Education of IoT in an industrial
context
Creating educational material for industrial workers
Karl Söderby
Interaktionsdesign Bachelor
22.5HP 2020
Abstract
As the rise of Industry 4.0 sheds light on many emerging technologies, our society will change with it. While it brings forth many positive aspects, it cannot be ignored the socio-economic problems we may face in the future. Many jobs will be transformed, manual labour such as order picking, forklift driving will be vanishing, and humanity will have to adapt, as we have for the previous industrial revolutions.
Educating the industrial workers that face unemployment due to automation is an important ethical matter, but can we as humans develop our knowledge with the technology, as opposed to adapting to it? This thesis uses methods of interaction design to create an alternative educational format, for industrial workers to learn about the Internet of Things, an essential component of Industry 4.0.
The result of this is TIOTTA (Teaching Internet of Things Through Application), a contextual learning material designed together with industrial workers.
Keywords: Internet of Things, Industry 4.0, Interaction design, ethical
Acknowledgements
I would like to thank the industrial workers that participated in various design activities. Their engagement and interest were admirable. I would also like to express sincere gratitude to my supervisor David Cuartielles for the incredible support throughout the thesis, to Susan for months of encouragement and emotional support, and to Lenard for his excellent consultation on design methodology.
Abstract ... 2 Acknowledgements ...3 1 Introduction ... 7 1.1 Purpose ... 7 1.2 Contribution ... 8 1.3 Limitations ... 8 1.4 Ethical considerations ... 8 1.5 Research question ... 9 2 Theoretical background ... 9
2.1 Past, present, and future of Internet of Things ... 9
2.2 Fourth industrial revolution – the new worker ... 10
2.3 Architecture of IoT... 11
2.4 Internet of Things in education ... 12
2.5 Physical and digital representation of things ... 13
2.6 Structuring learning conditions ... 13
2.7 Motivating learning ... 14
2.8 Summarising background theory ... 15
3 Related designs ... 15
3.1 Blokdots ... 15
3.2 LittleBits ... 16
3.3 Grove starter kit ... 17
3.4 Summary ... 18
4 Methodology ... 19
4.1 Research through design ... 19
4.2 Interaction design research... 20
4.3 Prototyping through an iterative cycle ... 20
4.4 Participatory activist research ... 22
4.5 Collecting data from participants ... 22
4.6 Summary of methodology ... 23
5.1 Analysis of existing workshop formats ... 24
5.2 Interviews ... 25
5.2.1 Interviewing workshop attendees ... 25
5.2.2 Interviewing industrial workers ... 25
5.2.3 Concluding first interviews ... 26
5.3 Educational format ... 26
5.4 Concept generation for physical design ... 29
5.4.1 The integrated box ... 29
5.4.2 The brain box... 30
5.4.3 The connector shield ... 31
5.4.4 Summary of first concepts ... 32
5.5 First iteration of physical prototype ... 33
5.5.1 The role of the prototype ... 33
5.5.2 Look and feel of the prototype ... 33
5.5.3 Implementation of the prototype ... 34
5.6 Digital platform... 36
5.6.1 Structure and digital tools ... 37
5.6.2 Selection of theory... 37
5.6.3 Preparation and practical activity ... 38
5.6.4 Testing, evaluate and discussion ... 39
5.7 Concluding first iteration of prototypes ... 39
5.8 Testing of TIOTTA with industrial workers ... 40
5.8.1 Selecting methods for workshop ... 40
5.8.2 Recruiting participants for a workshop ... 40
5.8.3 Hosting the workshop ... 41
5.8.4 Analysing data recorded from observation ... 41
5.8.5 Analysing data recorded from interviews... 43
5.8.6 Conclusion of workshop and interviews ... 43
5.9 Second iteration of prototype ... 44
5.9.1 Second iteration of physical prototype ... 44
5.9.2 Second round of user testing ...47
6 Evaluating results of TIOTTA ... 48
6.1 Educational format ... 48
6.1.2 Structure of the educational format... 49
6.1.3 Facilitating learning conditions ... 49
6.2 Reaching an understanding of use ... 50
6.2.1 Understanding how to use it ... 50
6.2.2 Understanding how it works ... 50
6.2.3 Understanding for what it can be used ... 50
6.3 Manifestation of things ... 51
6.3.1 Things in digital and physical realms ... 51
6.3.2 Physical design ... 51
6.4 Evaluation of methods... 52
7 Discussion ...53
7.1 Societal challenge cloaked in technical brilliance ...53
7.2 Balancing technical knowledge and conceptualisation ... 54
7.3 Prototyping educational material ... 54
7.4 Future work ... 54
8 Conclusion ... 55
9 References ... 57
1 Introduction
This thesis explores several aspects of teaching Internet of Things (IoT) related concepts at an introductory level. It aims to provide an educational format that incorporates theoretical background on IoT concepts, learning through practical application, and a contextual learning approach. The focus group of this thesis are industrial workers with no post-secondary education. I will investigate several notions, such as motivation to learn new technologies, learning climates, educational models, and the materials needed to teach IoT concepts. This thesis respects the 4th industrial revolution as a major challenge for industrial workers, and aims to provide knowledge nuggets that can be built upon to shape future careers that will require a more technological set of skills (Benešová & Tupa, 2017).
The research and design activities are conducted following ethical considerations presented in the work Pedagogy of the Oppressed by Freire (1996). Social groups, defined as oppressed, may lack power in society and cannot be liberated by forcing them to learn. The educational material needs to be co-created with representatives of the group, since without their reflection and participation, the liberator sees the group as a mass that can be manipulated, thus an oppressive pedagogical approach is applied (Freire, 1996).
In the context of the overlap between IoT and interaction design, this thesis will explore the correlation between digital and physical representations of things of the Internet. Through two iterations of sketches and prototypes it will aim at creating a seamless interaction between what is commonly referred to as things by Oriwoh and Conrad (2015), and how a thing is represented in both physical and digital format. The aim is to create a simple understanding of what can be the form and function of an Internet-thing, and how they form complex IoT systems, together.
1.1 Purpose
The main intention of this thesis is to develop an educational format that provides an engaging and motivating learning experience, teaching different IoT concepts through theory and application, that can be appropriated for industrial workers lacking a post-secondary education. The contents of the educational format is of an introductory level, and aims to provide a background of technical skills, theory, and practice that an industrial worker may use as a stepping stone to further develop technological skills that can be applied in a working context.
