PRE2024 3 Group9

Group members

Group member	Student number	Study program
Amélie van Rossum	1743929	BPT
Annika Valkering	1890395	BPT
Koen Jaartsveld	1855786	BPT
Lisa van Nuland	1778528	BPT
Max te Brake	1534017	BPT
Carmen Knepper	1884069	BPT

State-of-the-art

Current research in dance therapy has seen multi-faced benefits of dance and DMT (dance movement therapy) on the physical, cognitive, and emotional well-being of elderly populations, and people with physical disabilities. Simply put, these findings posit that dance therapy has much more to offer than just a rehabilitative practice on a physical level.

State of the art in dance therapy

Koch et al. (2019) claimed that dance and DMT have been sufficiently established to reduce symptoms of depression, anxiety, and stress and to enhance the general well-being of all populations, including patients with mental illnesses, and the elderly. This meta-analysis demonstrated that dance is effective as a therapeutic intervention for  psychological health improvement, thus highlighting the need for further exploration into the specific mechanisms that facilitate these benefits .

McGonigal (2019) explains that activities of a physical nature like dance can bring about changes at the physiological level that elevate mood and state of mind. Exercise constitute to the release of endorphins, dopamine, and endocannabinoids which help in the reduction of stress and tension, reinforce social ties, and bring resilience. This is in line with the idea that group dancing not only sharpens physical fitness but also promotes socialization and emotional ties, thereby yielding an all-encompassing therapeutic impact.

Swaine et al. (2020) discuss the role of dance therapy as complementary to conventional physical rehabilitation. It helps improve mobility and participation among people with disabilities. Directly, it did not produce a significant treatment effect on mood. However, the level of engagement evidenced by improvements in the Flow State Scale suggests very high levels of psychological immersion and enjoyment, which are critical to mood improvement.

Gouvêa et al. (2017) provides insight that dance programs have been shown to impact the quality of life of the elderly positively, with improvements in physical parameters such as balance, agility, and reduction of state anxiety. Findings imply that involvement in these activities can bring about positive changes in the physical and environmental quality of life aspects for the elderly, thereby promoting general well-being and possibly reducing symptoms of anxiety and depression

Koch, S. C., Riege, R. F. F., Tisborn, K., Biondo, J., Martin, L., & Beelmann, A. (2019). Effects of dance movement therapy and dance on health-related psychological outcomes: A meta-analysis update. Frontiers in Psychology, 10, 1806. https://doi.org/10.3389/fpsyg.2019.01806

McGonigal, K. (2019). The joy of movement: How exercise helps us find happiness, hope, connection, and courage. Avery

Swaine, B., Poncet, F., Lachance, B., Proulx-Goulet, C., Bergeron, V., Brousse É., Lamoureux, J., & McKinley, P. (2020). The Effectiveness of Dance Therapy as an Adjunct to Rehabilitation of Adults With a Physical Disability. Frontiers in Psychology, 11, 1963. https://doi.org/10.3389/fpsyg.2020.01963

Gouvêa, J. A. G., Antunes, M. D., Bortolozzi, F., Marques, A. G., Bertolini, S. M. M. G. (2017). Impact of Senior Dance on emotional and motor parameters and quality of life of the elderly. Rev Rene, 18–1, 51–58. http://repositorio.ufc.br/handle/riufc/22169

Robotics in therapeutic settings, especially dance therapy, is an emerging research area at the intersection of rehabilitation, robotics and psychosocial care. Some pioneering studies have shown that robots can play a very important role in physical rehabilitation and social interaction.

State of the art in robotics for therapy

Bevilacqua et al. (2025) proved both the technical feasibility and user acceptance of social robots when applied as dance therapy for patients with Parkinson's disease. From this study, it was interpreted that motor function and social interaction improved; hence, robots can deliver effective therapeutic interventions.

