Part one to provide for additional learning. The case allows teachers the option to stop at multiple points for questions and/or activities as desired.
Learners have the opportunity to:
analyse the ethical dimensions of an engineering situation;
identify professional responsibilities of engineers in an ethical dilemma;
determine and defend a course of action in response to an ethical dilemma;
practise professional communication;
debate viable solutions to an ethical dilemma.
Teachers have the opportunity to:
highlight professional codes of ethics and their relevance to engineering situations;
address approaches to resolve interpersonal and/or professional conflict;
integrate technical content on software and/or cybersecurity;
informally evaluate students’ critical thinking and communication skills.
Brain-computer interfaces (BCIs) detect brain activity and utilise advanced signal analysis to identify features in the data that may be relevant to specific applications. These features might provide information about people’s thoughts and intentions or about their psychological traits or potential disorders, and may be interpreted for various purposes such as for medical diagnosis, for providing real-time feedback, or for interacting with external devices such as a computer. Some current non-invasive BCIs employ unobtrusive electroencephalography headsets or even optical (near-infrared) sensors to detect brain function and can be safe and convenient to use.
Evidence shows that the brains of people with specialised expertise have identifiable functional characteristics. Biomedical technology may translate this knowledge soon into BCIs that can be used for objectively assessing professional skills. Researchers already know that neural signals support features linked to levels of expertise, which may enable the assessment of job applicants or candidates for promotion or certification.
BCI technology would potentially benefit people by improving the match between people and their jobs, and allowing better and more nuanced career support. However, the BCI has access to additional information that may be sensitive or even troubling. For example, it could reveal a person’s health status (such as epilepsy or stroke), or it may suggest psychological traits ranging from unconscious racial bias to psychopathy. Someone sensitive about their privacy may be reluctant to consent to wearing a BCI.
In everyday life, we show what is on our minds through language and behaviour, which are normally under our control, and provide a buffer of privacy. BCIs with direct access to the brain and increasing capability to decode its activity may breach this buffer. Information collected by BCIs could be of interest not only to employers who will decide whether to hire and invest in a new employee, but also to health insurers, advertising agencies, or governments.
Optional STOP for questions and activities:
1. Activity: Risks of brain activity decoding – Identify the physical, ethical, and social difficulties that could result from the use of devices that have the ability to directly access the brain and decipher some of its psychological content such as thoughts, beliefs, and emotions.
2. Activity: Regulatory oversight – Investigate which organisations and regulatory bodies currently monitor and are responsible for the safe and ethical use of BCIs.
3. Activity: Technical integration – Investigate how BCIs work to translate brain activity into interpretable data.
Dilemma – Part one:
After the company, Neuraltrix, deployed their BCI and it had been in use for a year in several hospitals, its lead developer Aziza became part of the customer support team. While remaining proud and supportive of the technology, she had misgivings about some of its unexpected ramifications. She received the following requests from people and institutions for system modifications or for data sharing:
1. A hospital asked Neuraltrix for a technical modification that would allow the HR department to send data to their clinical neurophysiologists for “further analysis,” claiming that this might benefit people by potentially revealing a medical abnormality that might otherwise be missed.
2. An Artificial Intelligence research group partnering with Neuraltrix requested access to the data to improve their signal analysis algorithms.
3. A private health insurance company requested Neuraltrix provide access to the scan of someone who had applied for insurance coverage; they stated that they have a right to examine the scan just as life insurance agencies are allowed to perform health checks on potential customers.
4. An advertising agency asked Neuraltrix for access to their data to use them to fine-tune their customer behavioural prediction algorithms.
5. A government agency demanded access to the data to investigate a suspected case of “radicalisation”.
6. A prosecutor asked for access to the scan of a specific person because she had recently been the defendant in an assault case, where the prosecutor is gathering evidence of potential aggressive tendencies.
7. A defence attorney requested data because they were gathering potentially exonerating evidence, to prove that the defendant’s autonomy had been compromised by their brain states, following a line of argument known as “My brain made me do it.”