1.2 Contribution
This thesis aims to provide an educational format to be used outside a higher education environment. It describes the major challenges people may face in the future due to technological gaps, and how they can be better prepared. It does not aim to deliver a full-scale educational model that provides a set of skills that can be directly applied in a work environment, but rather as an introduction to a range of topics that could prove essential in the future. It aims to motivate workers to adopt a “life-long learning” approach that will both benefit the worker and the employer. It also aims to motivate workers to further educate themselves through self-learning, workshops, or other programs that do not necessarily have to be part of a post-secondary educational programme.
A contribution to the field of interaction design is how this thesis explores the design of educational material that can bridge the gap between educational material and real-life context, thus increasing the perceived value of a learning opportunity. It focuses mainly on how the material can be designed to create an engaging and motivating experience of learning IoT concepts on an introductory level. It presents three criteria that can be used as guidelines for designing educational material within technology: conceptual, technical and, practical.
1.3 Limitations
The main limitation of this thesis is the choice of target group. The field of IoT encompasses many different industries and professions, and the results would have been more refined if the prototypes were tested in more than one context. The time frame of this thesis also limits the possibility to further follow the progression of skill of the participants, which is one of the main purposes, to see the development of an individual's knowledge.
Another limitation is that due to the COVID-19 crisis of 2020, interactions with people were limited, resulting in fewer design activities conducted with the social group studied.
1.4 Ethical considerations
The ethical considerations in this thesis project are two-fold. Firstly, it follows the guidelines for ethical research conduct from Vetenskapsrådet (2017) on how to safely collect data from participants of a study and informing them of the purpose of the study. The interviews, observations, and tests were conducted under full transparency, and each participant in this study were granted anonymity.
Secondly, the educational format that is presented as a result of this thesis, follows the work Pedagogy of the Oppressed by Freire (1996). He presents several ethical guidelines on how educational approaches to social groups
that may lack power in society should be co-created with its representatives. Otherwise, we apply an oppressive educational approach. The social group in this study are industrial workers that may face socio-economic struggles due to Industry 4.0.
1.5 Research question
How might we design educational material for learning IoT concepts for industrial workers?
2 Theoretical background
2.1 Past, present, and future of Internet of Things
The rise of the Internet of Things was inspired by the emerging Radio Frequency Identification (RFID) technology, which brought the possibility of obtaining information about an object by tagging it. Furthermore the Internet of Things brought the possibility of bringing physical objects into the cyber world, by using technologies such as RFID and Near Field Communication (NFC) to identify and share information about physical objects on the Internet (Madakam, Ramaswamy & Tripathi, 2015).
The “Internet of Things” was conceptualized in the article titled the Computer for the 21st Century by Weiser (1991), as he talks about machines and devices as being part of a larger ubiquitous network. At the time Weiser’s article was written, the Internet had not yet reached a general audience, and devices were still being designed to execute tasks and store data locally. Heading into the 21st century, things connected to the Internet have grown exponentially. Much of the credit goes to rapid innovation and development of smartphones, increased accessibility to the Internet, and millions of online services that are currently operating worldwide. It was not until 1999 that the term “Internet of Things” was coined by Kevin Ashton (Ashton, 2009).
Statistics presented in 2003 by Evans (2011) show that the ratio of connected devices per person was 0.08 (500 million devices, 6.3 billion people), whereas, in 2010, the number had grown to be 1.84 (12.5 billion devices, 6.8 billion people). It is predicted that in 2020 the number has almost quadrupled to reach 38 billion devices (Smith, 2020), and is estimated to generate 344 billion US dollars in revenue, with investments in the IoT sector reaching 1.4 trillion US dollars by 2021 (Banafa, 2020).
Madakam et al. present an overview of technologies closest associated with which is depicted in Figure 1 (2015).
Radio Frequency Identification (RFID)
Internet Protocol (IP) Electronic Product Code (EPC
Wireless Fidelity (Wi-Fi) Bluetooth ZigBee
Actuators 4.10. Wireless Sensor Networks (WSN) Artificial Intelligence (AI) Near Field Communication (NFC) Barcode
Figure 1. Technologies associated with the Internet of Things.
There has been an increasing popularity of using Wireless Sensor Networks (WSN) in manufacturing, logistics, environment tracking and other areas akin (Hunkeler, 2008). These networks are different from other traditional networks such as the Internet, as they typically are never directly connected to the Internet, but send data to gateways, using techniques such as Messaging Query Telemetry Transport (MQTT). These gateways can then communicate with other applications.
The IoT sector is expected a continuation of growth in the future. Perera et al. presents statistics that several markets, such as RFID and Smart Cities, are growing rapidly (2014). But as they grow, many industries and professions will inevitably change. In the next 20 years, the rapid development of IoT will cause a loss of jobs to computerisation as Frey and Osborne describe it (2017). This will bring a major socio-economic problem (2017).
While today’s society is becoming increasingly aware of the technologies surrounding them, the need for understanding them is increasing significantly at the same time. It is predicted that many jobs such as carpenters, chefs, bakers, or machine operators are likely to be almost fully replaced by automated machines. Instead, the need for computer scientists, technicians, maintenance workers, and data analysts is going to increase even further. Thus, education needs to shift focus towards emerging technologies shaping the workers of the future. Workers that may not have completed a post-secondary education.
2.2 Fourth industrial revolution – the new worker
It was in Germany, 2011, that Industry 4.0 was first mentioned, as an economic policy concept (Roblek, Meško & Krapež, 2016). It follows the first (mechanical), second (electrical) and third (digital) industrial revolutions, where the fourth revolution separates itself from its predecessors by incorporating many different technologies, including artificial intelligence,
renewable energy, autonomous transportation & manufacturing, and IoT (Pfeiffer, 2015).
Benešová and Tupa predict the need for a four-phase process to happen over the span of several decades transforming industrial manufacturing and logistics (2017). The first phase will be the digital representation of a factory in real-time; the second will introduce re-engineered automated machines as part of a production flow; the third will use data recorded in the first and second phase to finally; the fourth phase will see an almost completely autonomous manufacturing and logistics process. The description is similar to the notion of a “technological adaptation curve”, initially presented by researchers at Iowa University in 1957, where the four stages of adopting a technology: innovators, early adopters, masses, and laggards (Greengard, 2015, p. xxii). According to Greengard, the present IoT lives somewhere between innovation and early adopters but will likely move swiftly to reach the masses stage (2015).