Gianluca, B., Francesco, M., & Barbara, L. (2023), which proposed the integration of dance therapy and social robots in long-term care settings, their study proved the feasibility of the robot supporting mobile, engaged, and emotionally well seniors. Results also proved robots could help carry on dance sessions that are therapeutic and participatory. In this capacity, robots may serve as valuable partners in providing engaging therapeutic interventions to older adults.

Therapeutic robots such as Cody, presented by Goldman in 2012, represent progress in robots meant to enhance balance and walking speed via dance. This project focuses on the advantages partner dance can bring to older adults with mobility problems and proposes robots as possible leaders in physical therapy sessions.

Gassert and Dietz (2018) focus on the neurophysiological principles that should underlie rehabilitation robots, paying particular attention to designing devices adaptive to patient needs based on neuroplasticity. This paper emphasizes the interdisciplinary collaboration necessary for developing robotic rehabilitation systems capable of providing standardized, adaptable, and intensive therapy.

The study by Karim, H. A., Lokman, A. M., & Redzuan, F. (2016) investigates older adults' emotional responses to robot interactions and reveals a general openness and interest in engaging with robotic companions among the elderly. The potential of robots to reduce loneliness and improve quality of life is underscored, as long as their design is emotionally engaging.

Dance Sessions with Robots Li et al.'s (2022) pilot study confirmed once more the warmly received robot-facilitated dance sessions in long-term care environments. The staff and residents showed very positive attitudes towards the robots, with the latter group being slightly more appreciative of the robots' user-friendliness, safety, and entertainment value. However, they mostly investigated the quantitative evaluations of user acceptance and demographic impacts but are noted for their emphasis on general interaction.

Bevilacqua, R., Maranesi, E., Benadduci, M., Cortellessa, G., Umbrico, A., Fracasso, F., Melone, G., Margaritini, A., La Forgia, A., Di Bitonto, P., Potenza, A., Fiorini, L., & La Viola, C. (2025). Exploring dance as a therapeutic approach for Parkinson disease through the Social Robotics for Active and Healthy Ageing (SI-Robotics): Results from a technical feasibility study. JMIR Aging, 8, e62930. https://doi.org/10.2196/62930

Li, Y., Liang, N., Effati, M., & Nejat, G. (2022). Dances with social robots: A pilot study at long-term care. Robotics, 11, 96. https://doi.org/10.3390/robotics11050096

Goldman, D. (2012). Letting the robot lead. Emory Medicine Magazine. Retrieved from https://emorymedicinemagazine.emory.edu/archives/issues/2012/spring/briefs/letting-the-robot-lead/index.html

Gassert, R., & Dietz, V. (2018). Rehabilitation robots for the treatment of sensorimotor deficits: A neurophysiological perspective. Journal of NeuroEngineering and Rehabilitation, 15, 46. https://doi.org/10.1186/s12984-018-0383-x

Karim, H. A., Lokman, A. M., & Redzuan, F. (2016). Older adults' perspective and emotional response to robot interaction. 4th International Conference on User Science and Engineering (i-USEr), Melaka, Malaysia, pp. 95-99. https://doi.org/10.1109/IUSER.2016.7857941

Li, Y., Liang, N., Effati, M., & Nejat, G. (2022). Dances with social robots: A pilot study at long-term care. Robotics, 11, 96. https://doi.org/10.3390/robotics11050096

Our contributions to the state of the art

Our project aims to bring a big step forward in the domain of robotic-assisted dance therapy by integrating several innovative features into the Pepper robot, which will help overcome the shortcomings regarding personalization, interaction quality, and overall therapeutic efficacy of social robots applied to elderly care environments. To outline the key contributions our project will make to the state of the art:

Greater Personalization

Existing therapeutic robots typically offer little to no personalization. Our project will harness the Pepper robot to deliver an exceptionally personalized therapeutic experience. Pepper will be programmed to modify dance movements and music selections according to different user profiles based on their physical capabilities and preferences of music. These modifications are much more than the simple overall changes that comprise typical therapeutic settings today. They help ensure that every session can be made maximally beneficial and engaging for each participant.

