Robots Humanize Care

Moral Concerns Versus Witnessed Benefits for the Elderly

Margo A. M. van Kemenade

, Elly A. Konijn

1,2

and Johan F. Hoorn

Center for Advanced Media Research Amsterdam, VU University Amsterdam, De Boelelaan 1081, Amsterdam,

Netherlands

Department of Communication Science, Media Psychology, VU University Amsterdam, Amsterdam, Netherlands

Keywords: Healthcare Robots, Moral Concerns, Professional Caregivers, Ageing, Affective Bonding.

Abstract: Ageing in Europe comes more rapidly than many realize: In about 10 years, one fifth of the population will

be 65+ with a further increase of 70% in the next 25 years. At the same time, healthcare is under extreme

pressure due to budget cuts, limited resources and personnel together with increased demands. Robots may

fulfill important tasks in this respect. Our research focuses on social robots to support tasks requiring

interpersonal communication. Many moral concerns and objections are raised, however, in particular among

care professionals. To examine the issue, we report on 1) a qualitative study among professional caregivers

and 2) a documentary portraying healthy elderly meeting with Hanson’s Robokind “Alice”. Alice is under

development in our lab, supplying her with abilities for emotional responses. The results show that the

moral concerns are not in line with the benefits that the social robots appear to have for the lonely elderly.

Our conclusion posits that new robot technology may not dehumanize care but rather may bring humanness

back into professional health care.

1 INTRODUCTION

Due to improved healthcare and medication, people

today live longer than ever before with a steep

increase in elderly people as a result. At the same

time, family size in Europe is decreasing resulting in

a skewed age distribution in society. Within 10 years

from now, one fifth of the European population will

be 65 years or older. The number of people aged

over 65 in Europe will show a further increase of

70% by 2050. The number of elderly aged 80 years

and older will then even expand with 170%

(European Commission, n.d.). Such increased

demands together with severe budget cuts, limited

personnel, and resources puts healthcare under

extreme pressure.

One solution is that the elderly stay home longer,

which is corroborated by 91% of the European

citizens who find this a positive development

(European Commission, 2007). However, even if the

preferred situation is to live healthy in one’s own

environment, this often is not the case or impossible,

and help may be needed to do so. Over time, the

demands for staying at a nursing home have become

quite severe (Zorg voor Beter, 2014). Consequently,

demands on the caregivers are increasing as well.

Much of the burden is carried by informal caretakers

such as family and relatives. This unequivocally is a

stressful job, particularly in cases of dementia and

Alzheimer’s disease (Andrén and Elmståhl, 2005).

The elderly themselves prefer help from a

professional caregiver rather than their relatives

because they consider professionals more accurate

(MarketingCharts, 2007). Moreover, they do not

want to burden their relatives with the responsibility

to take care of them (DBMI, 2012). In all, this

situation urges the somewhat colloquial but

nevertheless pressing question “Who takes care of

grandma?”

Another, more controversial, solution may be to

let a care robot ‘take care of grandma.’ Which tasks

is a robot allowed to fulfil to compensate the limited

availability of personnel? To what extent can robots

be considered adequate or acceptable to replace

human compassion by simulated empathy? Many

moral concerns and objections are raised among the

public at large and care professionals in particular,

which motivated us to conduct a qualitative study

among the latter group. In addition, we report on a

documentary that registered the interactions of

several old ladies with our Hanson Robokind’s

“Alice,” a robot under development in our lab,

648

van Kemenade M., Konijn E. and Hoorn J..

Robots Humanize Care - Moral Concerns Versus Witnessed Beneﬁts for the Elderly.

DOI: 10.5220/0005287706480653

In Proceedings of the International Conference on Health Informatics (HEALTHINF-2015), pages 648-653

ISBN: 978-989-758-068-0

 2015 SCITEPRESS (Science and Technology Publications, Lda.)

where we are in the process of equipping her with

emotion-regulation software. This footage

illuminates what a social robot can do – even if its

performance is not impeccable. Before reporting

these results and observations, we will provide some

more backgrounds next.