Where today a worker in logistics might manually drive a forklift, print labels for shipping goods, or manually pick orders, future workers will have to adapt to roles involving maintenance, analysis of data, troubleshooting, and more advanced problem-solving. To adapt the current workforce to the fourth industrial revolution is a task of great magnitude, but if it fails, it can result in catastrophic socio-economic outcomes. Therefore, the need for education of relevant topics, tools, and skills for the workforce lacking a post-secondary education is essential to the success of the future.
As described by Frey and Osborne, the “potential extent of technological unemployment caused by a rapid technological progression rate” will lead to many jobs lost in the process (2017). Roblek et al. also supports this prediction, stating that the gains in productivity by advanced technologies will destroy less-demanding jobs (2016).
2.3 Architecture of IoT
While there are many ways an IoT system can be designed, the core infrastructure contains a few essential components that are needed to create what is called an IoT vertical. We can begin by looking at the lowest level of the infrastructure: the sensors or actuators that record real-time data. These sensors are typically part of a WSN, which could scale from a few sensors up to millions of individual sensors. These sensors send data through different protocols, such as MQTT, to a router that is connected to the internet. The router can be recognized as middleware, which operates between systems and applications (Alkhabbas, Spalazzese & Davidsson, 2019).
The data is then analysed by using different services and can then either visualise the data in a dashboard, or act upon it. An example of this can be a storage unit in a warehouse detecting that the supply of screws is running out of stock. The data is sent through the cloud and visualised in a dashboard,
notifying whoever is watching to act. Furthermore, in a highly intelligent system, the sensor might notify a machine that produces these screws to prioritise the manufacturing, and even start production autonomously. By being connected to the internet, these physical actions have no requirements to occur in the same facility, and thus a global autonomous process can be implemented.
Figure 2. Example of a possible IoT ecosystem. A sensor in a warehouse in Sweden sends data to a router (middleware) that connects to the internet, which sends data to a cloud service that analyses the data. It then sends a request to a machine in China to start manufacturing. The data is also accessible through a dashboard and can be viewed from an office in Germany.
Visualising data in dashboards is an integral part of IoT, as it is a way for humans to interpret, gain insights and present conclusions from data recorded (Matheus, Janssen & Maheshwari, 2018). Dashboards are essential in any IoT system, as they help humans understand the data. Dashboards in cloud services are typically considered the central point of any system: many affords a two-way communication, where a human can retrieve and send data to different IoT devices.
Matheus et al. describes visual dashboards as an instrument to reduce information asymmetry (2018). By incorporating visual dashboards in an IoT system, data can easily be distributed globally.
2.4 Internet of Things in education
As stated by Gul et al., Internet of Things in an educational context consists of two main criteria: enhancement for academic infrastructure and teaching fundamental concepts of computer science (2017). IoT as a tool for education has already reached a semi-ubiquitous stage, such as booking classrooms, managing attendance and the use of VR and QR codes is now technology widely used. An example of this is a study made by Gómez et al. where
students of a computer science course used RFID and NFC tags to obtain information about certain elements of a computer (2013). Students could interact with these objects with their smartphones, accessing a wide range of information regarding a specific part of a computer without any help from the teacher.
As for learning fundamental IoT concepts, it is framed by Kortuem et al. that some of the main topics while learning IoT involves programming, algorithms, distribution and collaboration, creative and collaborative design, and computing in society (2012). While a technological set of skills is required such as programming and writing algorithms, the design process and understanding the impact in society is of equal importance. There are many clear benefits of IoT technology, but the disadvantages cannot be ignored. Topics such as data security, privacy, and cyberterrorism to name a few are highly relevant as they can have dire consequences if not accounted for.
2.5 Physical and digital representation of things
Things in IoT can be defined in a variety of ways. As stated by Oriwoh and Conrad, a thing can be a living or non-living entity (2015). On the other hand, it could also represent a combination of IoT elements in a process, what we would then call the Internet of Processes. Contextual information —also known as metadata— will in turn create an Internet of Related Things, or the network of things containing information about things.
The characterizing of the Internet of Things made by Alkhabbas et al. defines ‘things’ as building blocks of IoT systems (2019). They identify several dimensions of a thing, such as the three types of things: smart things, sensors and actuators, and gateways. The first ones are physical objects that have processing and communication capabilities. Sensors and actuators sense or perform physical actions. Gateways link various sensors and actuators, and process and communicate the information further to a middleware component, such as a router.
This vision is partial as it is not representing the full vertical explained in section 2.3. It is lacking a representation of the digital version of things. This thesis focuses on the use of the service Arduino IoT cloud, a service allows the user to connect a Wi-Fi compatible microcontroller board and set up ways of collecting and sending data to the board (Arduino, n.d.-a).
2.6 Structuring learning conditions
Many manufacturing and logistics operations are completely reliant on the knowledge, skills, and presence of the worker, and without them, the operation is halted. This limitation in the use of time, constraints the amount of learning opportunities for the workers. Any educational action must be carefully planned. As Moon explains that if a learning opportunity is short,
the design needs to be focused, and all elements need to be considered to achieve an impact on the learner (2004).
Learning opportunities involving programming, microcontrollers, and wiring electronics, can be overwhelming. Blikstein and Sipitakiat points out that high school and non-technical undergraduate students often fails at understanding basic electronic concepts and infrastructure of basic components, and anxiety is often generated amongst learners (2011). It is estimated in a study by Gul et al., that the average American student spends 1025 hours in school, out of which 308 hours are spent on preparation of materials, transition or at the beginning or end of a class (roughly 30%), which can be paralleled with a workshop that uses complex material, the number naturally increases (2017).
According to Moon (2004), shorter learning opportunities are often “divorced” from their context, where the content often does not relate to the context of a participants’ workplace, and therefore the impact of the learning opportunity is damaged, resulting in a poor outcome. She also states that reflection plays a big part in determining the impact the learning opportunity has on the participant. Reflection helps to get current practice integrated with new knowledge and the re-organizing of new ideas to adapt to a work routine.
2.7 Motivating learning
Motivation is always a vital ingredient in the recipe for successful learning, where Jenkins and Davy see it as the function of two factors: expectancy and value (2002). They frame it as a mathematical equation: motivation =
expectancy x value, whereas if one is lacking, the result is zero.