Optional STOP for questions and activities:
1. Activity: Identify legal issues – Students could research what laws or regulations apply to each case and consider various ways in which Neuraltrix could lawfully meet some of the above requests while rejecting others, and how their responses should be communicated within the company and to the requestor.
2. Activity: Identify ethical issues – Students could reflect on what might be the immediate ethical concerns related to sharing the data as requested.
3. Activity: Discussion or Reflection – Possible prompts:
Do you, as a biomedical engineer, have any duty to the people who have been scanned? Do you have more or less of a responsibility to these people or to Neuraltrix?
If you find that a fellow employee has already shared the data without telling others, how should you act? Should you worry that revealing this employee’s actions might cause distress or create distrust in the integrity of the entire system?Is there anyone else you should inform? Are there any risks you may be able to mitigate immediately?
Do you think the reasons and justifications given for the data requests listed above are legitimate?
Who owns the data collected by the BCI? Should it be protected? How, and for how long? Who should maintain it?
Dilemma – Part two:
The Neuraltrix BCI has an interface which allows users to provide informed consent before being scanned. The biomedical engineer developing the system was informed about a customer complaint which stated that the user had felt pressured to provide consent as the scan was part of a job interview. The complaint also stated that the user had not been aware of the extent of information gleaned from their brains, and that they would not have provided consent had been made aware of it.
Optional STOP for questions and activities:
1. Activity: Technical analysis – Students might try to determine if it is possible to design the BCI consent system and/or consent process to eliminate the difficulties cited in the complaint. Could the device be designed to automatically detect sensitive psychological content or allow the subject to stop the scan or retroactively erase the recording?
2. Activity: Determine the broader societal impact and the wider ethical context – Students should consider what issues are raised by the widespread availability of brain scans. This could be done in small groups or a larger classroom discussion.
Possible prompts:
On the one hand, human assessors can be subject to bias and inconsistencies and, from this point of view, algorithmic assessment leaving human assessors out of the loop may be viewed as progress. On the other hand, some “black-box” algorithms used by the BCI have been criticised for opacity, hidden biases, and the difficulty of scrutinising their decisions. If a user is dissatisfied with the BCI-enhanced assessment, should they be able to opt out of it?
If use of the Neuraltrix BCI became widespread, do you believe that humans could eventually irreversibly lose their assessment skills? Compare this with the potential loss of map-reading skills due to the easy access to Satellite Navigation systems.
Can we dispense with human opinion and make assessment processes entirely “objective”?
“Goodhart’s law,” named after the economist Charles Goodhart, states that when a measure is used as a tool, it becomes vulnerable to manipulation. Would Neuraltrix BCI create new opportunities for candidates to “game” the BCIs, and how would they do it?
Dilemma – Part three:
Neuraltrix BCI is about to launch its updated version, which features all data processing and storage moved to the cloud to facilitate interactive and mobile applications. This upgrade attracted investors and a major deal is about to be signed. The board is requesting a fast deployment from the management team and Aziza faces pressure from her managers to run final security checks and go live with the cloud version. During these checks, Aziza discovers a critical security issue which can be exploited once the BCI runs in the cloud, risking breaches in the database and algorithm. Managers believe this can be fixed after launch and request the engineer to start deployment and identify subsequent solutions to fix the security issue.
Optional STOP for questions and activities:
1. Activity: Students should consider if it is advisable for Aziza to follow requests from managers and the Neuraltrix BCI board and discuss possible consequences, or halt the new version deployment which may put at risk the new investment deal and possibly the future of the company.
2. Activity: Apply an analysis based on “Duty-Ethics” and “Rights Ethics.” This could be done in small groups (who would argue for management position and engineer position, respectively) or a larger classroom discussion. A tabulation approach with detailed pros and cons is recommended.
Should you, as a biomedical engineer, follow company rules and go ahead with manager’s requests or risk the future of the company (and possibly your job) and put deployment on hold until the security issue is fixed?
Act utilitarianism principle, as advocated by John Stuart Mill, focuses on individual actions rather than on rule, therefore, actions should be judged based on whether they resulted in the most good outcome in a certain situation. Should the Neuraltrix BCI management be guided by this principle or rather by a cost-benefit approach?
