Robot Zora : friend or foe? : an exploratory study about the emotional attachment of elderly to robot Zora

(1)

ROBOT ZORA; FRIEND O R FOE?

AN EXPLORA TORY STUDY ABOUT THE EMO TION AL A TTACHMEN T O F ELDER LY TO ROBO T

ZORA

MSc Medical Anthropology and Sociology Author: Els Dekker Student Number: 10764666 Supervisor: Jeannette Pols Second Reader: Stuart Blume

(2)

ABSTRACT

Zora is a humanoid robot invented for revalidation purposes in elderly care. Zora is a tool used by physiotherapist (among others) that aims to provide them with more possibilities for individual assistance during a group session. In this study I investigate if and how elderly interact with robot Zora and how this possible affective relationship works. My research took place in a residential care home in Belgium. Through observations and Interviews I found that Zora is more than a tool. The interaction between Zora and the observed users is complex and fluid. The object becomes objectified as well as humanized and gender is given and changed over time. When the robot is used for animation activities the elderly respond differently than when the robot is showing exercises. How the robot is perceived thus depends on the way in which it is used. There seems to be a fragile boundary between the robot as a tool and the robot as a friend. I found that we should not look at robot Zora as a finished object, but have to see the robot as a progress.

(3)

AKNOWLEDGEMENT

I want to start by thanking my supervisor Jeannette Pols for her guidance and patience during the process of this thesis. I am very grateful that you stood by my side during the difficulties which I stumbled upon. Your guidance and constructive criticism is most appreciated. Your expertise on the topic was enlightening, thank you for all your help.

Next I want to thank Tommy Deblieck from QBMT for his help with introducing me to my field of study. Without you it would have been very difficult, if not impossible, to conduct this research.

This gratitude is extended to the direction and staff of Sint-Anna for letting me snoop around, you were all very welcoming which made it easier for me to participate and conduct my research. I am especially very grateful for the effort the staff of Kiné and Ergotherapy made. Special thanks to E. who guided me (and introduced me to everyone) during my time in Sint-Anna WZC. My personal feeling is that someone is very lucky to be under your professional, warm and enthusiastic care when living in Sint-Anna WZC.

Although the residents I observed and interviewed probably won’t read this paper, I still want to take a moment here to thank them for their participation in this study. Sometimes it was difficult to understand your dialect or non-verbal communication (I think that goes both ways), but I nonetheless enjoyed being around you and been given the opportunity to see how you live your daily lives.

Last but definitely not least I want to thank all people who are close to me for their patience during my self-inflicted social isolation while working on this thesis. I will come out of this cave soon!

(4)

1) INTRODUCTION

With an increase in the use of technology in our daily lives, we also see an increase in the use of medical technology. Technology used in the medical world consists not only of operating robots and monitoring tools within hospitals and laboratories, but gradually also made an entrance in the day-to-day life of patients. Patients are now, for instance, able to monitor their blood pressure at home. These patient devices make it possible for patients to monitor their health, keep in contact with medical professionals and/or can provide them the possibility to act fast in case of emergency.

The multiplicity of medical technological devices, which are used by patients without the assistance of a professional, are often called ‘Telecare devices’ (Hartvigsen&Botsen, 2008; Pols, 2011). These telecare devices are different in nature and purpose. For instance we have

‘personal alarms’ which are portable alarms that elderly can wear on a cord around their neck. If they fall they can press the alarm for help. This device thus is made for safety purposes. The alarm won’t prevent the fall, but it can make the patient feel more safe and ensures that help will arrive quickly when they fall, which is an uncertainty without the alarm.

Another example of a telecare device is a blood-pressure measure device, which is used, as the name says, to measure the blood-pressure of the patient, the device then digitally sends the results to the hospital. Medical staff can then monitor if the patient is measuring on a regular basis and if the blood pressure is as it should be. The medical staff can contact the patient when there are abnormalities, which can prevent hospital take-in. This device thus monitors the patient’s health and is made for prevention purposes; to prevent other health problems that can result from a too high or too low blood pressure. These two devices are already established in the medical world.

A new form of technology which still is gaining ground are medical robots. There are robots used in dangerous areas to save people, robots that can perform surgery, but also robots that can assist medical professionals and help people when at home. Robot Zora is such a robot which is used in elderly care for revalidation and animation purposes. Telecare devices and robots can make life easier for the patient and the caregiver, they can prevent hospital take-ins, can provide safety and can help the patient with staying fit. In this research I want to look at how robot Zora is used. I also want to know how people interact with Zora and which effects this has on their emotional wellbeing. I want to study which emotions people attach to their interaction with robot Zora. Does robot Zora then becomes a new friend for the elderly?

A big concern in society with telecare devices in elderly care is that the devices might isolate people, because there is less need for interaction between the patient and the nurse or

(7)

doctor (Perry et al, 2010). Does robot Zora interfere in the relationship between the elderly and the staff? On the other hand medical technology can also be seen as a possible solution for an aging society(Botsis et al., 2008). With less government money for medical healthcare in the Netherlands, a declining number of people who work in healthcare and a growing aging population, many people are worried that healthcare for the elderly will become unaffordable for the average person. Elderly are then depended upon their relatives. Relatives will have to invite the elderly into their own homes to care of them (Eindhovens Dagblad, 2014; Wind, 2013; Tonkens, 2014). The question is that if technology, such as robot Zora, can bridge this gap, then how does this change elderly care. Will elderly bond with the technology or will it make elderly care impersonal? If the elderly are able to bond with robot Zora, they might not see the care as less personal. How elderly bond with such a technological object could thus have a big influence on how they are perceiving the provided care. This leads me to the following research question;

Research question

How can we understand human-robot interaction in institutional medical care for elderly people, and how is this interaction shaped by the affective relationship the elderly possibly have with Robot Zora?

(8)

2) THEORETICAL FRAMEWORK

In this theoretical framework I want to give an idea of the technology and field that I will explore with this study. The theories used are meant to bring guidance and will help to shape ideas about technology and robots in elderly care.

2.1 THE TECHNOLOGY

In this first part of the theoretical framework I want to look at the technology used in elderly care and robot Zora in particular. I will also take into account the connotations tied to the technology that exists now and to the technology of the future.

2.1.1 INTRODUCING ZORA

First I will give a general introduction about how robot Zora came into existence, the appearance of the robot and for which purposes (in elderly care) it can be used.

In Belgium robot Zora (Zora, 2014) was developed by Tommy Deblieck and Fabrice Goffin from QBMT. The robot is used in Belgian nursing homes and hospitals. The robot also had a trial in a Dutch nursing home.

The shell of robot Zora, the hardware so to say, is made by a different company called; Aldebaran Robotics, located in France. This robot there is called NAO (A. Moon, 2014). Since mid march 2014 NAO came available for the consumer market with a price starting from € 5687,-. QBMT bought only the hardware of robot NAO. QBMT developed software specializing on revalidation for elderly, and by doing so thus created a new product called Zora. The name ZORA is an acronym, it means Zorg Ouderen Revalidatie en Activatie {translated; Care Elderly

Revalidation and Activation}. This acronym is later changed to Zorg Ouderen Revalidatie en Animatie {translated; Care Elderly Revalidation and Animation}. QMBT found that animation was an unintended, but a most appreciated function of Zora.