Advanced Response and Engagement

By enhancing the simple feedback we present in the dancing robots, our plan will furnish Pepper with the ability to give more subtle interactions. Pepper will give instant responses and feedback based on people’s actions during dance sessions. This will encompass spoken encouragements and changes to the dance routines to enhance the emotional and bodily support for the participants during therapy sessions, helping create a more supportive and engaging therapeutic setting.

Group Dynamics and Social Interaction

Our project will take place at the level of group dynamics, focusing on the dance sessions, since we consider social interactions to be a very important way to enhance well-being. Pepper will lead group-based activities that motivate participants to move physically and socially interact with each other. This element is very important for creating a feeling of community and togetherness for elderly users, especially in long-term care situations where there is often social isolation.

Our project will boost Pepper’s ability to interact, making it as natural and human-like as possible without using AI. By setting up predefined responses that expect typical interaction patterns during therapy sessions, Pepper will engage users in a way that feels intuitive and emotionally resonant, thereby enhancing the overall user experience.

Low-Impact Dance Steps

Pepper will be programmed with dance steps suitable for elderly users, focusing on low-impact and gentle exercises that can be done sitting or standing, catering to users with different physical abilities.

Mood Enhancement Through Music and Dance

The project will seek to further personalize the mood of participants by providing personalized music choices and intermittently light-intensity dance exercises. In this way, it tries to take advantage of the therapeutic benefits of music and movement for a happy, engaging experience.

Personalization of Movement Intensity

To make sure that everyone participates safely and reaps the benefits of dance therapy, Pepper will modify the intensity and type of movements according to the capacity of each participant.

Hands-Free Controls and Prompt Guidance

In consideration of users with impairments in mobility or cognitive functions, Pepper will implement voice command functionality and sharp visual and auditory cues for better accessibility and ease of use.

This will help our project stretch the potential of social robots in therapeutic settings and play a significant role in the emotional and social aspects that are very important for the elderly participants' well-being. Making sure that therapy is effective, enjoyable, and emotionally rich, makes a big and new contribution to the areas of human-robot interaction.

Problem statement and objectives

Problem statement

Globally populations are aging, making it increasingly important to help elderly people maintain physical health and emotional well-being . Limited mobility in older adults impacts both their physical functioning as well as their emotional health, which often leads to decreased social interaction and reduced quality of life . Traditional physical rehabilitation methods that are effective in improving physical abilities often lack the motivational and emotional engagement necessary for sustained participation . One effective intervention for enhancing mobility, cognitive function, and emotional well-being in older adults is dance movement therapy . More generally, it has also been shown that increased mobility overall significantly improves quality of life in elderly people. However, challenges remain such as accessibility and constant participation, especially for individuals with limited mobility .

The field of social robotics has seen recent advancements that could present a promising solution to these challenges by offering engaging platforms for physical exercise. Specifically, humanoid robots like Pepper show the potential to be dance companions or instructors. In this way, physical activity can be promoted while also enhancing emotional well-being through social interaction and engagement with a humanoid robot, as well as maybe other older people and caregivers . These robots have the ability to provide personalized, adaptable, and consistent dance-based interventions, which are crucial for users with limited ability such as older people. Research has indicated that integrating robots in dance therapy not only improves physical outcomes such as balance and mobility but also enhances psychological well-being by increasing social engagement as well as emotional satisfaction .

Despite these promising findings, at the moment there is limited research on how effective humanoid robots are as dance companions or instructors for older adults with mobility limitations. Specifically, it is still unclear how it could influence both physical and emotional outcomes including social connectedness, overall quality of life, and mood enhancement. Furthermore, user acceptance, long-term adherence, and engagement in robot-assisted dance programs need further investigation .

Therefore in this study, we aim to explore the feasibility and effectiveness of using the Pepper robot as a dance companion and instructor for elderly individuals with limited mobility. We will investigate the impact of robot-assisted dance interventions on emotional well-being (mood enhancement), physical health, and social engagement. In this way, this research seeks to address the gap in literature regarding the use of humanoid robots in dance-based therapy, as well as contribute to the development of accessible and effective interventions to enhance the quality of life in older adults with limited mobility.