Any views, thoughts, and opinions expressed herein are solely that of the author(s) and do not necessarily reflect the views, opinions, policies, or position of the Engineering Professors’ Council or the Toolkit sponsors and supporters.
Authors: Dr Yujia Zhai (University of Hertfordshire); Associate Professor Scarlett Xiao (University of Hertfordshire).
Professional situations: Rigour; Informed consent; Misuse of data.
Educational level: Intermediate.
Educational aim: Gaining ethical knowledge. Knowing the sets of rules, theories, concepts, frameworks, and statements of duty, rights, or obligations that inform ethical attitudes, behaviours, and practices.
Learning and teaching notes:
This case study involves an engineer hired to develop and install an Industrial Internet of Things (IIoT) online machine monitoring system for a manufacturing company. The developments include designing the infrastructure of hardware and software, writing the operation manuals and setting policies. The project incorporates a variety of ethical components including law and policy, stakeholders, and risk analysis.
This case study addresses three of the themes from the Accreditation of Higher Education Programmes fourth edition (AHEP4): Design and Innovation (significant technical and intellectual challenges commensurate the level of study), the Engineer and Society (acknowledging that engineering activity can have a significant societal impact) and Engineering Practice (the practical application of engineering concepts, tools, and professional skills). To map this case study to AHEP outcomes specific to a programme under these themes, access AHEP 4 here and navigate to pages 30-31 and 35-37.
The dilemma in this case is presented in three parts. If desired, a teacher can use Part one in isolation, but Part two and Part three develop and complicate the concepts presented in Part one to provide for additional learning. The case study allows teachers the option to stop at multiple points for questions and/or activities as desired.
Learners have the opportunity to:
apply their ethical judgement relating to privacy and consent on the use of machine data;
determine the societal impact of a technical solution to a complex problem;
analyse risks associated with ethical concerns and justify their ethical decisions;
communicate these risks and judgements to both technical and non-technical audiences.
Teachers have the opportunity to:
highlight a range of ethical considerations within the scope of a complex engineering project;
introduce methods for risk analysis and ethical decision-making;
link Engineering Council statements of ethical principles with real world situations;
IIoT is a new technology that can provide accurate condition monitoring and predict component wear rates to optimise machine performance, thereby improving the machining precision of the workpiece and reducing the production cost.
Oxconn is a company that produces auto parts. The robotic manipulators and other automation machines on the production line have been developed at considerable cost and investment, and regular production line maintenance is essential to ensure its effective operation. The current maintenance scheme is based on routine check tests which are not reliable and efficient. Therefore Oxconn has decided to install an IIoT-based machine condition monitoring system. To achieve fast responses to any machine operation issues, the machine condition data collected in real time will be transferred to a cloud server for analysis, decision making, and predictive maintenance in the future.
Dilemma – Part one – Data protection on customers’ machines:
You are a leading engineer who has been hired by Oxconn to take charge of the project on the IIoT-based machine monitoring system, including designing the infrastructure of hardware and software, writing the operation manuals, setting policies, and getting the system up and running. With your background in robotic engineering and automation, you are expected to act as a technical advisor to Oxconn and liaise with the Facilities, Security, Operation, and Maintenance departments to ensure a smooth deployment. This is the first time you have worked on a project that involves real time data collection. So as part of your preparation for the project, you need to do some preliminary research as to what best practices, guidance, and regulations apply.
Optional STOP for questions and activities:
1. Discussion: What are the legal issues relating to machine condition monitoring? Machines’ real-time data allows for the identification of production status in a factory and is therefore considered as commercial data under GDPR and the Data Protection Act (2018). Are there rules specifically for IIoT, or are they the same no matter what technology is being used? Should IIoT regulations differ in any way? Why?
2. Discussion: Sharing data is a legally and ethically complex field. Are there any stakeholders with which the data could be shared? For instance, is it acceptable to share the data with an artificial intelligence research group or with the public? Why, or why not?
3. Discussion: Under GDPR, individuals must normally consent to their personal data being processed. For machine condition data, how should consent be handled in this case?