Zora is a humanoid robot, the robot is 57 cm tall. Zora has a white hard skin, with coloured accents (blue, red, green or orange). The robot has two speakers on the sides of the head and has two eyes which can follow movement and have little coloured lights in them which can flicker. Zora has a small mouth and a small hole in the middle of its head. This hole is a camera which can be used to see and record the environment of the robot. Zora has a female

(9)

voice and can make hand- and arm movements while talking. The voice Zora uses is in Dutch (Netherlands) dialect.

As mentioned before Zora is originally made for revalidation purposes. Zora can therefore perform tasks such as showing people how to perform exercises with verbal and nonverbal instructions. It can also be put to use for social activities as such as bingo and dancing. Zora can read the news and tell what time it is.

The same as with a Smartphone or tablet Zora is controlled through an application (app) wherein the different functions of the robot can be updated, added and removed. In the beginning Zora was controlled with a remote control, now the robot is controlled with a tablet. This enables more possibilities and freedom in choosing and composing exercises. Robot Zora is maybe best described as a social empathetic robot, always in a good mood, always ready for a dance or a game. Zora can follow faces and can also notice change in an environment. Because the robot is made with all these supposedly ‘human’ characteristics it brings up the question how people see the robot and whether people attach emotions to this interaction.

2.1.2 TECHNOLOGY IN ELDERLY CARE

In elderly care many different types of technologies are used. As mentioned earlier these telecare devices can be used to quantify health and healthcare. There are devices which can be used to give the residents more autonomy and more social contact with their relatives. For example video calling technology can be used to get elderly in contact with distant family

members. Robot Zora offers a similar function. This function is not yet been used, but this type of technology can help prevent or lessen loneliness or social isolation.

Social isolation is an underlying red thread in this research. If elderly experience an affective relationship with Zora then this can have an impact on their experience of togetherness or loneliness. Social isolation happens when a person has little or no contact with other people or groups in society, and is unpleased with this lack of social contact (Gardner et al., 1999, as described in Findlay, 2003). People with ill health become less independent and are therefore less able to go outside and interact with others (Perry et al., 2010). The same technology is however also seen as having the opposite effect on people. Elderly who experience social isolation might value the face-to-face contact with their medical caregivers. A telecare device does not completely end this contact, but the contact becomes digitalized. The quantity of contact might not outweigh the quality of face-to-face contact. For this reason the feeling of

(10)

The fear of social isolation can thus also be a reason why elderly choose not to use a given telecare device. If elderly are afraid that robot Zora is a symbol for less contact with their caregivers then why would they emotionally invest in such technology? The elderly will then have to be motivated in order to use the robot. Pols &Willems (2011) speak of taming the technology. Patients have to find a device interesting enough in order to work with it. This acquires effort from the people who are providing the device. If the patients don’t know why the device might be helpful to them, and what benefits it can bring, then why would they start using it in the way the provider wants them to. This is not only an explanation for why people won’t use a device, but also for misuse of a device. For robots this can be similar. Is robot Zora

then made interesting enough for the elderly to use or will they abstain from using the robot? The multiplicity of telecare devices makes it difficult to say whether a telecare device is a good choice for a patient. It is very difficult to decide as there can be multiple positive and negative effects for both patient and health institution. The positive effects might be: less health problems, less health costs, more control over health problems, more independence, more (digital) contact with others, feelings of safety and so on. While the negative effects could be: less (offline) contact with others, feelings of isolation, feeling unsafe, feelings of being controlled. Some of the possible positive effects are similar to some of the negative effects. A device can thus bring one person a feeling of safety and control over his/her health, while another person might feel unsafe because of the constant awareness of being ill or the fear of losing social ties. Which effect will robot Zora have on the elderly who use robot Zora?

How the technology is created and how the supposed user is seen by the creator is also of importance. In turn how the end product is seen and used by the user is also something to be aware of.

Technology, in this case robot Zora, has elements which all carry a contingency of symbols. These elements are put together which in turn can then create new meanings that are attached to the object as a whole. Akrich, M. (1992, as described in Bijker & Law, 1992)

describes how we have to go back and forward between the social and the technical elements of an object. When we look at an object, Zora in this case, from a technical point of view we see that Zora exist of already determined elements; screws, a plastic case, a speaker, a camera. If we look from a sociological perspective we can attach meanings to these elements, what does each element represent? For example a certain type of camera lens can be seen as expensive, which then can be seen as luxurious, as an item only available for the richest of society. Akrich (ibid.) thus says we have to shift between looking at an item as what is (as well as loose parts and as a whole) and at which symbols those lose parts and the object has a whole embodies. According to

(11)

Akrich (ibid.) we also have to look at how people can relate to the technology. Can elderly relate to robot Zora and how can we understand the role of their own background and their current life and disabilities in the use of Zora?

Among elderly dementia is a common illness. Botsis, T., &Hartvigsen (2008) explain why people with dementia or similar illnesses won’t benefit from telecare devices. The authors make a distinction between two categories of telecare: the use of telecommunication and the use of telemonitoring. Where telecommunication is about technological devices that allow the patient to virtually come in contact with a caregiver and where telemonitoring devices are seen as devices with which the patient can monitor his/her own health. The study shows how patients with cognitive diseases such as Alzheimer and Dementia do not benefit from both types of telecare devices as it is too difficult for those patients to learn and remember how to use the devices (Botsis, T., & Hartvigsen, 2008). Robot Zora cannot be subdivided into these categories. The robot is not made for communication with a caregiver, nor is it designed for monitoring the health of the resident. Robot Zora can nevertheless be used in the communication between caregiver and resident. Will elderly who experience Alzheimer or Dementia interact with robot Zora and how does this interaction work?

2.1.3 APPEARANCE; GOOD LOOKS ARE NOT EVERYTHING

Humanoid robots are made in a way that they have to look and act similar to humans.

Nevertheless Zora has no soft skin or human skin colour, does not have hair, has no nose that comes out of its face and doesn’t have lips similar to those of humans. The reason for this is that the company that created NAO (the hardware of Zora) kept in mind the ‘Uncanny Valley’ theory. When making this humanoid robot, they made sure the robot still looks very different from a human.

According to the Uncanny Valley theory creators of robots should make the appearance of the robot not too much as that of a human being, as this can affect how people perceive the

robot and thus if and how they interact with the robot

(Mori, M., 1970, as described in

Walters et al., 2008)

. Mori explains that when robots look a lot like real people, but differ in subtleties, that this brings the biggest form of resistance against a robot. According to this research, by showing people different types of robots, the more a robot looks and acts like a human, the more resistance there is towards the robot. This is an example of an ideology the makers of NAO (hardware Zora) followed.