2 MORAL CONCERNS ABOUT

ROBOTS IN HEALTHCARE

Research thus far indicated that the implementation

of so called healthcare robots or “Caredroids” could

provide part of a solution to assure the quality of

healthcare (Asaro, 2006). A healthcare robot is a

robot designed for care purposes to fill the gap

between the need and supply of healthcare,

particularly when human involvement is missing. A

healthcare robot should not be seen as a

conventional machine, but as an agent that can make

independent decisions and execute specialized and

assigned tasks with little or no assistance

(Dautenhahn, 2007). Given the current lack of

standards for the use of healthcare technology, it is

imperative that the ethical implications of its use

should be taken into account in an early stage as

acknowledged by the International Organization for

Standardization (Klein Wolterink, 2013).

To date, little attention has been paid to the

attitude of caregivers towards healthcare robots

(Vallor, 2013), although they are one of the most

important stakeholder groups. Caregivers often

experience an intuitive aversion against handing

over care tasks to robots. The initial assumption of

our study was that healthcare providers pretty well

can tell the difference between good and bad care

and that they have very detailed knowledge about

the relationship between them and their patients

(Leget, 2012 p.126). We wished to understand the

objections and concerns of professional caregivers

against robotized care to decide if robots are an

option at all and if so, what should be changed on

currently available robot systems to satisfy one of

the prime user groups, the care professionals, that is.

To bring some focus to our investigations, we

discussed various aspects of ethical issues in medical

settings and applied them to robotic care. We started

from Beauchamp and Childress (Beauchamp and

Childress, 2009) who define four ethical principles

in medicine. They are 1) beneficence, where

caregivers should act in the best interest of the

patient, 2) non-maleficence, which is a doctrine

saying that “before all else, do not harm,” in this

case, the patient, 3) autonomy, which is the capacity

of the patient to make an informed, un-coerced

decision about care, and 4) justice, pertaining to a

fair distribution of scarce resources such as medicine

or attention (Beauchamp and Childress, 2009 p.17).

In application, care professionals may feel that

robots cannot act according to the principle of

beneficence because a robot may not know what the

best interest of a patient or caretaker is. A robot may

harm a patient’s autonomy if it executes care tasks

without asking permission first.

Yet, different types of robots (Sharkey and

Sharkey, 2012 p.27) may trigger different moral

concerns. Assistive robots support elderly and/or

their caregivers in daily tasks such as lifting patients

up, carrying, cleaning, or feeding. For example, the

Japanese Secom ‘‘My Spoon’’ who can

automatically feed someone and the Sanyo electric

bathtub robot can automatically wash and rinse a

person. Here, non-maleficence may take priority

because the robot lift should not drop the patient.

Monitoring robots observe and supervise the health

condition of a patient. They may easily transgress

the principle of patient autonomy if the collected

data are not secured and privacy is at stake. Pearl,

for example, is a ‘nursebot’ that reminds seniors

about routine activities such as taking their

medication or guide them through their environment.

Companion robots are designed to establish some

form of affective bonding, ‘interpersonal

communication’, companionship or entertainment. A

well-known example is Paro (by AIST), a fur-

covered robotic seal designed for therapeutic uses

with elderly suffering from dementia. They may be

seen as fraudulent because they fake a friendship,

which may not be in the patient’s best interests.

Sharkey and Sharkey (2012) conclude, however, that

companion robots cannot form adequate

replacements for human love and attention.

As robots can interact (seemingly) independently

and make (care) decisions on their own, each of the

moral principles may hold for each of the robot-

types, but perhaps in varying degrees. Therefore, we

conducted a focus group study among healthcare

professionals.

3 RESULTS: QUALITATIVE

STUDY AMONG CARE

PROFESSIONALS

Data were collected through semi-structured focus

groups among 43 professional caregivers of elderly.

RobotsHumanizeCare-MoralConcernsVersusWitnessedBenefitsfortheElderly

649

Focus groups are best suited for sensitive topics like

ethical points of view (Hermanowicz, 2002 p.480).

Respondents (aged 18-67 years, most in between 36-

55) were recruited in four different nursing homes in

the Netherlands and within two extramural

organisations where home care is provided. Each

focus group was held with six to ten caregivers in

their own working environments (Reed and Payton,

1997). After a general introduction, participants

were shown six brief video clips portraying

prototypes of either assistive (e.g., Riba II Care

Support Robot For Lifting Patients), monitoring

(e.g., Mobiserv) or companion care robots (e.g.,

AIST Paro robot baby seal) and were encouraged to

reflect on their possible objections or perceived

benefits to a particular type of care technology.