4. Discussion: What ethical codes relate to data security and privacy in an IIoT scenario?
5. Activity: Undertake a technical activity that relates to how IIoT-based machine monitoring systems are engineered.
6. Discussion: Based on your understanding of how IIoT-based machine monitoring systems are engineered, consider what additional risks, and what kind of risks (such as financial or operational), Oxconn might incur if depending on an entirely cloud-based system. How might these risks be mitigated from a technical and non-technical perspective?
Dilemma – Part two – Computer networks security issue brought by online monitoring systems:
The project has kicked off and a senior manager requests that a user interface (UI) be established specifically for the senior management team (SMT). Through this UI, the SMT members can have access to all the real-time data via their computers or mobiles and obtain the analysis result provided by artificial intelligence technology. You realise this has implications on the risk of accessing internal operating systems via the external information interface and networks. So as part of your preparation for the project, you need to investigate what platforms can be used and what risk analysis must be taken in implementation.
Optional STOP for questions and activities:
The following activities focus on macro-ethics. They address the wider ethical contexts of projects like the industrial data acquisition system.
1. Activity: Explore different manufacturers and their approaches to safety for both machines and operators.
2. Activity: Technical integration – Undertake a technical activity related to automation engineering and information engineering.
3. Activity: Research what happens with the data collected by IIoT. Who can access this data and how can the data analysis module manipulate the data?
4. Activity: Develop a risk management register, taking considerations of the findings from Activity 3 as well as the aspect of putting in place data security protocols and relevant training for SMT.
5. Discussion/activity: Use information in the Ethical Risk Assessment guide to help students consider how ethical issues are related to the risks they have just identified.
6. Discussion: In addition to cost-benefit analysis, how can the ethical factors be considered in designing the data analysis module?
7. Activity: Debate the appropriateness of installing and using the system for the SMT.
8. Discussion: What responsibilities do engineers have in developing these technologies?
Dilemma – Part three – Security breach and legal responsibility:
At the beginning of operation, the IIoT system with AI algorithms improved the efficiency of production lines by updating the parameters in robot operation and product recipes automatically. Recently, however, the efficiency degradation was observed, and after investigation, there were suspicions that the rules/data in AI algorithms have been subtly changed. Developers, contractors, operators, technicians and managers were all brought in to find out what’s going on.
Optional STOP for questions and activities:
1. Discussion: If there has been an illegal hack of the system, what might be the motive of cyber criminals?
2. Discussion: What are the impacts on company business? How could the impact of cyber-attacks on businesses be minimised?
3. Discussion: How could threats that come from internal employees, vendors, contractors or partners be prevented?
4. Discussion: When a security breach happens, what are the legal responsibilities for developers, contractors, operators, technicians and managers?
Any views, thoughts, and opinions expressed herein are solely that of the author(s) and do not necessarily reflect the views, opinions, policies, or position of the Engineering Professors’ Council or the Toolkit sponsors and supporters.
Activity: Prompts to facilitate discussion activities.
Author: Sarah Jayne Hitt, Ph.D. SFHEA (NMITE, Edinburgh Napier University).
Overview:
There are several points in this case during which an educator can facilitate a class discussion about relevant issues. Below are prompts for discussion questions and activities that can be used. These correspond with the stopping points outlined in the case. Each prompt could take up as little or as much time as the educator wishes, depending on where they want the focus of the discussion to be. The discussion prompts for Dilemma Part three are already well developed in the case study, so this enhancement focuses on expanding the prompts in Parts one and two.
Dilemma Part one – Discussion prompts:
1. Legal Issues. Give students ten minutes to individually or in groups do some online research on GDPR and the Data Protection Act (2018). In either small groups or as a large class, discuss the following prompts. You can explain that even if a person is not an expert in the law, it is important to try to understand the legal context. Indeed, an engineer is likely to have to interpret law and policy in their work. These questions invite critical thinking and informed consideration, but they do not necessarily have “right” answers and are suggestions that can help get a conversation started.
a. Are legal policies clear about how images of living persons should be managed when they are collected by technology of this kind?
b. What aspects of these laws might an engineer designing or deploying this system need to be aware of?
c. Do you think these laws are relevant when almost everyone walking around has a digital camera connected to the internet?
d. How could engineers help address legal or policy gaps through design choices?