Objectives

The primary objective of this project is to develop and evaluate the effectiveness of the Pepper robot as a dance instructor for older adults and individuals with limited mobility. The project aims to build upon existing research on robotic-assisted dance therapy for older adults and provide insights into the feasibility, user acceptance, and impact of such interventions.

This objective involves programming Pepper to guide a music- and dance-based activity that promotes physical activity, emotional well-being, and social engagement through human-robot interaction while remaining accessible for those with limited mobility. It is essential that the robot instructs people in a way that is engaging and encourages participation, thus it is important the movements and directions the robot demonstrates are within the physical capabilities of the elderly users. Beyond that, the robot's actions should feel natural and emotionally engaging, as research has shown that older adults are generally more open to robotic companionship under these conditions.

Additionally, an experiment is to be carried out to determine the effectiveness of this implementation of Pepper's capabilities. To this end, we aim to construct an experiment and interview with a select cohort of participants according to our target population criteria. Prior to the dance session the participants will engage in with Pepper, they will be interviewed to determine their expectations and current mood, and afterwards they will be interviewed once more to determine the effectiveness of the activity in improving their mood and physical well-being.

Each participant will be properly informed and sign an informed consent form, ensuring they understand the purpose of the research and their role in the experiment.

User description

The users of this project are older people who may no longer move easily. They need gentle and safe activities to stay active and happy. The dance programme is designed to be easy to perform, both sitting and standing, and it aims to prevent them from getting injured while exercising. The aim is to help them feel better, both physically and emotionally, through simple dance moves and fun music they like. In this way, they can have fun, stay healthy and feel connected to others.

User requirements

1.    Low-impact dance movements for elderly users

Create dance programmes that include low-impact movements specifically designed for older users. Incorporate well-controlled ballet and tap dance movements that can be performed by the robot, focusing on arm, hand and head movements that the Pepper robot can demonstrate. With such an approach, older participants can perform the activities sitting or standing, improving their physical endurance, strength and functionality in a safe way. Movement that maintains flexibility and reduces pain does not increase the risk of injury, but has been shown to encourage very high levels of adherence among participants to demonstrate effectiveness and also something worth doing . Including aspects of aerobics that reduce psychological stress and possible mood enhancement would be valuable, as indicated by a study in which low-impact aerobic dance exercises significantly improved stress levels among sedentary women in Malaysia .

2.     Mood enhancement

Intermittent dance exercises

The robot should have intermittent aerobic dance exercises of light intensity (I-LADE) as its core activity. This exercise is considered the best to improve mood in older adults because it is very fun and engaging. According to a study published in 2021 in Frontiers in Aging Neuroscience, both forms of exercise interventions (continuous and intermittent) had a positive impact on mood and executive functions. However, the I-LADE (intermittent) intervention was particularly beneficial because it was more fun, which made it most likely to increase participation for longer. Participants reported that this intervention was clearly more pleasurable and enjoyable than continuous physical activity, allowing this intervention to be included fairly regularly in treatment programmes for older adults, especially as a morale booster .

Classical music

The robot should include classical music with a happy sound, specifically pieces by Mozart, to help improve the mood and cognitive functioning of elderly users. Such music would use the general inherent characteristics of classical compositions to evoke happiness and would therefore be suitable dancing programs to boost mood. Research suggests that listening to happy-sounding music, such as Mozart's, can significantly increase the level of arousal and mood in older people. This kind of music is not only good for mood but also for cognitive functions, working memory and skills needed for voluntary mental activity in older people to maintain overall mental health .