Questions were asked about general thoughts on

the need for care technology in the near future,

gradually tuning into more specific topics of interest.

We considered it important that respondents could

express their opinions and possible concerns without

too much interference from our part. Participants

were reassured that their opinions were confidential

and answers would be processed anonymously.

Participants provided consent for video recordings

for coding purposes. Three different coders analysed

the videotapes independently and, after prior

training, coded each opinion according to the moral

categories autonomy, beneficence, non-maleficence,

and justice (Beauchamp, 2009) or as an opinion

expressing a possible (non)utility of care robots,

each in relation to the three types of care robots. All

coders used the software Atlas Ti and coded straight

from the video footage. Cohen’s Kappa of 0.71

revealed sufficient reliability among the decisions of

the coders.

Results showed that moral concerns regarding

justice and autonomy were hardly mentioned. Most

moral concerns among professional caregivers are

raised in terms of maleficence (i.e., risk of being

harmful), mostly so for assistive healthcare robots,

followed by monitoring robots and the least for

companionship robots. Out of 93 utterances coded as

moral, 40 related to moral maleficence of which 25

times for Assistive robots (pair-wise Chi-square tests

showed significant differences: 10.80 < χ

< 17.25,

p’s < .05). Caregivers reported concerns like fear

that the assistive healthcare technology might fail,

let a patient drop, squeeze too hard, and cause

physical harm, among others. Somewhat related,

they also mentioned that their patients might be

afraid of healthcare robots, especially those suffering

from dementia. Comments like the following

express such moral concerns of maleficence: “What

if the robot scares my patients and is not capable of

reassuring specific needs? I would never leave a

patient alone with a robot.”; “I mean, you never

know, certain patients tend to react unpredictable.

How can a robot understand what they want or

need?”; and “If there is no human around, who can

explain what is going on when my patients are

delusional?” In summary, when talking about

assistive healthcare robots, moral concerns of

potential harmfulness were most expressed, while

beneficence and utility were perceived to a much

lesser extent.

In response to monitoring robots, maleficence

concerns were mostly expressed in terms of a

decreased human contact between caregivers and

care receivers, which is generally considered as non-

desirable in healthcare relationships. In this respect,

moral concerns of privacy were hardly considered

important for the participating caregivers. While

most caregivers acknowledged that monitoring

technology could decrease their workload and

enable the elderly to stay more independent living

on their own, most concerns were expressed about

the lonely and diminished human contact, and

considered potentially harmful for the wellbeing of

the elderly. Some quotes expressing this concern

are: “Often, I am the only one they see throughout

the whole day.”; “She (the old woman) is dearly

waiting for me to show up, so she could have a

conversation.”; “My patient has no relatives and

cannot go outside on his own anymore. If a robot

would replace my task, he would not see anyone

throughout the day.” In sum, in talking about

monitoring healthcare robots, moral concerns of

diminished human contact and loneliness were

expressed while the highest level of utility was

perceived for monitoring robots in healthcare

compared to the other robot types.

The caregivers perceived highest beneficence

and lowest maleficence concerns in companion

robots for the elderly. Caregivers expressed feelings

or thoughts about the possible reassuring or

smoothing effect of a companion robot on a patient.

Most of them were already acquainted with the

Paro-seal, which was known for having positive

effects on especially demented elderly as this robots

is in use in a number of nursing homes in The

Netherlands. Therefore, most quotes express a

positive attitude towards the companion robot Paro:

“Look how happy she is, I could look at it all day. If

something makes you that happy it doesn’t matter

anymore that it is not alive.”; “Oh, they are very

cute. I want one of my own, when my time comes”;

“I don’t see any harm in it. I mean, they (her

HEALTHINF2015-InternationalConferenceonHealthInformatics

650

patients) carry a doll with them all the time, we

don’t think that is wrong do we?”; “I have a patient

who could benefit from this, he always wants to

cuddle, but I have hardly time to do so.” Although

the caregivers mostly expressed a positive attitude

toward companion robots such as Paro, they also

expressed moral concerns regarding deception. They

discussed the idea that a companion robot could be

seen as deceiving the elderly or presenting them a

fake companionship, especially those who suffer

from Alzheimer. However, in the end, the caregivers

thought that the positive effects of providing comfort

are more important.