The first factor, expectancy, is the interpretation from the student: does the student expect to succeed in the learning activity, will she meet the goal that is set out by the teacher? This is an interesting notion, particularly in introductory programming, where a student may not understand the goal itself from the start. The goal of any learning activity must be clear and simple, else the student may lose one of the two major factors.
The second factor, value, is determined by either the student’s personal goal —to either succeed or to avoid failure— or the students’ interest in increasing their knowledge in a specific field. Schiefele compares the interest in studying with learning strategies, the researcher suggests that interest is strongly correlated with the seeking of information, elaboration, and critical thinking (1991).
Norman’s view on popular classroom teaching models, is that the way we teach with textbooks and lectures is just the easiest method, benefiting the teacher of the class (2004, p. 205). He suggests that inspiring students with topics and letting them struggle with the concepts is a far more useful approach: struggle is dynamic, and the struggle can be enjoyable if the topic is intriguing.
2.8 Summarising background theory
The background theory on Internet of Things (IoT) and Industry 4.0 presents two large phenomena that are currently changing the world, and how industrial workers may face socio-economic struggle in the future (Frey & Osborne, 2017). Furthermore, the digital and physical representations and general understanding of things are explained by the definitions of Alkhabbas et al. (2019) and Oriwoh and Conrad (2015). The perspective of what a thing is, what characteristics it may have and how it fits in an ecosystem is particularly important, when creating educational material within the field, and aids the process of designing for an object that traverses between physical and virtual spaces.
IoT presents a basic problem of abstraction that may represent a challenge to the learner. It is a multi-layered technology not only at the infrastructural level (Alkhabbas et al., 2019) but also at the conceptual one (Oriwoh & Conrad, 2015). This must be considered when designing engaging educational experiences (Norman, 2004, p. 205).
I will attempt to create an educational format that focuses on achieving “maximum impact” during a learning opportunity, described by Moon (2004), following the principle of Jenkins and Davy, the formula for motivation being the product of expectancy times value (2002). I will build my educational framework following Kortuem et al.’s one and cover aspects within programming, algorithms, distribution and collaboration, creative and collaborative design, and computing in society (2012).
3 Related designs
This section presents three designs that inspired the physical design activities in this thesis project.
3.1 Blokdots
Blokdots is a project that created at the university of Schwäbisch Gmünd in Germany. They use modular sensor and actuator blocks to build circuitry, and data recorded is streamed to a digital dashboard (Blokdots, 2020). The project uses an Arduino microcontroller as processing power and leverages the use of auxiliary connectors instead of jumper wires to create circuitry, while simplifying the coding aspect using block programming methods. Figure 3 depicts the setup of a Blokdots circuit. A motor and a potentiometer are connected to a central unit powered by an Arduino, which is then connected to a computer. The blocks are then programmed using Blokdots
own application, and data from components can then be visualised in their dashboard.
Figure 3. Physical setup of a Blokdots circuit. Courtesy of Blokdots.
Blokdots achieves a visually pleasing physical design of complex material and makes it more accessible for non-technical users.
3.2 LittleBits
LittleBits is an educational platform for electronics that use simple plug and play connections, using magnets instead of wires to build circuits. They offer over 60 electronic building blocks, which they call bits. One of which is called
CloudBit, which provides a possibility to connect to a cloud service and send
information to the Internet (Singh & Kapoor, 2017). The bits are connected to each other form circuits, without the need to program them, as each bit has its own circuit and functionality.
LittleBits use colour-coding as a method to quickly identify the primary function of a bit. There are four types of bits: power, input, output, and extension, each marked clearly with a distinct colour. This categorisation helps the user quickly identify the functionality.
Figure 4. A variety of LittleBits sensors and actuators. Retrieved from WikiMedia commons. LittleBits is suited for children over the age of 8, and their first programming experiences. They describe it as ideal for elementary school curriculums, where in high school, the student can start coding user more advanced tools, such as Arduino.
The LittleBits’ learning experience does not emphasise writing text-based programs and operates almost fully in the physical realm. The bits are already pre-programmed, where each bit has its own logic.
3.3 Grove starter kit
The Grove starter kit, designed by Seeed Studios (2020) uses sensors, actuators, and displays to create circuits. This kit is designed to be used with an Arduino board (n.d.-b). The kit includes components that can detect and produce sound, measure light intensity, and emit light. They focus on simple connections, using 4-wire connectors with a mechanical fail-safe to reduce the risk of faulty connections and minimise the time spent building circuits.
Figure 5. An ultrasonic sensor with a Grove connector. Retrieved from WikiMedia commons. This physical configuration allows the user to focus on the programming aspects as opposed to the assembling of circuits, removing many preparation elements. The Grove starter kit comes with a set number of digital and analog components, but also components that use I2C, a serial protocol using two signal wires. Digital temperature and humidity sensors and displays are examples using the I2C protocol.
This kit provides a shield, which can be directly mounted on an Arduino board. This allows the easy connections of components to be directly attached to the Arduino, without the use of a breadboard. This configuration is helpful for beginners to get started, and diverts time spent on circuitry to instead focus on learning the software.
3.4 Summary
These designs are presented as educational materials that currently exist for learning physical computing. They present different approaches in their physical design but share a common feature. They all use a modular approach, which removes the time spent on standard approaches to build circuits (soldering or using breadboards). LittleBits however, does not place any emphasis on the software, as their building blocks already come pre-programmed.
LittleBits does however provide a great distinction between their different “bits”. Their colour-coding technique helps the learner quickly identify a type of component. As seen in figure 5, the Grove starter kit’s components does not have such system. For the untrained eye, making a distinction between the different electronic components can be a difficult task.
Blokdots provides both an easy way of connecting the blocks, but also focuses on the use of dashboards, a vital part of an IoT system. The Blokdots configuration is designed to make circuitry an enjoyable task, by replacing multiple wires with the use of auxiliary connectors. Using these creates an intuitive feeling when building circuits, as these connectors are universally recognised, and are used in any home (headphones, speaker systems and cars are just some examples). By being familiar with just one part of the circuit it can help empower a novice user, as a regular electronic component can invoke insecurity and frustration.
4 Methodology
This section introduces the methods that were used throughout the design process of this thesis.
It introduces the general Research through Design (RtD) approach, the main strategies of interaction design research that influenced the design process and presents different authors’ views on what a prototype is, and how it can be evaluated. It also presents the Participatory Activist Research (PATR) method used in the design process, and methods used to collect and analyse data.