Personalized music

Personalised music interventions could include using favourite genres and songs to improve mood in older users, especially those with stroke or dementia. The robot will facilitate music sessions that are personalised to evoke positive emotional responses that improve mood. Teppo Särkämö's research proves that personalised music, especially well-known and favourite songs, plays a crucial role in improving the mood and emotional well-being of elderly people . Such interventions take advantage of the very close emotional relationships users have with music they like, and activate positive memories and feelings that are exceptionally strong for improving mood. It emphasizes that familiar music evokes stronger emotional responses and activates reward circuits in the brain more effectively than unfamiliar music.

3.     Customization of movement intensity to ensure safety and comfort for all users

The system should be able to adjust movement intensity and style so that all users, particularly older users, feel safe and comfortable. Older users may have varying levels of physical ability. Balance and type of movements should be modified to ensure the safety and comfort of the programme for older participants. This modification is due to the fact older individuals do not have comparable physical capabilities, as inscribed in the synthesis evidence from recent studies on health programmes in dance. Such customized approaches minimize injury risk in dance activities while promoting greater overall participation by making movements accessible and enjoyable for all participants, irrespective of their physical condition .

4.     Hands-Free Controls

The robot must feature hands-free controls, such as voice commands, so that users with mobility or cognitive limitations can still use the robot. According to Zhao et al. (2023), human-robot interaction, perception-based systems and prediction-based systems design to adapt to the user's needs while ensuring safety and comfort. Dance therapy is usually defined as artistic therapy, i.e. coordinated movements and balanced use of music; hands-free control allows the user to interact with the robot without diverting their attention or requiring manual input. In addition, voice controls also enable interaction where users can start, stop or tune dance routines without the burden of physical control input that can be difficult for those with less dexterity or cognitive limitations. Another advantage is that voice guidance can provide real-time encouragement, corrections or instructions that can increase motivation and support dance sessions. Psychological safety has also been shown to greatly aid user acceptance of robotic systems . Letting users control the robot with words rather than by moving their body reduces worry and increases confidence, especially among people who are not so used to technology.

5.     Clear visual and audio guidance

Providing clear and straightforward visual and audio instructions is essential for elderly users, particularly those with cognitive impairments. Research indicates that older adults with cognitive decline often face challenges in comprehension and communication, which can impact their decision-making processes. Utilizing visual methods, such as infographics or pictures, has been shown to support understanding and facilitate communication in this population .

Additionally, sensory impairments, including vision and hearing loss, are common among the elderly and can further hinder information processing. Studies have found that visual and hearing impairments are associated with a higher risk of cognitive decline. The use of visual aids can mitigate some of these challenges by enhancing comprehension and supporting cognitive functions . Incorporating clear visual and audio instructions can also improve working memory performance in older adults. Research examining the effect of auditory-visual speech stimuli on working memory found that combining auditory and visual information can enhance cognitive performance in older adults with hearing impairments .

By integrating clear visual and audio instructions, assistive technologies can better support elderly users, particularly those with cognitive and sensory impairments, leading to improved comprehension, communication, and overall user experience. Research on interactive robotic feedback and balance therapy demonstrates that customized movement intensity, real-time feedback, and hands-free controls significantly improve adherence, engagement, and motor functions in older people .

6.   Encouraging feedback

The dance therapy robot should provide positive reinforcement, acknowledging the progress of the user and encouraging further effort. Physical activities require motivation, and enjoyment, and hence long-term adherence to be successful. Positive reinforcement holds such an important value in these aspects. Research also made it clear that people’s perceptions of their coach’s behaviours, like goal setting, technical guidance, and rapport-building, are the key predictors of exercise engagement over time. For a dance therapy robot, this means incorporating real-time, personalized positive feedback that will enhance the users' intrinsic motivation and enjoyment in activity. The robot should be able to:

• Recognize progress made in moving, balancing, and enduring so that it can help foster a sense of competence
• Give motivational encouragement to add to enjoyment and engagement
• Involve goal-setting mechanisms that track user achievement with reinforcement when milestones are reached

As intrinsic motivation and enjoyment are key for adherence, the feedback system has to ensure that users feel able, supported, and engaged during their dance therapy sessions .