(12)

(1998) writes about developing a robot by looking at how children learn to interact with adults. The idea is that the robot ‘learns’ these things as if it were a child. Most social robots are now able to show certain basic emotions, like happiness and sadness. We can thus not look at robots as machines with completely numb faces and objective responses. We have to understand that the robots are in a way not merely robotic, because they are created by humans. Humans who will always have certain ideologies and a certain social & cultural background which influences their creations.

The ideas with which a robot is created could in turn also influence how the patients perceive the robot, or at least their perception of what the underlying ideas are. A person with a dark skin colour might perceive robot Zora, which has a white skin colour, differently than a person with a light skin colour. It would be interesting to see if we can find out how cultural perceptions of the self play a role in the way elderly look at robot Zora.

Carpenter et al. (2009) state that the gender of a robot creates certain expectations that are connected to that gender. Those associations differ per society. The robot can thus be a reflection of dominant social culture and reflect the inequalities of a society. It can therefore also be seen as partly human because it is initially driven by the same forces which we as humans are driven by. Besides the appearance of the robot, there also seems to be a difference in gender of the user in connection to the way a robot is perceived. Kuo et al. (2009) found in their research that male participants reacted more positive than female respondents when interacting with a robot. The exact reason of this difference is unclear. Similar to Akrich’s (1992, as described in Bijker & Law, 1992)notions on the reciprocal relation between the technical and the social, when doing research on this topic, we have to be aware of the appearance of the robot, the appearance of the user and the social assumptions and notions which are attached to the appearance of robot Zora.

2.1.4 THE FUTURE: NEW ROBOTS

In this short chapter I want to describe why some people fear robots. This fear can be another reason why people choose not to work with robot Zora.

The industry of robotics is growing. New robots are developed and improved. In scientific and non-scientific papers we can read about this fear. There are many authors (Nolfi&Floreano; 2001, Eiben, Bredeche&Haasdijk; 2014) whom argue that new evolutionary robotics are just around the corner. The evolutionary robots are installed with certain

algorithms, which enables them to have the freedom to do what they want and to replicate themselves. The human than is no longer needed as the robots can take care of themselves. This could also mean we can create an army of nurses, who can replicate and fix each other. No one

(13)

leaves, dies, gets sick, and if they get broken than the will be fixed by a co-robot. There will never be a shortage of nursing staff again.

In newspapers authors talk about jobs that will be performed by robots. The fear exists that robots with artificial intelligence, endanger our human race. The robots might take over jobs and our entire world. In most cases the robots people refer to, are robots which contain the function of artificial intelligence. Artificial intelligence means that when a certain algorithm is implemented in a technological object, the technological object itself is able to learn and able to use the new knowledge. For instance a robot arm exists which learnt how to reproduce itself. The fear for robots can be found in different national and international news articles;

“In the Netherlands people will lose their jobs to robots. If you want to do something about this then you will need to make a drastic change in our education. So that youngsters will learn skills that can be used in the future. Minister Asscher says we have to change course” [translated] (NOS, 2014)

“Robots may be autonomous, but they cannot be morally responsible as humans must be. The ambition to control them is as profoundly human as it is right ” (The Guardian, 2015)

June 22th (2015) in Japan robot Pepper sold out within a minute. Pepper is a robot who works through a collective artificial intelligence (Nu.nl, 2015). Although there are a lot of positive sounds these are almost drowned by comments about the possible negative consequences of using such a robot;

(3); “Am I the only one who thinks this is a scary development?” (Nujij.nl, 2015)

(5); “Creepy, especially since they are in permanent connection with each other through a cloud. Who knows what they will share...” (Nujij.nl, 2015)

(22); “If this grows big in the future and somebody hacks it, it is possible because it is digital and then one has the power over a whole robot army.” (Nujij.nl, 2015)

(50); “Collective learning robots? Empty the cutlery drawer and the kitchen knives under lock and key.” (Nujij.nl, 2015)

This fear for robots taking over our world is also explored in the fictive television show ‘Real Humans’ about humanoid robots (called Hubots). These robots are bought and used by people to perform all different sort of tasks, like doing the housekeeping, to watch the children, to be a romantic partner end so on. A small group of hubots find that they have a free will. They run away from their human owners and try to find a way to live without being controlled by humans (Theatrical, 2012).

(14)

2.2 THE USER(S)

In this second part of the theoretical framework I will pay attention to the user of technology, in this case the elderly. Here I will explore the difference between interacting with a robot and with a human. Adjoining I will look at how these differences could influence the user’s perception of robot Zora.

2.2.1 THE ‘SUPPOSED’USER

Akrich (1992, as described in Bijker & Law, 1992)describes how there is a ‘supposed user’ and a ‘ real user’ of technology. Where the supposed user is the person who will use the technology according to the manufacturers idea of how the item should be used. The real user is the person who, when the product is finished and delivered, is using the technology. The supposed user is in such a way often different from the real user. For example robot Zora is created for revalidation purposes, the supposed user will then be assumed to be a physiotherapist using the robot for revalidation exercises. The real user however can be a volunteer that uses the robot as a

microphone to perform stand-up comedy for the residents of a nursing home. The real user then changed the purpose of the robot. Akrich (ibid.) also refers to this initial plan as ‘The Script’. When the script has been changed, the symbols attached to the technological object may then also have changed. How is Zora used by staff and residents. Did the script of the designers change and how does this influence the interaction?

As implied in the above question the user of robot Zora is not only the staff who controls the robot, but also the resident who interacts with it. Following Akrich’s (1992, as described in Bijker & Law, 1992) theory on the script, we can argue that there are two scripts written by the makers of robot Zora; one for the residents and one for the staff (of a nursing home). This dichotomy thus means that how Zora is used differs for the staff members and residents involved. Even when both are in the same place at the same time, the robot is used in two different ways, while the robot itself is performing only one action. For instance, when everything goes according to Zora’s script; and the staff makes Zora play bingo then the staff uses the robot to provide the residents with an afternoon activity. While for the residents when Zora plays bingo, they are following the numbers and enjoying the jokes Zora makes in-between, they are entertained by the robot. The staff can thus be seen as the choirmaster, while the residents are the choir. How the residents respond to robot Zora is thus also influenced not only by the creators of Zora, but also by the people who control Zora.

(15)

2.2.2 DEMENTIA

Most residents in Sint-Anna WZC experience a mild to severe form of dementia. Here I will concisely appoint the striking features of dementia for those not familiar with the disease category. I will give a short explanation about what dementia entails and how people are affected by this mental illness.

Dementia is an umbrella term used for diseases whereby the nerve cells or the connections between the nerve cells are destroyed (Alzheimer Nederland, 2015). Dementia is in most cases connected to old age. The disease is most common in people who are older than 90 years. 40% of elderly in The Netherlands, who are older than 90 years old, experience dementia. The prognoses of Alzheimer Nederland [translation: Alzheimer Foundation Netherlands] is that the number of people with dementia will exponentially grow in the next 25 years due to an aging population and the knowledge that those people will reach an older age than the current generation of elderly(ibid.).