4.1 Research through design
This thesis project follows a Research through Design (RtD) approach. This framework by Zimmerman, Forlizzi and Evenson’s presents several different aspects to consider when conducting design research (2007). They quote that “the process of criticizing potential solutions, is to reframe the problem to create the right thing” (2007). By their definition, it is to ground a problem using empirical research, gaining several aspects of what a problem is, followed by an ideation process through several iterations (2007). Gaver’s view on RtD is that while we build our knowledge on others’ work, we should also suggest alternatives and new constructions (2012). He also states that we are “not developing theories that are never wrong but developing theories that sometimes are right” (2012).
Zimmerman et al., also suggests using the method “four critical lenses”, process, invention, relevance, and extensibility. These are described as:
• Process, the process used in the project is detailed and can be reproduced by others, while justifying the selection of tools and methods used in the process.
• Invention, the ability to weave together various topics that addresses the situation.
• Relevance, to deliver an impact on the world, the solution must be relevant to the problem it is attempting to solve and.
• Extensibility: the ability to build upon the knowledge presented, e.g. using it in a future research project.
These four critical lenses helped evaluate many aspects of this thesis project (2007).
4.2 Interaction design research
One strategy to conduct interaction design research is described by Jonas Löwgren as “exploring the potentials of a design material, design ideal or technology” (2007, pp. 6-7). This strategy explores the possible space of a future artefact, as in this thesis, is the educational material to learn IoT in an industrial context.
This is further informed by another strategy of Löwgren, to conduct a participatory design process, where future users are viewed as the experts of the field and not objects of a study (2007). This is further elaborated on in section 4.4.
4.3 Prototyping through an iterative cycle
There are numerous definitions of what a prototype is. As Houde and Hill describes it: for a programmer, it is a test program, for interaction designers, it is the behaviour of an object on a screen (1997). The way a prototype can be defined is not confined to its physical appearance, as Houde and Hill explains that even a brick can be a prototype if it represents a future artefact (1997). Stolterman presents another view on prototypes by explaining that prototypes “traverses a design space” creating meaningful knowledge in the process and are formed manifestations of design ideas (2008).
Mogensen phrases prototyping as being the construction of the future, and that the process can be described as making guesses at different solutions, and through iterations improve these guesses (1992). Mogensen also highlights the danger of ‘becoming blind’, meaning that a prototype is made concrete and the designer does not question the foundations of it (1992). In a model presented by Houde and Hill, dimensions of a prototype are defined as role, look and feel and implementation depicted in figure 6 (1997). Role can best be described by the functionality of a prototype, look and feel are what the users see, feels and hears - the sensory experience - and implementation is described as how the prototype actually works, how it is designed to function.
Figure 6. The three dimensions of a prototype (Houde & Hill, 1997).
Throughout the design process, several sketches and prototypes were produced, following an “iterative cycle approach” that is commonly used in human-centred design (Norman, 2013, pp. 222). This method is useful when applying human-centred design, as concrete design decisions are rarely made, and can be changed by each iteration.
Figure 7. The iterative cycle design process, illustration based on a model from Norman (2013, pp. 222).
The prototypes were also later evaluated, using the framework “three levels of design”, visceral, behavioural, and reflective (Norman, 2013, pp. 50-54). The visceral level of processing is the immediate response to an artefact: what we see, touch and feel are what matters as they are subconscious responses. The behavioural level also lives in the “immediate” processing of thoughts, and it can best be described as the actions that we perform before we reflect consciously. Norman describes all three criteria as essential but suggests that the reflective level is the most powerful. It is the reflective process that we
experience both during and after an interaction with the artefact. This is the level that may affect conversations made about the artefact even when it is not present or the one that may determine the longevity in history.
4.4 Participatory activist research
Participatory Activist Research (PATR), belongs to the Action Research (AR) research family (Emerald & Martin, 2012, p. 1). The methodology focuses on the facilitation of social change through research, where the researcher has an active interest in changing the context. PATR is practiced through being a participant, activist, and researcher at the same time, and they are not exercised separately (2012, pp. 8-9).
As this thesis is aiming to design educational material for use in an industrial context, I identify myself as the designer that researches the topic, while also identifying myself as being part of the social group studied. This is due to my own experience of working in the same industrial facility in which the workers participating in this study work in. Some core principles of PATR can be identified (2012, p. 4).
• Inclusiveness of all stakeholders. • Collective and collaborative action. • Passionate acting.
• Producing an environment in which trust and respect flourish. The PATR and AR methodology is applied in this project by not conducting research on participants, but rather for and with. PATR presents the value of looking at the participants as the key to identify problems, collection and analysis of data and implementing solutions, as Emerald and Martin describes it: “it explores and expresses the values important to the target group’s lives” (2012, p. 7).
Participatory types of research highlight the importance and value of telling unwelcome truths and how only investigation of reality can help us transform it (Kemmis, 2006). Kemmis states that if we do not encounter any unwelcome truths, we have not applied any critical research. As this thesis investigates possible future socio-economic challenges and aims to provide a solution to these, it is immensely important that this approach is practiced.
Eva Brandt also suggests in the International Handbook of Participatory
Design to engage participants in future possible scenarios. This presents an
opportunity to be part of a solution (Simonsen & Robertson, 2013, p. 176).
4.5 Collecting data from participants
Interviewing was used to gain an understanding of problems that exist in physical computing workshops and were used as the main method to gather qualitative data from industrial workers.
Interviews were conducted using an in-depth interviewing style, based on the methodology presented by Muratovski (2015, pp. 82-83). This method uses a conversational approach, starting with casual dialogue and reducing the sense of the interviewee being interrogated. The interviews also followed the guidelines of Sanders and Stappers, where they suggest not becoming a teacher in an interview, and that your role as an interviewer is to ensure that participants express themselves (2012, p. 171).
Observation also helped identify several issues concerning tangibility, emotional states, and spatial choices: how the participants interact with the prototypes presented; what emotional states they go through at which specific point of the experience.
Analysis of this data was mainly done by using affinity diagrams (Wilson, 2013, pp. 34-35). This method is used to organise qualitative data from field studies, by grouping them based on their natural relationship.
4.6 Summary of methodology
The methods presented in this section will be used in various stages of the design process. The overall design process uses the iterative cycle approach presented by Norman (2013, p 222.), using the RtD approach, following the “four critical lenses” by Zimmerman et al. (2007) and Gaver’s suggestions to create alternative constructions based on other’s works (2012).