7.     Companion-like interactions

The interaction with the robot has to be friendly, sympathetic, and engaging so that the robot acts more like a dance partner than just an instructor. It should include social and emotional support elements, for example, active listening, personalization, and expressive communication through voice and face. The robot needs to recognize the different users' conversational requirements by providing active engagement to spur motivation and passive companionship when the user requires it. This will encourage older adults to increase motivation, enjoyment, and adherence to activities like dance therapy when done correctly .

The Interview Study

In order to get a better understanding of what the requirements for a robot must be, an interview study will be conducted. Different stakeholders and people give different valuable insights into the desires and needs of the target population, therefore it is important to interview a wide range of people, and making sure that the questions are also complete and valuable.

The participants

Three distinct groups are to be interviewed. First, there are the elders. As this is the target population for which this robot is designed, keeping their wants and needs in mind is vital.

The second group are the caretakers. This group has a lot of knowledge on the abilities and struggles of the target group of this research, and so interviewing them about this robot also has the potential to give a lot of valuable information. They are also often well acquainted with modern technology, which gives the designers the possibility to ask more directly about the robot.

The last group is the generation before our current elders. If this robot is ever developed and used in a care home, it will be a few more years down the line. Therefore, the current older working adults or the recently retired population have a larger chance of getting to interact with the robot than other generations. Therefore, knowing what they want from such an intervention can provide insight that can be utilized in the design process. They also often have more knowledge of the current state of tech and robotics and are more capable of providing detailed directions than the current elders.

The questionsUsThe sets of questions are used more so as a guideline than as a predefined list. They can be found in appendix A. The interviewers were encouraged to ask further questions if they felt there were interesting insights to be gained, or ask entirely different questions if they felt compelled to do so.

For the current elders, the questions were mostly focused on their health and desires for future health. The questions asked about if they felt like they moved enough at the moment, and if they wanted to move more. The questions were also geared towards exploring the interests of the elder, in order to drive engagement once the robot is realized. These questions asked about their music preference, and whether they appreciated the possibility for exercise or the possibility for social interaction more.

The caretakers were asked more technical questions, as well as questions to explore the behavior of the elders. The questions were about how the elders interact with modern technology, skepticism against robots and advanced technology. Other questions were about common movement limitations, and about the importance of a social element in their daily lives. An important point of discussion was whether personalization at a group level would suffice or if individual customization is necessary to get the best results for the elders.

For the future elders, the questions were a combination of the aforementioned categories. The questions were about music taste, interest in exercise, but also about their understanding of technology.

The original questions can be found in appendix A, but all the questions can be found in the transcripts, where they are in bold.

Thematic analysis

After the interviews, a thematic analysis will be conducted, where all the answers to the questions and other interesting observations will be divided into themes that can aid the designers in setting up a program of requirements for the robot. The results of the thematic analysis can be found below.

A link to the Miro board where the analysis was performed can be found below: https://miro.com/app/board/uXjVIO22Vuk=/?share_link_id=111601739888

From this thematic analysis, some key considerations will be addressed in the program of requirements.

Program of requirements

User testing

In order to test the effectiveness of our mood-enhancing dance robot, several user tests have been defined and tested.

User test 1: Basic dialogue

In this user test a user will be asked to interact naturally with Pepper without having prior experience with the dialogue. In this way, the dialogue will be assessed on clarity, comprehension and engagement. The procedure of this user test is to let the participant initiate an interaction with Pepper and observe how the user responds to the dialogue. Specific attention should be paid to any confusion or stumbling blocks in the interaction. After the interaction, the users will be asked about how they assess the interaction based on clarity, comprehension, and engagement. The success criteria for this user test are when the participant can complete the dialogue interaction smoothly without frustration and when Pepper responds appropriately to the user input and the users' assessment of the dialogue.