Due to the breakdown of nerve cells or connections, a person with dementia loses grip on reality and will progressively need more and more assistance to live his or her life. The time in which the nerve cells are destroyed differs per person, it is therefore difficult to say how fast someone will need daily care (ibid.). In the Sint-Anna WZC a scale is used to measure how severe the dementia of a person is, which then determines if this person can be placed on this list for admittance and if this person has a priority compared to others.

There are different symptoms of dementia which can be visible in the behavior of people. The symptoms can be; restlessness, aggression, changing moods, impulsive behavior and in the case of a particular type of dementia also hallucinations can occur. Dementia could thus play a role in how the residents respond to robot Zora.

(16)

2.2.3 IS HUMAN-HUMAN INTERACTION THAT DIFFERENT FROM HUMAN-ROBOT INTERACTION?

In this chapter I want to explore the main differences between human-human and human-robot interaction. This is important because it tells us how robots and thus also Zora might give different care then a human does, hence it also continuous upon the previous topic whether robots, in this Zora, can be a threat to the warmth of elderly care.

Many studies about human-robot interaction are focused on how the interaction works while neglecting the underlying social and cultural notions (Dautenhahn, 2006). The researchers and the makers of robots strive for improving the human-robot interaction. However, by

doing so they do not reflect upon the complexity of human-human interaction, but are focused more on the interaction part of the robot towards the human. This means that most robots today, cannot grasp the subtlety of human interactions (ibid.). How are these human subtleties in our interactions important for the interaction?

That the subtleties in human-human interactions are important is something that Goffman (1955) shows in his work about the daily life of people. He wrote about how we interact and how we make these interactions work. Not only we, but also the other(s) with whom we are interacting, ensure that this interaction works. Goffman shows how people who are interacting help each other with the interaction. He speaks of ‘saving face’ as a situation where something happens which can cause an interruption in the interaction. If this happens in an interaction the other person(s) will help you by reducing the ‘loss of face’ (ibid.). For example when you are speaking with someone you recently met and you accidentally call him/her by the wrong name, the other person tells you this and your face turns red. The person can help you by saying; “I too find it difficult to remember all the names, I had the same problem this morning”. This will make you feel less embarrassed and enables you to continue the conversation. These are examples of the subtle gestures we can find when looking at human-human interaction (Goffman, 1955).

These subtle gestures might be difficult to program into a robot. There seems to be certain feelings of empathy behind human gestures, we make excuses when we think we might harm someone’s feelings. The gestures we make are placed in a certain context. If you would not show signs of being ashamed by using the wrong name, then the other person probably will not have said anything about it because then there is no saving necessary. This again brings us back to the debate about whether technology, and robots in particular can dehumanize care. When robots will be used instead of people, this could dehumanize care because current robots cannot perform the interaction in such a subtle way. Can the elderly connect with robot Zora in a way where they can attach emotions to that interaction, or is the interaction mechanical and cold? Will the elderly see Zora as a living thing; a friend, or will they see Zora as an object; as a tool?

(17)

3) METHODS

3.1 OVERVIEW FIELDWORK

Below I present a quick overview of the performed fieldwork. With observations as my main research tool, followed up by interviews and informal talks with the residents. The interviews and informal talks with the staff are

Observations Kiné (90 minutes) Observations Animation (90 minutes)

Closed Floor: 1) 01/04/2015 2) 08/04/2015 3) 21/04/2015 4) 22/04/2015 Open Floor(s); 1) 01/04/2015 2) 08/01/2015 3) 21/04/2015 4) 22/04/2015 Closed Floor; 1) 21/04/2015 2) 26/05/2015 Open Floor(s); 1) 26/05/2015 2) 27/05/2015 Cafeteria*; 1) 22/04/2015 * 45 minutes

Interviews & informal talks Staff1 _{Zora talking with residents; recorded & Informal}2

Interview (45 minutes); 1) Managing Director 2) Anna Informal conversations 1) Eline 2) Fabian 3) Anna 3) Marcel 1) Sophia 2) Jerome 3) Stephaan 4) Achiel 5) Irma 6) Dina 7) Hortense 8) Julienne 9) Elodie

(18)

3.2 FIELD SITE: LOCATION AND CONTEXT

This research is conducted in a social care home (nursing home) in Ichtegem in Belgium called Sint-Anna Woon- en Zorg Centrum (Sint-Anna Living and Care Centrum).

This location was suggested to me by Tommy Denblieck after contacting him about wanting to do a study about robot Zora.

Sint-Anna WZC offers social housing and complete care apartments. Elderly who can still maintain a household but need a bit of help are living in the social housing. Elderly who need constant care are living in the complete care apartments. Sint-Anna WZC aims to provide these elderly a feeling of individual warm care.

My research is conducted in the common areas of the complete care apartments and in the complete care apartments of a few residents, always under supervision of staff. The people who live in the complete care apartments are divided among three floors. Each floor has a common area where there are tables with chairs, lounge chairs, television, radio and a kitchen. All group activities take place in the common areas.

The people who live in Sint-Anna WZC in the complete care apartments are in need of daily care, which means that most have a starting or severe mental or physical health problem. Dementia is most common among the residents. The severity of dementia among residents differs per floor. The ground floor houses people with severe dementia, this floor is locked with an electronic alarm to prevent that residents walk away. The other two floors are inhabited by elderly with less severe dementia and are not locked.

All observations took place in the common areas, the observations with staff took place in the cafeteria on the ground floor and the talks with residents took place in the comfort of the residents’ home (under supervision of a staff member). I will provide a thick description of the location and its residents in my analysis.

(19)

3.3 METHODS &ETHICS

3.3.1 METHODS

With this study I want to get a better understanding of how the interactions with robot Zora work, and how this is influenced by affective feelings toward the robot. For this reason my research was conducted with a qualitative approach; by doing an ethnography.

“An ethnography is written representation of a culture (or selected aspects of a culture).”(Clifford, J. , 1969, p).

This ethnography includes observations, informal interviews and in-depth interviews.

Observing is my main research mechanism as this enables me to include everyone who interacts with robot Zora. By doing observations I can see what the actual behaviour is towards Zora. In my observations I for instance looked at the facial expression of people when Zora tries to interact with them and how the elderly non-verbally responded to the robot.

I started by doing non-participating observations in the nursing home. By doing so I wanted to get a better understanding of the context of the field. This was necessary to

understand what is considered ‘normal’ in the nursing home. How often do people interact with Zora, where is Zora placed and how is interaction started? Starting with non-participating observations is also be helpful for the elderly as they can slowly adapt to my presence. Fast it became clear that the participations were immediately turning from non-participant to

participant observations, as I was not invisible and residents as well as the staff included me in the classes. During the observations I took notes of the atmosphere, interactions, the subtleties in interactions and things that stood out.

Secondly, I had informal contact with the elderly, guided by staff and with Zora present and/or by talking through Zora. During these interactions I conducted short informal interviews among the elderly. These interviews were recorded with a tape recorder, the tape recorder was placed next to me, out of sight so that it did not hinder the conversation. The residents were informed in advance about the recording. The recordings were deleted after transcription. With these informal talks I got a feel of how the elderly individually interact with Zora and what they think about Zora.