2. Sharing Data. Before entering into a verbal discussion, either pass out the suggested questions listed in the case study on a worksheet or project on a screen. Have students spend five or ten minutes jotting down their personal responses. To understand the complexity of the issue, students could even create a quick mind map to show how different entities (police, security company, university, research group, etc.) interact on this issue. After the students spend some time in this personal reflection, educators could ask them to pair/share—turn to the person next to them and share what they wrote down. After about five minutes of this, each pair could amalgamate with another pair, with the educator giving them the prompt to report back to the full class on where they agree or disagree about the issues and why.
3. GDPR Consent. Before discussing this case particularly, ask students to describe a situation in which they had to give GDPR consent. Did they understand what they were doing, what the implications of consent are, and why? How did they feel about the process? Do they think it’s an appropriate system? This could be done as a large group, small group, or through individual reflection. Then turn the attention to this case and describe the change of perspective required here. Now instead of being the person who is asked for consent, you are the person requiring consent. Engineers are not lawyers, but engineers often are responsible for delivering legally compliant systems. If you were the engineer in charge in this case, what steps might you take to ensure consent is handled appropriately? This question could be answered in small groups, and then each group could report back to the larger class and a discussion could follow the report-backs.
4. Institutional Complexity. The questions listed in the case study relate to the fact that the building in which the facial recognition system will be used accommodates many different stakeholders. To help students with these questions, educators could divide the class into small groups, with each group representing one of the institutions or stakeholder groups (college, hospital, MTU, students, patients, public, etc.). Have each group investigate whether regulations related to captured images are different for their stakeholders, and debate if they should be different. What considerations will the engineer in the case have to account for related to that group? The findings can then be discussed as a large class.
Dilemma Part two – Discussion prompts:
The following questions relate to macroethical concerns, which means that the focus is on wider ethical contexts such as fairness, equality, responsibility, and implications.
1. Benefits and Burdens. To prepare to discuss the questions listed in the case study, students could make a chart of potential harms and potential benefits of the facial recognition system. They could do this individually, in pairs or small groups, or as a large class. Educators should encourage them to think deeply and broadly on this topic, and not just focus on the immediate, short-term implications. Once this chart is made, the questions listed in the case study could be discussed as a group, and students asked to weigh up these burdens and benefits. How did they make the choices as to when a burden should outweigh a benefit or vice versa?
2. Equality and Utility. To address the questions listed in the case study, students could do some preliminary individual or small group research on the accuracy of facial recognition systems for various population groups. The questions could then be discussed in pairs, small groups, or as a large class.
3. Engineer Responsibility. Engineers are experts that have much more specific technical knowledge and understanding than the general public. Indeed, the vast majority of people have no idea how a facial recognition system works and what the legal requirements are related to it, even if they are asked to give their consent. Does an engineer therefore have more of a responsibility to make people aware and reassure them? Or is an engineer just fulfilling their duty by doing what their boss says and making the system work? What could be problematic about taking either of those approaches?
Any views, thoughts, and opinions expressed herein are solely that of the author(s) and do not necessarily reflect the views, opinions, policies, or position of the Engineering Professors’ Council or the Toolkit sponsors and supporters.
Authors: Dr Nicola Whitehead (University of Wales Trinity Saint David); Professor Sarah Hitt (NMITE); Emma Crichton (Engineers Without Borders UK); Dr Sarah Junaid (Aston University); Professor Mike Sutcliffe (TEDI-London), Isobel Grimley (Engineering Professors’ Council).
Topic: Development and use of a facial recognition system.
Professional situations: Rigour, Informed consent, Misuse of data, Conflicts with leadership / management.
Educational level: Advanced.
Educational aim: To encourage ethical motivation. Ethical motivation occurs when a person is moved by a moral judgement, or when a moral judgement is a spur to a course of action.
Learning and teaching notes:
This case involves an engineer hired to manage the development and installation of a facial recognition project at a building used by university students, businesses and the public. It incorporates a variety of components including law and policy, stakeholder and risk analysis, and both macro- and micro-ethical elements. This example is UK-based: however, the instructor can adapt the content to better fit the laws and regulations surrounding facial recognition technology in other countries, if this would be beneficial.