User test 2: Dance level selection and execution

The aim of this user test is to evaluate how well users can select and experience their preferred dancing level. The procedure is for the user to respond to Pepper's question about which dancing level the user wishes, and then execute the dancing moves. After this, the user will be asked about the ease of use of the selection dialogue and about the enjoyableness/speed & ease to follow the robot's dancing moves. The success criteria for this user test are a successful selection of dance level, and the users' ease and enjoyment of following along with the dances.

User test 3: Elderly does not understand Pepper's dialogue

In this scenario the elderly person is engaged with Pepper's dialogue but does not understand it. The procedure for the user then is to say a sentence like 'I do not understand', to which Pepper should repeat/simplify/rephrase the statement and ask the user if the explanation is clear. An expected Pepper response would be “I’m sorry, let me say that in a simpler way.” [Pepper rephrases the statement] or “Did I say that too fast? I can repeat it more slowly if you’d like.”. Success criteria are when participant's express improved understanding after Pepper's response and it does not cause further frustration or confusion.

User test 4: Off-topic response

It could happen that elderly people start talking about topics irrelevant to Pepper's dialogue even when Pepper has been programmed to ask questions with clear options (like 'yes'/ 'no'/'level 1/2/3'). Pepper should be able to handle such situations too. The test procedure includes the user interacting with Pepper in a dialogue and giving an unexpected response such as 'what time is lunch?', as well as observing Pepper's reaction to this. Pepper could respond to this with sentences such as “I think I didn’t understand that. Could you try saying it in another way?” or “That’s interesting! But I was asking how you’re feeling today.”. Success criteria are when Pepper does not get stuck, the user does not feel ignored, and the conversation can continue smoothly.

User test 5: Elderly user forgets what they were doing

Elderly people can be forgetful. Pepper should be able to navigate this too. Such a scenario could be the participant starting an interaction, but then getting distracted and forgetting what it was about. The test procedure for this is to let a user initiate a conversation with Pepper, getting distracted, and then having forgotten what the conversation was about so hesitating to respond or expressing this forgetfulness. Peppers' responses to this could be “It looks like we were about to choose a dance level. Do you still want to do that?”, “If you forgot what we were talking about, don’t worry! We were picking a fun dance.” or “I can help! Would you like me to repeat what we were doing?”. Success criteria are when the user does not feel embarrassed or confused and Pepper can successfully guide the user back in the interaction.

User test	Objective	Procedure	Success criteria
1	assess clarity dialogue	let users engage naturally with the robot	smooth interaction without confusion where the user rates the dialogue as clear
2	assess ease of use dance level selection and execution	let users select a level and dance along	successful selection of dance level, and ease and enjoyment in following along with the dances
3	assess Pepper's ability to navigate users not understanding the dialogue	let users express them not understanding	improved understanding after Pepper's response without confusion or frustration
4	assess Pepper's ability to navigate off-topic and unexpected responses	let users say sentences Pepper does not expect or understand	Pepper does not get stuck, the user does not feel ignored, and the conversation can continue smoothly
5	assess Pepper's ability to navigate forgetful/distracted users	let users engage with Pepper, then get distracted and not remembering what the conversation was about	Pepper guides user back in the interaction and the user does not feel embarrassed or confused

Discussion

- A robot that can also do individual dances and can come up to elderly

- A robot that can give real-time feedback based on the performance of elderly people (in color: green, orange, red)

- A robot in which Max his python script can be connected to the robot

Planning (approach, milestones and deliverables)

Each week, there will be a mentor meeting on Monday morning to ask questions and get feedback. Furthermore, we have two additional group meetings every week, one directly after the mentor meeting and one on Friday during the break. The Wiki will be updated every Sunday (weekly deliverable).

On a weekly basis, we will evaluate what tasks need to be done and assign the tasks according to skills and interests of the group members.