Thirdly, I conducted a few semi-structured in-depth interviews (along with informal talks) among the staff and direction. By doing these recorded interviews and non-recorded talks, I wanted to learn how the staff thinks about the interaction, but also how they feel about Zora and how they interact with Zora. What is the interpretation of the staff when looking at the interactions of robot Zora with the elderly and what do they of the robot? The in-depth

(20)

Because my research focuses on how the elderly interact with Zora, I made the conscious choice to start by talking with the elderly. The interviews with the staff were conducted after these talks. This to prevent that my interviews with the residents are clouded by the stories of the staff. Each interview was recorded and transcribed. During and after each interview I wanted to take notes. This to include the non-verbal communication showed during the interview and also because I wanted to record my feelings about the interview and the person interviewed. This appeared to be too difficult during the interviews, as I believe an interview is a conversation between two people, wherein both subjects should talk as freely as possible. Making notes during an interview resulted in me not being able to have a fluent conversation, therefore I choose to focus merely on the conversation and make notes over the overall conversation afterwards.

During and after my fieldwork I transcribed all interviews and digitalized all field notes. After transcription the recordings were deleted. With the help of the analyzing program Atlas.ti. I did this by following the grounded theory approach (

Glaser & Strauss, 2009).

I coded, re-coded and grouped my data to get a clear picture of what I have seen in the past months.

The ethnographic nature of this study means that this work is a sketch of my interpretations on the field I studied. In my field notes I tried to stay as close to the facts as possible. Finding a balance between storytelling and factual data. However my own field notes and my interpretation of those field notes are of course still coloured by my own perspectives. This is something that cannot be prevented, but of which the writer and reader(s) should be made aware of.

3.3.2 ETHICS

All participants were asked for informal consent, where I explained who I am, why I conducted this study (broadly) and how the information I gathered would be used.

Most residents of Sint-Anna WZC were however not able to fully understand my

intentions and can therefore not give consent to participating. Sint-Anna WZC kindly offered me the opportunity to observe and speak with people, under the constriction that I would ensure the privacy of the residents. It would be highly unethical if I would use any personal information of the residents. To make sure that the residents are not harmed by this study, I will therefore

not use the real namesand/or other personal information of the residents, instead I will ascribe

fake names to each resident and exclude all information that can be traced back to the resident. Regarding the staff of Sint-Anna, everyone was individually asked if their name could be used and published. After transcription the recorded interviews were deleted immediately. The staff have been offered the choice of choosing a fake name. They all agreed upon the use of their

(21)

first name. Even though at this moment I cannot think of reasons why this study would harm them, this does not mean this cannot be the case in the future. To be absolutely sure that I will not harm anyone I therefore decided to change their names.

During this study I was aware that speaking about life in general might bring up past

memories and emotions. I did my best to prevent this without taking away the agency of a person. The staff of Sint-Anna WZC was very helpful. They have introduced me to the residents and provided me with information about topics I could better avoid.

3.3.3 LIMITATIONS

Although I conducted my research in a neighboring country, in an area where people speak the Dutch language, it was often still difficult to understand the dialect that was spoken by the residents. By doing observations most of this difficulty was overcome. In other cases the staff of Sint-Anna WZC helped me by translating.

The second difficulty I stumbled upon had a technological nature; the WiFi-network . Due to problems with the WiFi it was not possible for me to do observations or any other activity with robot Zora. This caused a delay of a few weeks in the time I calculated for my fieldwork .

(22)

4. LIVING IN SINT-ANNA WZC

4.1. LOCATION

Sint-Anna WZC is located in Ichtegem, Belgium. It is hidden in a street with houses that are connected together. To find the entrance I had to drive with my car through a short alley in-between two rows of brick houses. When on the other side of the alley a wide space showed, with a large L-shaped residential building in front of me. On the right side of the building there are parking spots, on the left side of the building there is a terrace with modern green and black plastic tables and chairs. The terrace is placed in front of the cafeteria of Sint-Anna WZC. Below there is a field note to visualize the location.

This is the office wherein the Kiné and the Ergo therapy staff is situated. It’s a small office with three treadmills, a few exercise bikes and bars for walking exercises among some other training equipments.

3

Fysiotherapy is called Kiné in Belgium, I will therefore sometimes refer to Kiné or use quotes wherein someone speaks of ‘Kiné’

Field note: “The first couple of day’s Kinesist3_{Fabian takes me along when he is giving}

gymnastic classes to two groups. One group is living on a closed of floor in the nursing home where people are living whom experience severe mental health problems (severe dementia f.e.).

There are light grey stone tiles on the floor. When we walk through the doors of this department we first walk through a corridor with doors left and right. Some doors are open and I can see that these are the patient’s rooms. The patients rooms are approximately 12m2 _{and offer a one persons bed, a closet and a table. At the end of}

the corridor we find the common area. In the common area on the left is one big square table, on the right side of the room there are more square-shaped tables. There are a few people in wheelchairs who are sitting at the tables. Some with a drink in front of them. In the back of the room, on the left side, there is a small living area with big wooden chairs with red leather seating’s. The wall on the left side and the adjoining wall on the right have many high windows through which the sun enlightens the room, this makes the common area a light space, even when the clouds appear there is enough daylight in the room. On the sides of the windows there are colour-blocked curtains (orange, red, yellow). They remind me of the curtains they often have in bungalow parks.”

(23)

Residents come to the office to exercise under the guidance of the staff. In the back of the room there are two desks placed against the wall, with two computers on them. Robot Zora is stored in the back of the office, in the right corner of the room, on or under a chair. There the robot is charged until it will be used again.

4.2 MEETING THE RESIDENTS

In both groups I was introduced, with my first name, as a student who is observing robot Zora for her thesis. In the closed department people shortly looked at me, I smiled at them and friendly nodded. One woman kept on looking at me and continued smiling at me. After a few weeks I tried to categorize the residents to give insight in what sort of people are living in Sint-Anna WZC. I found that all residents I have seen have a white skin-colour4_{, most residents are} between 70 and 90 years of age. Another thing I noticed is that most residents are dressed well, meaning that they are wearing clothes that are not pajamas, sloppy or worn down, but instead fit them nicely, clothes in which they could go outside without being stared at. The residents

smelled clean. Every time I met the residents I made sure to be dressed appropriately, according to my standards this means wearing non-revealing clothes. I am a female, 29 years of age, I have a white skin, red hair, I am short (164cm) with an average body type. The impact of my

characteristics on the residents is difficult to measure. Most residents are female, but on the

Field note: “After a few minutes we head to the second group on the second floor. We take the

elevator. This group of people are in better mental health and some of the people are only here for a short stay. When we arrive on the second floor. When we get of the elevator we turn right and walk into the common area. In front of us and on the right there are tables. In the back there is a jukebox and a cupboard with an old looking radio on it. On the left side of the room there are more white tables, there is a television hanging in the corner and two red chairs stand next to the television. There is a kitchen on the left side. There are window on two walls with the same colour-blocked curtains. There are some people sitting around the television, most in wheelchairs.