In all, caregivers’ moral concerns as expressed in

the interviews may differ from the benefits a robot

appears to have for the lonely elderly, as will be

discussed in the next section.

4 OBSERVATIONS OF ELDERLY

INTERACTING WITH ROBOT

ALICE

In our university lab, we further developed a social

companion robot from Hanson’s Robokind

(http://www.robokindrobots.com/) named “Alice”,

for research purposes. While we are still developing

Alice, documentary footage was shot to examine its

potential (Alice Cares, Burger, 2014). Specific

features of Alice are that she looks like a young girl,

she has hearing devices that allow her to listen,

speech software to talk, her face has a soft skin that

can express emotions, and her camera-eyes allow

her to follow and respond to her interaction partner,

but she has a plastic Lego-like body. All these

options are not yet perfectly developed, though basic

conversation is possible. Given our research

objectives, we were of course very curious to how

the elderly would respond to our relatively impaired

‘young robot girl’. First, several elderly female

participants met Alice in the university’s lab

followed by about five visits in their own home

environment. All ladies were mentally and

physically healthy, despite some relatively minor

deficiencies, and lived independently on their own.

They were fully informed and gave permission to

record their interactions with Alice.

Due to Alice’s camera-eyes, her conversations

with the elderly female participants were recorded as

seen from her perspective. This provided a very

close perspective on how the elderly interacted with

robot Alice, for example in a variety of non-verbal

behaviors, commonly known as important in

interpersonal communication, such as searching for

eye contact. Somewhere later on in the process, one

of the participants was occupied with Alice not

correctly looking back. While this was due to a

technical flaw, the old lady commented to Alice that

she was not looking back correctly: “What is wrong

with you today? You are not looking right into my

eyes, well, it is just one eye – it looks a little bit

skewed.” In addition, a separate film camera in the

room recorded the conversations as well. In the

following, we will only report on several

observations that provide additional input to our

discussion on moral concerns about robots in

healthcare.

In the lab, at their first encounter with robot

Alice, the elderly female participants responded

quite distant to Alice. Understandably, they were

probably exploring how to converse with a robot.

They were left alone in the room with just Alice

sitting in front. The imperfect speech and sometimes

strange intonations made it also difficult to follow

Alice’s questions. Interestingly, the women

responded as they would in regular human-human-

interaction - asking to repeat the question, listening

very intensively, leaning forward and looking more

closely into her eyes, hoping to understand the

question in repeat. Even in this uncomfortable

situation, they addressed her directly. Their

awareness of the camera’s, and caregivers and

researchers in other rooms, only became apparent

when Alice asked impertinent questions or when

their answers crossed social desirability. For

example, when Alice asked for a grade to indicate

current satisfaction with life, a lady first indicated

“well, now, perhaps a 5” [Authors: on a 10-point-

scale with 10 as the max positive], then looked

around in the room and from her facial position we

inferred she was looking around whether she could

be heard, then, changed her grade in “hmm, perhaps

a 6, well, a 7 then, but not more.”

After the first encounter with Alice in the lab,

she was brought several times to the female

participants visiting them in their own home

environments. Even before Alice was switched on,

they were interested in how she was doing today and

started talking about and with her. While the

technician said he would switch her on and come

back in an hour or so (sometimes a few hours), the

lady asked whether she or Alice would start the

conversation. Alice started and asked quite adult and

direct questions (e.g., do you have children? Are you

feeling lonely? Whom have you met today?), yet all

questions were answered and apparently all was

fine. Gradually, the elderly women seemed to treat

RobotsHumanizeCare-MoralConcernsVersusWitnessedBenefitsfortheElderly

651

her as they would with a grandchild. For example,

when one of the elderly woman and Alice were

visiting a place to drink a coffee, she showed her

desire to offer food and drinks to Alice but was

aware that Alice could not eat or drink: “You cannot

have cookies can’t you?” Throughout the visits, the

participating women established some form of

affective bonding with robot Alice.