The prototyping activities follows Stolterman’s views on prototypes: prototypes are manifestations of design ideas (2008). It also follows the three dimensions of a prototype by Houde and Hill: role, look and feel and implementation (1997). The prototypes are also evaluated using the three levels of design: visceral, behavioural, and reflective. (Norman, 2013, pp. 50-54).
The PATR approach does not identify the target group as users but as
participants since they are considered co-creators in this study (Emerald &
Martin, 2012). This is further emphasised by Kemmis, as he suggests attempting to uncover welcome truths by investigating reality, and how these are valuable to the impact a design may have on society (2006). The main method to gather qualitative data is through in-depth interviews (Muratovski, 2015, pp. 82-83). Affinity diagrams are then used to group data based on their natural relationship (Wilson, 2013, pp. 34-35).
5 Design process
The design process is divided into 9 smaller subsections.
5.2 interviews former participants of these workshops to identify pain points
and their perceived learning experience
5.3 presents a new educational format that includes feedback from design
activities in 5.1 and 5.2
5.4 shows sketches of three physical design proposals 5.5 explains the first iteration of a physical prototype 5.6 explains the creation of a digital learning platform
5.7 summarises the prototyping activities of the first iteration 5.8 explains the first test of the educational format
5.9 incorporates feedback from the first iteration and a second test is
conducted
5.1 Analysis of existing workshop formats
As Mogensen describes it, to make initial ‘guesses’ of what to prototype, one must have a basic idea of what to create (1992). Therefore, the design process began with an analysis and evaluation of different physical computing workshops at Malmö University.
The material used in these workshops was hardware and software from Arduino, an electronic prototyping platform to write basic programs and, read and control electronic circuits (2020). Based on my experience from these workshops, I created a diagram of the different elements of the workshop, seen in Figure 8.
5.2 Interviews
5.2.1 Interviewing workshop attendees
Following the creation of the workflow diagram, I conducted interviews with people that had attended physical computing workshops at Malmö University. These workshops used Arduino hardware and software as part of a prototyping course.
In total 5 people were interviewed, using an in-depth interviewing technique to generate more qualitative data (Muratovski, 2015, pp. 82-83). They were asked a series of question regarding their experiences in learning different physical computing concepts, and what pain points they could identify. Through the interviews, participants were able to identify both positive and negative experiences with the educational format they were using. Multiple participants expressed a clear frustration in learning hardware and software simultaneously and had become overwhelmed as a result. Some identified that the learning outcome was unclear, and that the relevancy of the topics introduced did not match their education. This was a divided opinion, however, as two of the participants were satisfied with the content and found the topics interesting.
Three out of five participants did however express a clear frustration when discussing electronic components and circuits. This ranged from preparational tasks to understanding the basic concepts of electronics. One clear issue was identifying a specific electronic component in their kit, which was described as near impossible in some cases.
They did, however, express that the learning experience was dynamic, had elements of playfulness and at the end, had resulted in a general positive experience. They did however point out that the need for supporting content was lacking and was often a frustrating element to locate. Programming and electronic concepts had to be found on the Internet, and as one participant expressed: it was hard to find an answer when you did not understand the question.
5.2.2 Interviewing industrial workers
The next set of interviews were conducted with three workers at an industrial facility in Sweden to gain an understanding of their work environment, their motivation to learn new skills, their perspective on how important it is to learn new skills and to what degree they interact with technology in their work.
Participants expressed that troubleshooting, machine operation and maintenance was a highly desired work tasks, as they presented a dynamic workday. One participant quoted:
“The troubleshooting and fixing common errors of a machine is boring and repetitive. Although when there is a new type of error with unknown cause, you need to work against the clock to identify and solve the error, often within minutes.”
All participants interacted daily with different technologies: digital dashboards, inventory systems, label printers and networks of sensors were some mentioned. Infrared sensors were highlighted as a vital cog of their operation, where monitoring these through digital dashboards and maintenance was a considered a normal work task.
2 out of 3 participants demonstrated a clear motivation to learn more about emerging technologies and were open to educational opportunities such as workplace training and online courses. The participants expected more tasks to be automated in their future work environment, and predicted monitoring, troubleshooting and maintenance would increase as manual labour such as forklift driving would decrease. A desire to learn programming was identified among all participants. None of the participants had heard the term IoT before.
5.2.3 Concluding first interviews
The interviews with participants of the physical computing workshop presented issues with working with electronic circuits, and the preparational tasks of identifying the right components in their kits. They also pointed out a clear lack of supporting content, as they were frustrated at finding information on the Internet themselves.
The interviews with industrial worker gave a good insight to their work routine, current interaction with technology and what tasks they enjoyed. The workers demonstrated a good understanding of how sensor networks work, and how their daily work routine involved the use of dashboards, which they used to monitor different elements of the factory. While they had not been exposed to the term IoT, several comments made by them demonstrated an understanding of many elements of an IoT system. Also expressed by participants was the desire to learning programming.
This suggests that education of IoT concepts can prove relevant for industrial workers. The next sections will describe the creation of an educational format that will be used to teach IoT concepts through practical application.
5.3 Educational format
The insights from previous activities lead to designing a new educational format. This format centres around the Arduino IoT cloud, and the practical application of an IoT system. This system includes sensors and actuators, which in this project are identified as things (Alkhabbas et al., 2019). The Arduino microcontroller serves as the middleware for these sensors,
processing incoming data and sends it further to the cloud, where data can be viewed in the Arduino IoT cloud dashboard (see figure 9).
Figure 9. The Arduino IoT Cloud dashboard. This screen capture shows the real-time data of several sensors.
The first iteration of the educational format is designed as a workshop to last approximately one hour. To fit in theoretical introduction, preparation, practical activity, and testing and evaluating the knowledge, I created a flow diagram of all elements. The red blocks represent physical action, purple digital actions, green are milestones and yellow are a segment of the workshop. The white blocks are the theoretical topics to be introduced.
Figure 11.The second part of the workshop: activity, test and evaluation, and discussion. The outcome of this activity provided new challenges: provide background theory for the workshop, create a learning activity for the workshop and create a physical prototype that:
(i) Can reduce the time spent on preparation and circuit building.
(ii)
Can manifest a virtual thing of the Internet in physical format(iii)
Can simulate a basic IoT system.5.4 Concept generation for physical design
In this section, I present three sketches that can accommodate the requirements in the last section.