We have to make space on the right side of the room by moving the tables to the back and the side of the room. Creating a big enough space for people (in

wheelchairs) to sit in a circle. Fabian leaves to collect patients. In the mean time I set up a chair in the front of the circle.”

(24)

open floors this balance is less skewed. When introduced on the open floors some residents asked me questions; do you have children? (no) How many children do you have? (none) Where do you live? (Rotterdam). Following up on my answers they brought up memories of cities or people. This made me feel more present and less invisible than on the closed floor.

(25)

5. USING ZORA

5.1 USING ZORA AS A TOOL; ACTION-ACTION-REACTION;

Zora has been put to use in Sint-Anna WZC as a tool for revalidation and animation. Here both staff and residents can be seen as users. The staff uses the robot consciously as a tool to help with revalidation exercises or animation. The staff decides which exercise comes next or which game will be played.

The residents use Zora in a less direct manner. They are inclined to follow the lead of the staff. How the staff steers the robot then also has an influence on how the residents can use (or interact with) the robot. The action of the staff leads to an action of Zora, which leads to a reaction of the residents. In the communication between staff member and a resident an extra component is thus added. To see how this works I will look at both groups separately. In the next sections I will first explore if and how residents use Zora.

5.1.1 THE RESIDENTS

Closed Floor

The residents in Sint-Anna WZC have a small or limited choice in using Zora. On the closed floor most residents were brought to the class and were (kindly) put in a chair. Some of whom were tied to the chair with a waistband to prevent them from falling out or walking away. Due to health problems and health risks their agency thus has been reduced. Hence there has been no clear avoidance among these residents in using Zora.

The robot introduces itself at the beginning of each Kiné class. This is a standard application in the menu of robot Zora, when this option is selected the robot will say its name and provides some details about itself and for which purposes it can be used. After the

introduction Fabian always chooses to let Zora perform a song. On the next page I provide a lay-out of the common area on the closed floor. This lay-lay-out shows the setting of the first

(26)

Picture 1: drawing of common area of the closed Floor Zora: blue dot

Fabian: orange smiley face

Me: green smiley face

(27)

Due to the short attention span of the residents, which is tied to the severe dementia most suffer from, it is difficult to see whether people, individually, disliked working with Zora. Opposite to disliking the robot, it was transparent whom of the residents did like Zora. How those residents responded to Zora and which effect this had on them is shown in the next field notes.

Field note; “Dina looked at Zora and started to smile at the robot, while making cuddly noises like how some people do when seeing a baby.”

Another resident was looking at Zora while the robot was performing a song;

Field note; “During the song Sofia looks at Zora and while copying some of the arm movements of the robot, she continuously smiles at Zora.”

Sofia thus displays emotions towards the robot, she kept looking at the robot and was not put off when Zora continued its dance routine. The moves Zora makes with its arms and the eyes that flicker, seem to be interpreted by the residents as a reaction to them.

Residents who interact with the robot are visible, because the interaction stands out during the classes. Not every resident responded to Zora in a visible way and some did not respond at all. It is not to say if this is a lack of interest or other factors. All residents awake however did use Zora, not in the way of wanting to interact with Zora, but as a speaker which played music. The musical option of the robot makes Zora interesting for almost all residents as most like to listen to music. The songs programmed in this version of Zora are especially chosen for the target group; the elderly. In summary, most responses of residents to Zora are

non-Field note: “We walk into the room and some residents are already waiting for the class to begin.

There are two women who smile at me when we walk in. One of them is still sitting at the table and stands up to move to the circle of chairs. I ask if I can help her move and she takes both my hands. When she stands up from the chair the lady wants to dance with me. Slowly and while dancing we move towards the chairs. I tell that she can sit down here as the class almost begins. She tries to convince me to dance a little more, but I kindly ask her to sit down. She keeps smiling at me and I smile back. Fabian is almost done with the set-up for today’s class, everyone who wants to join and who is able to join is now sitting in the circle. The class starts with Zora

performing a song. This is the song ‘Eviva Espana’. One woman is looking and smiling at Zora while she moves her feet along with the music. Others are looking shortly at Zora and then at Fabian. Fabian is making the same arm movements (dance

movements) as Zora, but he enthusiastically does this a bit bolder, to persuade the residents to join. The attention span of the residents is short, some fall asleep during exercises or suddenly stop with an exercise. Fabian continuously calls out their names to enable them to reconnect with the class...

(28)

Open floor(s)

There is a difference between the group of residents on the closed floor and the residents on the open floors. In summary, the residents on the closed floor almost all responded to Zora during the Kiné classes. The residents on the open floors however seemed to care less if Zora was brought along during the Kiné classes;

There is no sign that this non-response occurs due to residents unwillingness to work with Zora. The above field note shows that the group of residents is larger than the group from the closed floor. The residents from the open floor experience less severe forms of dementia, thence their verbal skills are better. Where in the group from the closed floor there was little or no

interaction among the residents, in the group of the open floors the residents did interact with each other.

Besides being used in the Kiné classes Zora has also been put the use for the animation classes organized by the ergo therapy staff. There is a noticeable difference between how people respond to Zora before the animation classes compared to the Kiné classes.

Field note; “... We walk in to the common area of the second floor. I am carrying Zora. There are residents sitting at tables, drinking coffee or watching television. One woman enthusiastically says; “óóó Zoraaaa, is Zora here this afternoon, that’s nice!”. A few others also make recognizing sounds towards the robot...”

I have never observed this type of responses before the start of the Kiné classes. When stepping into the room for the animation activity, people responded in an enthusiastic manner on the

Field note: “Beforehand we connect Zora to the electricity to charge it. We start by moving the

tables and chairs. Fabian collects the residents for the Kiné class. When a couple of residents are sitting down they start chatting with each other. One man, Julienne, starts joking around, he pays no attention to Zora, nor do the others. They are laughing about a joke he makes and are focussed on who else joins the group. Others are greeted. Often a resident walks in with a stroller and is asked by another resident (who is already sitting in the circle) to sit next to him/her, everyone seems to know each other well. No one looks at Zora. We start when everybody is sitting down. There 17 residents joining the class. Zora does its introduction and then Fabian tells the residents which exercise will be showed by Zora. Zora shows how to perform the exercise, and Fabian does the same. During the exercise Fabian walks around and helps a few residents who have difficulties bending their arm in the right way. Zora continuous to perform the exercise for the group to follow. The residents look at Zora, however when Fabian is not looking they often stop and start chatting with their neighbour.

(29)

presence of Zora. I asked Anna, an ergotherapist who is working in Sint-Anna WZC, about the difference between the two groups of residents;

Anna: “ Yes upstairs [on the open floor] people are more critical. Yes well how can I tell

this, upstairs the people are a bit more on guard, because of the dementia people let go

of their boundaries. They are more spontaneous, yes I think so. Yes they will show their

feelings faster. Emotions come to the surface more than upstairs.