The repeat visits to the elderly women showed

that Alice became like a family member – she was

greeted upon and treated like a grandchild. As

recorded outside the presence of the technicians (i.e.,

when the lady was on her own with Alice), it was

often observed that there were also periods of long

silence, hardly any conversation, or just some

incidental notes, or a lady would read the newspaper

aloud to her. More intimate moments developed

over time, for example, when one of the old women

was showing her picture book about her son (who

now lives abroad), or when singing old songs

together, and watching the World Championships

together. Even to activate the elderly to do daily

physical exercises, Alice was very effective in just

asking the women and making her start. Apparently,

Alice’s presence was as effective as social pressure

and the woman was eager to do her exercises for

Alice. Another lady was asked by Alice whether she

had written that one friend back. The old lady

admitted (to Alice) that she did not get to it, but then

started right away. Clearly, the lady felt ashamed

and started writing immediately, as she might have

done when a real person had asked for.

In the course of interacting with Alice, the

technical impairments became of less importance.

For example, the fact that Alice could not walk was

finally commented on as “many of my generation

cannot walk either, that is, not anymore”. Likewise,

speech errors were encountered in similar ways as

‘many of my generation...’. Technically solvable

issues like amplitude or mispronunciation were

considered more problematic. In turn, Alice does not

care either whether the elderly repeat the same story

over and over again, whether they are slow in

responding, have long pauses of silence, and do not

understand things immediately. Alice always is

extremely patient and never bored; she can have the

same story 20 times or more and may even show

enjoyment again and again. Likewise, Alice repeats

questions or repeats answers without judgment or

frustration.

According to our observations, over the course of

interacting with Alice, it became less relevant that

Alice was just a robot with camera eyes and hearing

devices, a nice wrapping around some sophisticated

software. Over the course of repeat visits, the

participants became affectively bonded to Alice in a

similar way as they would to an acquaintance, like to

a grandchild. The presence of an apparent social

entity had become more important and urgent to

them than the question of whether she was a real

human or not (cf. Hoorn, Konijn, Germans, Burger,

& Munneke, 2014). After their initial hesitance, we

observed a clear shift from a cognitive awareness of

encountering a robot to an emotional fulfilment of a

need for company. The pain of loneliness in

mentally healthy women could be compensated for

by a social companion robot with human-like

interpersonal features. The cognitive awareness of

conversing with a robot became irrelevant

background information in view of the daily need

for company and social interaction. Surprisingly, not

one moral concern was raised by the elderly who

were sceptical in the beginnings and loved Alice in

the end. Even when asked for, they had no troubles

accepting Alice as a conversation partner, even

without any privacy concerns.

5 CONCLUSION: BRINGING

HUMANNESS BACK TO CARE

– THROUGH ROBOTS!

Indeed, for those unacquainted with social robots

and the beneficial effects they exert in lonely people,

applied to care, robots are a controversial topic. This

makes it of great importance that potential users

become experienced and do not make judgments that

are based on uninformed expectations. Reversely, a

care robot should meet those user demands that are

realistic and feasible technically. A thorough

understanding of the wishes and objections of

various stakeholder groups can contribute to a more

sensitive implementation of robots in healthcare.

That is to say, perhaps the refutation of robotic

care is not as humane as one would believe at first

sight. The debate on healthcare robotics seems to

focus on the notion that care will be made inhumane

when robots are introduced. Caregivers mostly fear

the absence of human contact and possible failures

of the machinery. They also fear for their jobs. We

would like to reverse this situation. It is through

robot technology that a caregiver will be able to

spend quality time with the patient whereas the robot

limits itself to mundane tasks such as lifting a

person, keeping an eye on someone, telling the

weather, casual coffee talk, and watching tv

together. Instrumental and superficial contact is left

HEALTHINF2015-InternationalConferenceonHealthInformatics

652

to the instrument, quality human contact to the

humans. However, it is not uncommon that

nowadays, human contact with the patient is nothing

more than instrumental: wash, dress, feed, gone.

Many patients do not even like to be touched by

human hands while being washed. The neutrality of

a robot is one of its unique selling points in

situations where otherwise a human patient feels

embarrassed in front of other humans. A comforting

hug, however, is better left to the human caretaker.