5.4.1 The integrated box
The “integrated box” (see figure 12) is a concept that completely removes the activity of circuit building, and presents instead a physical box that includes
the microcontroller and electronic components, thus shifting the focus to the software part.
Figure 12: Sketch of the integrated box
5.4.2 The brain box
The “brain box” (see figure 13) presents a brain unit and several sensor and actuator blocks. These blocks can connect to the brain unit, using one single connection instead of using a breadboard, jumper wires and raw electronic components.
Figure 13: Sketch of the brain box
5.4.3 The connector shield
The “connector shield” (see figure 14) involves the design of a new circuit board, where like the brain box, it has connections that components can directly be plugged into. This concept uses no other physical material than bare electronics.
Figure 14: The connector shield
5.4.4 Summary of first concepts
The integrated box is an all-inclusive design. With no physical assembly required, the focus can shift towards the digital aspects. It does not allow any addition of external material, which limits extensibility i.e. adding extra sensors. It also does not provide a great link between digital and physical properties, as the device includes everything (thus creating custom dashboards is rendered useless, since there is only one possible outcome). The brain box encapsulates the electronic components in different blocks. They are then connected to a brain block. This provides a dynamic experience as different blocks can be chosen for different types of activities, while highlighting each block as its own thing, with its own function.
The third concept, the connector shield, follows a common approach when designing educational material for physical computing. This sketch includes a programmable circuit board (PCB) that has multiple connector slots, where a microcontroller can connect to it. Instead of blocks, the raw electronics components can be connected directly to the PCB, similarly to the brain
block. This concept is currently a very popular approach to learning electronics and is inspired by the Grove starter kit.
The Brain Box aligned best with the requirements set out. By making it modular, it increases the possible variations of sensors and actuators, and through encapsulating them in blocks, the physical design can be altered to aid the learning of IoT concepts.
5.5 First iteration of physical prototype
The prototyping process began by consulting Houde and Hills’ three dimensions of a prototype: role, look and feel, and implementation (1997). The role that the prototype has in this educational format is to simplify circuit building and represent an IoT architecture. The look and feel regards how a thing of the Internet is manifested in a physical format. The implementation is how the prototype will be constructed, and how it will function.
5.5.1 The role of the prototype
The circuit building needed to be simplified to reduce time spent on preparational tasks. Since the target group has no prior experience in building circuits, I decided to use Molex connectors to reduce time spent on circuit building. The design of the connector does not allow any faulty connections, where up to four wires can be connected simultaneously with ease, where the circuits can be made inside each component block.
Two types of blocks were designed: the brain block and the component blocks. The component blocks represent things, where the brain block represents a middleware component (Alkhabbas et al., 2019).
5.5.2 Look and feel of the prototype
To manifest a thing of the Internet in physical format, an understanding of which formats a thing existed is required. A sensor in the IoT cloud environment manifests in three instances: the code, the dashboard, and the physical realm, see figure 15.
Figure 15. The three categories: physical, code and visualisation.
The next step is to identify a link between these instances. Depicted in figure 16, marked in yellow are the segments where an ultrasonic sensor is involved. A clear link between the dashboard and code is the variable name, which contains the data the component is recording.
Figure 16. Ultrasonic sensor marked in yellow across three instances.
To link the physical realm to the dashboard and code, I decided to use the variable name as the link, by making it clearly visible in the physical design of the component blocks.
5.5.3 Implementation of the prototype
The purpose of the first physical prototype is to incorporate the results from the analysis of the platform, previous interviews, and the concept generation. The first prototype uses a medium-density fibreboard (MDF), and using laser
cutting and CAD software to produce the different blocks. The first prototype included five blocks, depicted below in figure 17.
Component used Type Functionality
distance block Ultrasonic sensor analog Measure distance
screen block OLED Screen digital Print messages or
values
wheel block Potentiometer analog Produce an analog
value
sound block Piezo digital Play different sounds
temperature block
DHT11 temperature & humidity sensor
digital Record temperature and humidity locally Figure 17. Description of component blocks
The physical form of the component blocks is square and rectangular, while the design of the brain block is larger to establish a hierarchical structure. The brain block have pin slots mounted at its top side, where an Arduino microcontroller can be mounted, and have 5 connectors on the side, each marked with what pin it is connected to (A1, A2, D1, D2, SCREEN).
The brain block is designed to mimic a router, a middleware widely recognised to be capable of connecting to the Internet.
The first iteration of the physical prototype can be seen in figure 18, and a schematic can be seen in figure 19.
Figure 18. Brain block with an Arduino microcontroller mounted to the left, component blocks to the right.
Figure 19. Schematic of the blocks.
5.6 Digital platform
To support using the prototype with the Arduino IoT cloud, and theoretical support, I created a digital platform. This digital platform provides descriptions on how to do a step-by-step assembly of each component block, and how to configure the cloud service to work together with the physical prototype. As Moon describes shorter learning opportunities as often divorced from their context, the content created is designed based on
industrial workers descriptions of their work routine (2004). This contextual approach can aid the process of creating a meaningful experience for the participants, as they can relate the knowledge to their own context.
The digital format is designed using a self-guided approach, to be used without a teacher, and is set at communicating only introductory knowledge of IoT.
5.6.1 Structure and digital tools
The structure of the digital format is divided into five sections: theory, preparation, practical activity, testing and evaluation, and discussion.
It follows the flow diagram of the educational format created in 5.3, and is designed as one page of content, to store information in the same place.
5.6.2 Selection of theory
The theoretical selection included introductions to the general understanding of IoT and Industry 4.0. The other selections were linked to the material used in the workshop: basic microcontroller theory, electronic concepts, and programming concepts (Arduino programming language, a subset of C++). Figure 20 depicts a part of the theoretical section.
5.6.3 Preparation and practical activity
In the preparation and practical activity, the learner identifies the necessary digital and physical material. The physical blocks are visualised in the content, with simplified schematic on how to connect them (see figure 21). The learner then receives step-by-step instructions on how to create a ‘thing’ in the Arduino IoT cloud interface. Once a thing is created, they need to create properties in the cloud. These properties represent variables in a code that is autogenerated after a property is created.
Figure 21. A visual representation of the physical actions required
The instructions prompt the learner to use specific variables, which are visualised on the physical component blocks. The learner is then required to copy code segments from the digital platform into the code in the Arduino IoT cloud platform, thus getting an experience of working with a text-based code editor. Once the code is complete, the code is compiled and uploaded to the Arduino, which starts communicating with the cloud service over Wi-Fi.