Anna explains how dementia has an influence on how the residents respond to Zora. This is

interesting as some argue that technology is something we have to understand in order to use it like Botsis, T., & Hartvigsen (2008) argue when speaking of people with dementia and

technology. In the case of technology like robot Zora, the understanding that is asked of the elderly is limited. Among the residents of the closed floor it is thus not true that they have to understand the technology in order to use it. People accept the presence of the robot and enjoy the songs the robot performs.

The controlling and steering of the robot is done by the staff and the functions are

specifically designed for elderly. On the one hand we see that people with more severe dementia seem to enjoy the music & dance functions of the robot more than the residents with less severe dementia. On the other hand the games where they have to guess and answer to Zora take up more time or are too difficult to complete.

Field note; “I walk with ergo-therapist Eline to the closed floor. With us we take Zora to play the game ‘Show me the animal’. This game is played by giving the residents a few cards with a drawing of an animal on it. Zora then asks to show a certain animal. When someone shows the correct animal, Zora will say ‘Well done’ or ‘Good job’ accompanied by a cheer and a tune. When the residents are sitting Eline starts the game. Zora says; show me the elephant. To decrease the level of difficulty Eline holds up three cards and says ‘can someone see the elephant’. No one responds. Eline than chooses a more individual approach by holding the three cards in front of a resident. With a questioning look on her face the residents stares at the cards. Eline helps her in picking the right card...

(30)

This game is too difficult for the residents of the closed floor. Only a few can recognize the animals on the card and only when the staff helps them by providing a context for.

How residents thus respond to a robot can be influenced by the functions of the robot and the way in which the staff provides help. When looking at robots in elderly care it will therefore also be necessary to look at the functions of a particular robot. In the case of Zora there is an action of the staff – then an action of Zora and then a reaction of the resident towards the robot. The intervention of a third party, in this case the staff, helps making the robots actions interpretable for the elderly.

5.1.2 THE STAFF

Robot Zora is used by the staff of Sint-Anna WZC for entertainment (ergo therapy) or activation (Kiné) purposes. I will first highlight how Zora is used by the Kiné staff and how this works. In the second part I will describe how the ergo therapist use Zora.

As mentioned before the Kiné staff uses Zora to help showing people the exercises, which enables the Kinesist to provide individual help when needed.

Using the robot sometimes takes effort. Robot Zora is dependent on the wireless internet connection (WiFi). When the WiFi doesn’t work then also robot Zora will not work. In Sint-Anna WZC there WiFi points were installed to ensure WiFi throughout the whole building. Despite this there were still problems with the WiFi signal. In crucial area’s the WiFi-signal was to weak for the robot to pick up on.

Field note: “...today there are problems with the WiFi network, In the office there is signal, but

when we step outside of the office the signal disappears. If it’s not possible for to connect Zora to the WiFi signal then it’s also not possible to realise a connection between the tablet and Zora, which means the robot cannot be used. We decide to just give it a try upstairs, where Fabian will organize the Kiné class, to see whether Zora can be used for the exercises. Fabian picks up Zora and walks with it into the hallway. He tries to find a connection standing in different places with robot Zora in one hand and the tablet in the other. The connection seems to be weak, only near the reception desks there is a stronger signal, but that’s in the hallway and it won’t be of any use for the class if Zora stays there. Fabian walks back into the common area. He turns the robot off and on again and reloads the interface of the robot on the tablet. After approximately 15 minutes Fabian is able to find a strong enough Wifi signal and is now able to use the tablet and thus to control Zora...”

(31)

Fabian thus tried patiently to ensure the connection. When the wireless connection is not strong enough it takes effort and patience from the staff member to find a solution or to continue without using Zora.

The ergo therapy staff uses Zora for animation activities, mainly with residents from the open floors. Zora is used as a tool for playing bingo and other games. When playing bingo without Zora, the ergo therapist is turning the wheel with the bingo balls and calls out the numbers. When playing bingo with Zora then the robot takes over these tasks;

As shown in the above field note, the bingo function of Zora is not working as it is supposed to do. The functions was meant to be improved by the use of a dialect which the residents

understand. The dialect indeed works, but because the robot cannot use its own voice anymore the volume is to weak. Problems with the technology the robot needs and the internal programs can make staff less willing to use robot Zora.

Field note: “ This afternoon we are going to play bingo with the residents. Sarah collects al

requisites and we then step into the elevator to the second floor. Sarah arranged a microphone because the sound coming from Zora is not loud enough to hear for the residents. Normally the sounds coming from Zora are loud enough, because the residents could not understand the bingo in the Dutch dialect, it was arranged that someone spoke in the numbers in their own West-Flemish dialect. Unfortunately it was not recorded loud enough. We set up the microphone and the speakers. Sarah installs the residents. The residents are sitting at four square tables that are shoved together. Every resident gets a plastic bingo card, above every number there is an orange slide that you can slide down when the number has been named. We start Zora and put the robot on the table. The microphone makes a high-pitched noise when we try to move it close to Zora. When Zora starts with the bingo there comes a horrible sound from the speakers. The sort of sound you hear when a microphone is not adjusted correctly. The next ten minutes we are busy removing all things that might interfere with the microphone, like phones and the tablet. We turn the buttons of the sound system but nothing seems to help. The noise is gone and then it

suddenly it is back again. After a while we figure how best to place the microphone. Now it is placed further away from Zora. Because the sound is not that loud anymore, Sarah repeats all the number that Zora says. Meanwhile the residents keep looking at their bingo cards to see whether they can cross a number off the card.

(32)

5.1.3 FRIEND OR TOOL?

There are different reasons why people choose not to use (or misuse) a telecare device; less contact with other humans; fear for social isolation or the fear of losing agency. Zora is implemented in Sint-Anna WZC as a tool for the Kiné and Ergo therapy staff. In social debates people are often afraid that robots will replace staff, and that this dehumanizes elderly care.

In Sint-Anna WZC I experienced no direct negative attitudes towards the robot. Most residents are willing to use Zora, or have accepted the presence of Zora. According to Anna there are a couple of residents however who make a conscious choice in avoiding Zora when possible; Anna: “…Yes yes, or they are afraid of it because they don’t recognize it. It is something that moves, talks, but they don’t know what it is, which can be threatening. We have a few residents who want nothing to do with it, if there is bingo with Zora then they won’t show up.”

The residents described above never changed their opinion about the robot, nonetheless they were present during bingo one afternoon, while knowing Zora was there. The choice of avoiding the robot can then also be something that differs per day.

With another resident I was warned that she did not like the robot and could get very angry at it;

Field note; Some residents were already sitting at the table waiting for dinner to begin. It is 16:20h and the staff start collecting people for the dinner, which starts at 17:00h. I make Zora walk around. The robot does not walk fast so it takes a while to get to a resident...

when I let Zora walk to another resident, who is sitting by the window, another woman looked at the robot. I steered the robot towards her, keeping some distance because the staff told me that she does not like the robot. Surprisingly she held out her hand for a handshake, I let Zora give her a handshake and made Zora say hello, and how are you today? She smiled and petted the robot on the head..