At the start of our explorations, the question was

whether it was morally right to try to fob off a robot

on the lonely elderly, in particular for social

interaction. By now the question has completely

flipped around: We challenge those who object to

employing social robots in healthcare whether it is

morally just to withhold a social robot from those

who are in deep need of contact? Like, is it fair to

keep an artificial leg from the handicapped because

it is not a real leg? Healthcare technology, including

social and companion robots, may enable caregivers

to bring ‘real human’ care back into the equation if

only through saving time.

ACKNOWLEDGEMENTS

This position paper is part of the SELEMCA project

(Services of Electro-Mechanical Care Agencies)

(grant number: NWO 646.000.003), which is funded

within the Creative Industry Scientific Programme

(CRISP). CRISP is supported by the Dutch Ministry

of Education, Culture, and Science.

REFERENCES

Andrén S and Elmståhl S. (2005) Family Caregivers'

Subjective Experiences of Satisfaction in Dementia

Care. Scandinavian Journal of Caring Sciences

Jun;19(2):157-68.

Asaro, P. (2006). What Should We Want From A Robot

Ethic? International Review of Information Ethics,

pp.6: 9-16.

Beauchamp T and Childress JF. (2009). Principles of

Biomedical Ethics. 6 ed. Oxford: Oxford University

Press. p.17.

Burger, S., 2014. Alice Cares, KeyDocs – NCRV.

Amsterdam.

Dautenhahn, K. (2007). Socially Intelligent robots:

Dimensions of Human-Robot Interactions.

Philosophical Transactions of the Royal Society of

London. Series B, Biological Sciences, Apr 29, 2007;

362(1480): 679–704. doi: 10.1098/rstb.2006.2004.

DBMI (2012). Delta Marktonderzoek. Nieuwe ouderen

eisen nieuwe benadering ouderenzorg. Jan. 13, 2012.

accessed June 2013. Retrieved http://www.dbmi.nl/

blog/item/nieuwe-ouderen-eisen-nieuwe-benadering-

ouderenzorg.

European Commission (2007) Health and Long-term

Care. Report Special Eurobarometer 283. December

2007.

European Commission. (n.d.) DG Health & Consumers,

Public health, Ageing, Policy. Retrieved June2014,

from http://ec.europa.eu/health/ageing/policy/index_

en.htm.

Hermanowicz, J. (2002) The Great Interview: 25

Strategies For Studying People in Bed. Qualitative

Sociology, p. 480.

Hoorn J.F, Konijn E.A, Germans D.M, Burger S. &

Munneke A. (2015). The In-Between Machine: The

Unique Value Proposition Of A Robot Or Why We Are

Modelling The Wrong Things. International

Conference on Agents and Artificial Intelligence,

ICAART, Lissabon, Portugal, Jan, 10-12, 2015.

Klein Wolterink G. (2013) Informatiestandaarden in de

Zorg. Den Haag Nictiz ID 13013.

Leget G. O, (2012) Menslievende Zorg in de Praktijk. Den

Haag: Boom/ Lemma, 2012.

MarketingCharts. (2007). Senior citizens fear moving into

a nursing home and losing their independence more

than they fear death. Accessed June 2013. Retrieved

from: http://www.marketingcharts.com/demographics-

and-audiences/boomers-and-older/seniors-fear-loss-of-

independence-nursing-homes-more-than-death-2343/

Reed J and Payton J. R. (1997). Focus Groups: Issues of

Analysing and Interpretation. Journal of advanced

nursing Oct; 26(4):765-71.

Sharkey A and Sharkey N. (2012) Granny and the robots;

ethical issues for robotcare in the elderly. Ethics and

Information Technology, Volume 14, Issue 1, pp 27-

40.

Vallor S. (2013) Carebots and Caregivers: Sustaining the

Ethical Ideal of Care in the Twenty-First Century.

Philosophy & Technology. Vol. 24, Issue 3 , pp 251-

268 doi: 10.1007/s13347-011-0015-x.

Zorg voor beter (2014).

Nieuwe toelatingseisen

verzorgingshuis. Jan. 13, 2014. accessed 22-11-2014.

Retrieved from http://www.zorgvoorbeter.nl/ouderen

zorg/Nieuwe-toelatingseisen-verzorgingshuis.html.

RobotsHumanizeCare-MoralConcernsVersusWitnessedBenefitsfortheElderly

653