This case study addresses two of AHEP 4’s themes: The Engineer and Society (acknowledging that engineering activity can have a significant societal impact) and Engineering Practice (the practical application of engineering concepts, tools and professional skills). To map this study to AHEP outcomes specific to a programme under these themes, access AHEP4here and navigate to pages 30-31 and 35-37.
This case is presented in three parts. If desired, a teacher can use Part one in isolation, but Part two (focusing on the wider ethical context of the case) and Part three (focusing on the potential actions the engineer could take)develop and complicate the concepts presented in Part one to provide for additional learning. The case study allows teachers the option to stop at multiple points for questions and / or activities as desired.
Learners have the opportunity to:
apply their ethical judgement to a case study relating to privacy and consent;
judge the societal impact of a technical solution to a complex problem;
make and justify an ethical decision;
analyse risks associated with micro-ethical and macro-ethical concerns;
communicate these risks and judgements to both technical and non-technical audiences.
Teachers have the opportunity to:
highlight a range of ethical considerations within the scope of a complex engineering project;
introduce methods for risk analysis and ethical decision-making;
evaluate critical thinking, argumentation, and communication skills;
Metropolitan Technical University (MTU), based in the UK, has an urban campus and many of its buildings are located in the city centre. A new student housing development in this area will be shared by MTU, a local college, and medical residents doing short rotations at the local hospital. The building has a public café on the ground floor and a couple of classrooms used by the university.
The housing development sits alongside a common route for parades and protests. In the wake of demonstrations by Extinction Rebellion and Black Lives Matter, students have raised concerns to the property manager about safety. Despite an existing system of CCTV cameras and swipe cards, the university decides to install an enhanced security system, built around facial recognition technology that would enable access to the building and cross-reference with crime databases. To comply with GDPR, building residents will be required to give explicit consent before the system is implemented. Visitors without a student ID (such as café customers) will be buzzed in, but their image will be captured and cross-referenced before entry. A side benefit of the system is that MTU’s department of Artificial Intelligence Research will help with the installation and maintenance, as well as studying how it works, in order to make improvements.
Dilemma – Part one:
You are an engineer who has been hired by MTU to take charge of the facial recognition system installation project, including setting policies and getting the system operational. With your background in AI engineering, you are expected to act as a technical advisor to MTU and liaise with the Facilities, Security and Computing departments to ensure a smooth deployment. This is the first time you have worked on a project that involves image capture. So as part of your preparation for the project, you need to do some preliminary research as to what best practices, guidance, and regulations apply.
Optional STOP for questions and activities:
1. Discussion: What are the legal issues relating to image capture? Images allow for the identification of living persons and are therefore considered as personal data under GDPR and theData Protection Act (2018).
2. Discussion: Sharing data is a legally and ethically complex field. Is it appropriate to share images captured with the police? If not the police, then whose crime database will you use? Is it acceptable to share the data with the Artificial Intelligence Research group? Why, or why not?
3. Discussion: Under GDPR, individuals must normally consent to their personal data being processed. How should consent be handled in this case?
4. Discussion: Does the fact that the building will accommodate students from three different institutions (MTU, the local college, and the hospital) complicate these issues? Are regulations related to students’ captured images different than those related to public image capture?
5. Activity: Undertake a technical activity that relates to how facial recognition systems are engineered.
Dilemma – Part two:
The project has kicked off, and one of its deliverables is to establish the policies and safeguards that will govern the system. You convened a meeting of project stakeholders to determine what rules need to be built into the system’s software and presented a list of questions to help you make technical decisions. The questions you asked were:
Should students be able to opt in or out of image capture?
Should visitors be told that their image will be captured?
What happens if a student living in the housing development decides that they no longer wish to take part in the image recognition project?
What you had thought would be a quick meeting to agree basic principles turned out to be very lengthy and complex. You were surprised at the variety of perspectives and how heated the discussions became. The discussions raised some questions in your own mind as to the risks of the facial recognition system.