This residents was thus known for not liking the robot, and for occasionally having a strong negative reaction towards robot Zora. The day described above there was no sign of those feelings. As described in the chapter about dementia, how people respond can be influenced by the disease (Alzheimer Nederland, 2015). Hence it is difficult to make an interpretation of how in general, the elderly respond to Zora. Overall, considering all observations, they accept Zora and have a positive attitude towards the robot.

(33)

5.2 INDIVIDUAL USE; THE AFFECTIVE RELATIONSHIP

In one-on-one talks with the residents it was easier to see how the elderly respond to the robot. These talks were informal and in the setting of their own room. The following field note will demonstrate how the residents were approached;

We thus talked and asked questions through Zora. Zora stood nearby the resident. Some residents petted Zora on its head. When talking with the residents through Zora it was

noticeable that they all kept looking at Zora. During the talk it was explained to each residents that Zora only said what the person with the tablet typed in and that it only did what the person with the tabled wanted it to do. This made no difference in how each of the residents spoke with Zora. They kept waiting expectantly for Zora’s next answer or question while looking the robot in the eyes.

Letting Zora speak took a while. Typing on a tablet goes slow, especially when you are not used to doing this and when you are sitting on a chair with the tablet on your lap. With one hand you have to hold the tablet while typing with the other one.

Below one of the conversations the ergo therapist and I had with a resident. Here ergo therapist Sarah did the typing; ;

Field note: “ I stand with Sarah in the hallway to see if the WiFi-signal is sufficient here. It takes

a few minutes, but then Zora works and we put Zora on the ground before entering the room. Sarah let it stand. When Zora reached its standing position, she lets Zora walk into the room. Sarah asks the resident if it is okay if we have a small chat with him, through Zora. She then lets the robot do the talking. The residents start asking Zora questions, looking only at Zora.

Chatting with Zora:

Zora walks into the room. Jerome; who came with you?

Zora; We have come to bring you a visit Jerome; Yes, but who did you bring with you? Zora: Sarah and Els

Jerome; do they behave well? Els: she has to think about it Zora; yes very well

Jerome; and what are you up to then? Zora; I cannot talk fast yet.

(34)

Here we see that Jerome directs his questions to Zora. Even when Sarah joins the conversation, he still continues to ask his questions to Zora. He also keeps looking at Zora while doing so.

Zora: Yes

Jerome: Do you like to go along with them? Zora: Yes very much so

Jerome: yes I would also enjoy that [We all laugh]

Zora: Els is here for her study Sarah; Do you understand that? Jerome; yes she is here for her study Sarah; Yes for school

Els: to look at the robot

Jerome; to see what we do with it.. Zora: Yes, especially what you are up to. Jerome; Well that’s not that much [laughs] that was about 20 years ago.

Zora; don’t tell me lies! Jerome; What?

Els; Yes, she knows everything Zora: that you cannot tell lies. Jerome; yes

[long silence]

Zora; I know everything

Jerome; Óóó, you know everything? Zora; Yes

(35)

Sometimes there is a long silence in the conversation because the typing on the tablet takes some time. In this time Jerome kept looking at Zora, patiently waiting for the next question or answer. This was the same for the other residents who participated in one-on-one chats with Zora. It is interesting that even when the conversation does not go fluently, that Jerome is not bothered by this and is not neglecting the robot by turning to us with his questions. Although a long silence might feel unpleasant fur us, this doesn’t seem to be true for the residents

interacting with Zora. Robot Zora stays the main thing he is interacting with. Even after Sarah explains that she is actually asking the questions, the questions were still not addressed to her, or to us.

We see here that after Jerome has been told that Sarah types in the questions, he still continuous his interaction with the robot. He asks Zora whether the robot likes Sarah. While, when thinking that Sarah is steering Zora, he is actually asking Sarah if she likes herself. I don’t mean to suggest that Jerome doesn’t realize that someone is steering the robot, because I think he understood that very well. I want to make clear that he is thus playing along with the game we offered him and he seems to enjoy this. Halfway through the conversation we realise that we didn’t turn the

Sarah: It is difficult to communicate like this Jerome. I Type everything and then she says it, but I cannot type quick enough.

Jerome; Do you like her? [the question is directed to Zora] Zora: Who?

Jerome; If you like her? Zora Who?

Sarah; She says; who?

Jerome: uhhm.. [then points at Sarah] Sarah; what is my name again? [smiles] Jerome; Yes don’t kill me for it.

Zora; Sarah

Jerome; á yes, a nice girl hè Sarah Zora; Me as well.

Jerome; Yes of course

Jerome; and what are you up to then? Zora; to behave well

Jerome; yes you have to behave well Zora; are you joining us in bingo today? Jerome; yes

(36)

face recognition option. We then decide to still do this during the conversation. When this option is turned on, Zora will follow the face of the person the robot is talking to.

After we turned the face recognition on, the robot looked up at Jerome’s face. Zora then kept on looking at Jerome. Non-verbal communication was thus added to the conversation. This resulted in Jerome showing his glasses to Zora.

Jerome sees the robot as a female. Because the robot chats a lot, which can be ascribed to femininity. With the other residents we have kept the face recognition on. All residents encaged in this form of non-verbal communication, they smiled at Zora and looked the robot in the eyes when talking to it.

Zora; Do you like me Jerome; Yes, very nice

Jerome; and what do you think of my glasses [takes of glasses and shows them to Zora] Zora: very nice

Jerome; yes? Zora; nice colour

Jerome; I bought it for Sarah Zora: Sarah does not wear glasses [Sarah receives a call]

[silence]

[Jerome pets the robot on the head} Els; She can chat a lot

Jerome; Yes it is a woman so that’s why Els: ? [language barrier]

Sarah; Els is from the Netherlands, she does not understand everything we say.

Sarah; it is a woman he said Els: yes is it a woman? Jerome; Yes

Sarah; that’s why she is good in chatting Els: Aaah! Well that’s true [laugh] Jerome; You notice that right away

(37)

Jerome did not hesitate in picking up the robot. Zora weighs approximately 11 kg, which is quit heavy, especially for elderly with less muscle strength. Jerome however enjoyed the physical contact with Zora.

Sarah; are you joining the bingo? Zora is going to do the bingo. Jerome; yes

Sarah; okay see you there, we will leave now.

[ Sarah makes Zora perform a programmed goodbye option] Zora; you can pick me up now

[ Jerome picks up the robot, and sets the robot on his lap]

Zora; if you want me to stay than say; sit and wait. if you want to pass me along then say; pass along.

Sarah; “Pass along”

Jerome; will she also be there this afternoon? Sarah: Yes

Zora; Can I kiss you?

[ Sarah and I did not expect this sudden affection of the robot, we laugh] [Jerome laughs]

[ Zora makes the sound of giving a kiss]

[while we pick up Zora and start walking away the robot suddenly laughs, this laugh sounds a bit strange, a mix of naughty and mean is perhaps the best way to describe it]

Robot Zora : friend or foe? : an exploratory study about the emotional attachment of elderly to robot Zora