5.6.4 Testing, evaluate and discussion
After completing the activity, the learner can interact with the physical component blocks, and see the data change in the cloud dashboard. They are invited to test out each block to understand how it works. This is also described in the content.
Figure 22. Expressing functionality of each block
The learner is then encouraged to identify problems in their working context, that can be solved using the basic IoT system they created. For this, a few practical examples of problem and solutions are provided.
5.7 Concluding first iteration of prototypes
The design activities generated a prototype that will be tested with industrial workers that have little to no experience with microcontrollers, programming and IoT systems. A link to the platform can be found in Appendix 1.
From this point of the thesis, the digital platform and physical prototype are referred to as one artefact named TIOTTA (Teaching Internet of Things Through Application). This definition is made to unify the different software, hardware, and digital content that TIOTTA constitutes. Figure 23 describes the different elements of TIOTTA.
Figure 23. Elements of TIOTTA
5.8 Testing of TIOTTA with industrial workers
5.8.1 Selecting methods for workshop
Observation was selected as a method to record participants’ performance, reactions and comments made during the testing of the prototype. This helped evaluate what elements of the prototype that was going through to the next iteration (Muratovski, 2015 p. 149). For the last part of the workshop, discussion, I was involved in the ideation, where I assumed the role of a participant. This was to help the other participants with ideation techniques, such as identifying problems, presenting a solution and the result of that solution.
Post testing, in-depth interviews were used to evaluate several aspects of the workshop. This technique requires a set of pre-structured questions to ask, but where the sequence of them is irrelevant, and to elaborate on a specific topic to uncover details that may not have occurred during the tests (Muratovski, 2015, pp 82-83).
5.8.2 Recruiting participants for a workshop
Two of the industrial workers that were interviewed earlier agreed to participate in the workshop. They had no prior knowledge in programming and electronics. The participants were expected to participate for around 1-2 hours, including an introduction to the workshop, a short break and post-activity interviews. Figure 24 describes each participant.
Age Role Years of
experience in the field
Participant 1 26 Forklift driver, machine operator, logistics 6 years Participant 2 27 Forklift driver, machine operator, logistics 7 years
Figure 24. Participants of the workshop
The participants were informed about the intention of the workshop and agreed to sign a letter of consent before any participation. This letter of consent followed Vetenskapsrådet’s guidelines on ethical conduct (2017). The participants were informed of:
• the contents of the workshop,
• their right to refuse further participation,
• their right to complete anonymity.
Permission to use images and quotes from participants were granted by signing the letter of consent. This can be found in Appendix 4.
5.8.3 Hosting the workshop
The workshop lasted for approximately one hour, where the participants were guided by the digital content to complete the activity designed. They were limited in receiving instructions and were asked to think out loud at any time during the test, to evaluate how well TIOTTA could be run without a teacher or educator present. The two participants worked in pair to complete the activity.
The data and insights recorded from the workshop were divided into two categories: findings through observation, and findings through post-workshop interviews.
5.8.4 Analysing data recorded from observation
Data from observations were analysed using an affinity diagram (Wilson, 2013, pp. 34-35). This resulted in six groups of data concerning the: digital platform, Arduino Create environment, contextualization, physical interaction, motivation, and emotions. See Figure 25 for an overview.
Figure 25. Overview of data from observing participants of the workshop, using an affinity diagram.
A main insight from the observation was how the participants performed the physical actions described in the digital platform. The participants managed to follow instructions from the platform to complete a circuit of 5 components using the component blocks in less than 30 seconds, but took over two minutes to mount the microcontroller on the brain block and connect the USB cable from the computer, since it had not been explained in the content. The participants displayed various levels of motivation during the workshop. It decreased during the theoretical part and increased during the practical activity. Most of the theory was labelled too complicated to understand, but they found the IoT and Industry 4.0 topics intriguing, as it was more conceptual and easier to digest.
While participant 1 was creating the digital properties in the cloud dashboard, participant 2 identified the physical component blocks, and an understanding of the connection between the two was achieved at an early stage.
The discussion helped the participants link the knowledge to their working context. In this activity, I actively participated in the discussion, as I had experience of their work environment. Together, we identified several small problems that could be improved by implementing an IoT system (the IoT system being the sensors and actuators of the prototype, and the use of the cloud dashboard). These problems circulated storage tracking, triggering alarms remotely and tracking temperature of different spaces in the facility. As Eva Brandt also explains, by engaging them in future possible scenarios, they can be part of the solution (Simonsen & Robertson, 2013, p. 176). Participants pointed out some desired functionalities with the prototype during the discussion. They identified the tracking of noise as an essential quality. The noises in their industrial facility indicates that something is malfunctioning (or is about to) and if a sensor could capture this information
it would be extremely useful. They also highlighted that simple button presses to indicate a “yes or no” situation could prove extremely useful.
5.8.5 Analysing data recorded from interviews
The participants described the workshop as satisfactory and when asked of its relevancy to their current occupation they described it as moderate to high. Participant 1 quoted:
“It felt like I was building the system we are working with daily, from scratch. I recognise some patterns between the dashboard from our system and the dashboard that we created.”
Both participants described the testing, evaluating, and discussion part as most engaging. While they performed well in the practical activity, they expressed that it turned into a boring and repetitive task. This was due to the requirement of manually creating properties in the cloud interface.
The participants demonstrated an understanding of how the logic of a basic program worked but expressed that they would fail in replicating it without any instruction, as expected. Both participants expressed that access to the digital platform should have been granted days prior to the testing, as they would have more time to reflect on it.
The circuit building was described as an easy activity but was frustrating, with how many cables were present when testing the activity. Participants were however aided by the variable name that was printed on each block. They expressed a desire to use it in their working context, which would be impossible due to the lack of portability of the prototype.
Participants were encouraged to discuss the physical qualities of the different blocks. While they were informed that they were testing a low-fidelity prototype, they pointed out some physical qualities needed to improve. They expressed a discomfort in touching the material (laser cutting MDF produces burned glue residue), and that each component block needed to be better distinguished from each other.
5.8.6 Conclusion of workshop and interviews
The observation and interviews from the workshop presented some issues that needed to be addressed:
• Skipping through the theoretical part. It is suggested by participants to be granted access to it in advance, so they could read and reflect on it better, as they were interested in the subject, but overwhelmed by the time frame.
• A large amount of cables reduces the portability of the prototype. This can be linked with their desire to use the prototype in a working context.