Optional STOP for questions and activities:
The following activities focus on macro-ethics. This seeks to understand the wider ethical contexts of projects like the facial recognition system.
1. Activity: Stakeholder mapping – Who are all the stakeholders and what might their positions and perspectives be? Is there a difference between the priorities of the different stakeholders?
2. Activity: There are many different values competing for priority here. Identify these values, discuss and debate how they should be weighed in the context of the project.
3. Activity: Risks can be understood as objective and / or subjective. Research the difference between these two types of risk, and identify which type(s) of risks exist related to the project.
4. Discussion: Which groups or individuals are potentially harmed by the technology and which potentially benefit? How should we go about setting priorities when there are competing harms and benefits?
5. Discussion: Does the technology used treat everyone from your stakeholders’ list equally? Should the needs of society as a whole outweigh the needs of the individual?
6. Activity: Make and defend an argument as to the appropriateness of installing and using the system.
7. Discussion: What responsibilities do engineers have in developing these technologies?
Dilemma – Part three:
A few days later, you were forwarded a screenshot of a social media post that heavily criticised the proposed facial recognition system. It was unclear where the post had originated, but it had clearly been shared and promoted among both students and the public raising concerns about privacy and transparency. Your boss believes this outcry endangers the project and has requested that you make a public statement on behalf of MTU, reaffirming its commitment to installing the system.
You share the concerns, but have been employed to complete the project. You understand that suggesting it should be abandoned, would most likely risk your job. What will you tell your boss? How will you prepare your public statement?
Optional STOP for questions and activities:
Micro-ethics concerns individuals and their responses to specific situations. The following steps are intended to help students develop their ability to practise moral analysis by considering the problem in a structured way and work towards possible solutions that they can analyse critically.
1. Discussion: What are the problems here?
You are an employee of MTU and have a responsibility to be a representative for its interests. However, you can see that the university’s actions create significant problems relating to privacy and consent and may be ethically or legally questionable.
2. Discussion: What are the possible courses of action you can take as an employee?
Students can be prompted to consider what different approaches they might adopt, such as the following, but can also develop their own possible responses.
You could take the university line and refuse to consider any compromise. After all, you have a duty of care towards the students.
You could act as a whistleblower and contact the Information Commissioner’s Office,or the press, with the university’s plans.
You could look for changes in the hardware setup for the system. Can the cameras be placed so that they only capture people coming into the building without recording anyone else?
You could look for changes in the software setup for the system. What level of accuracy is needed to declare a match between the image and the reference image before the doors will open?
You could look to make changes in the data management processes. How long will the data be stored? Which database(s) will images be checked against? What are the data security implications of implementing this system?
Are there other alternatives available to you?
3. Discussion: Which is the best approach and why? – Interrogate the pros and cons of each possible course of action including the ethical, practical, cost, local relationship and the reputational damage implications. Students should decide on their own preferred course of action and explain why the balance of pros and cons is preferable to other options. The students may wish to consider this from other perspectives, such as:
What would the best outcome be if cost was no object?
What course of action would be taken if different perspectives were chosen as the priority. For example, if the personal privacy perspective was the main priority, what action would be taken, compared with action taken if the cost to the university were the main priority?
What are the wider implications of the use of image recognition in public spaces and how can these be mitigated?
Are there any other technologies that would solve the security problem without the ethical implications?
What are the possible solutions open to you?
Are there any short-term solutions versus longer-term solutions?
4. Activity: Public Communication – Students can practise writing a press release, giving an interview, or making a public statement about the case and the decision that they make.
5. Activity: Reflection – Students can reflect on how this case study has enabled them to see the situation from different angles. Has it motivated them to understand the ethical concerns and to come to an acceptable conclusion.
Enhancements:
An enhancement for this case study can be found here.
Any views, thoughts, and opinions expressed herein are solely that of the author(s) and do not necessarily reflect the views, opinions, policies, or position of the Engineering Professors’ Council or the Toolkit sponsors and supporters.
Authors: Ahmet Omurtag (Nottingham Trent University); Andrei Dragomir (National University of Singapore / University of Houston).