Week 1

• Introduction to the course and forming of teams - all
• Brainstorm to come up with ideas for a project and select one - all
• Communicate the idea with the course coordinater - Annika
• Conduct literature review (5 pp) - all
• Write problem statement - Lisa
• Write objectives - Max
• Write user description - Amélie
• Write user requirements - Koen + Amélie
• Make planning (containing approaches, milestones, deliverables) for the project - Annika
• Select technology (Pepper, other TU/e robot) - all

Week 2

• Contact the nursing home - Annika
• Contact with robot lab and make appointment - Annika
• Create interview questions for elderly regarding user needs for design - Koen
• State-of-the art update - Amélie
• Look at program opportunities of Pepper - Max
• Create survey about mood for pre- and post-test - Lisa
• ERB form for approval of qualitative experiment - Carmen

Week 3

• Program the robot - Max, Lisa en Annika
• Make consent form - Carmen
• Start finding participants for experiment - everyone

Week 4

• Conduct interview with Vitalis caregiver - Annika & Koen
• Conduct interview with elderly/parents - everyone
• Programming the robot and its dances - Max, Lisa, Annika and Amélie

Week 5

• Test robot - everyone
• Analyze findings of interviews and incorporate in design - everyone
• Incorporate design recommendations in dialog - Max
• Incorporate design recommendations in dances - Lisa
• Program of requirements - Koen, Carmen en Annika

Week 6

• Write results section
• Write discussion section
• Write conclusion section
• Create slides for final presentation - Lisa

Week 7

• Finalize report (milestone 3)
• Prepare final presentation + demo
• Give final presentation (milestone 4)
• Fill in peer review

Individual effort per week

Week 1

Group member	Total hours	Tasks
Amélie van Rossum	11,5	Intro lecture (2h), literature review (2h), group meeting (1h), user needs (4h), lecture robots tue (2h), user description (0,5h)
Annika Valkering	8	Intro lecture (2h), literature review (2h), Wiki layout and draft planning (2h), group meeting (1h), references (1h)
Koen Jaartsveld	8,5	Intro lecture (2h), literature review (2h), group meeting (1h), reading the sources of others (3h), draft Interview questions (0,5h)
Lisa van Nuland	9	Intro lecture (2h), literature review (2h), draft problem definition, user needs and approach (2h), group meeting (1h), problem statement (2h)
Max te Brake	7	Intro lecture (2h), group meeting (1h), literature review (2h), reading the others' literature analysis (1h), objectives & minor adjustments to problem statement (1h)
Carmen Knepper	4	catching up the intro lexture (2h), reading literature analysis of others (2h)

Week 2

Group member	Total hours	Tasks
Amélie van Rossum	9,5	Tutor meeting (0,5h), group meeting (1,5h), research on state of art (4h), group meeting (1h), writing state of art (1,5h), robot dancing research (1h)
Annika Valkering	3,5	Tutor meeting (0,5h), group meeting (1,5h), contact with nursing home (0,5h), group meeting (1h)
Koen Jaartsveld	5,5	Tutor meeting (0,5h), group meeting (1,5h), group meeting (1h), Reading about state of the art (2h), Refine and add to research questions (0,5h)
Lisa van Nuland	6,5	Tutor meeting (0,5h), group meeting (1,5h), research on mood surveys and selecting one (3,5h), group meeting (1h)
Max te Brake	8	Tutor meeting (0,5h), group meeting (1,5h), group meeting (1h), pepper documentation first look & installing programs (4h), pepper capabilities videos (1h)
Carmen Knepper	5,5	Tutor meeting (0,5h), group meeting (1,5h), group meeting (1h), making the erb form (2,5h)

Week 3

Group member	Total hours	Tasks
Amélie van Rossum	2	group meeting (2h)
Annika Valkering	7	Tutor meeting (0,5h), group meeting (1,5h), attempt programming Pepper (2h), Spotify API (1h), group meeting (2h)
Koen Jaartsveld	6	group meeting (2h), Contacting elderly homes (1h), Finding sources to justify user testing (3h)
Lisa van Nuland	5	Tutor meeting (0,5h), group meeting (1,5h), experiment with programming robot dance moves (3h)
Max te Brake	6	Tutor meeting (0,5h), group meeting (1,5h), attempt programming pepper, programming Spotify API (4h)
Carmen Knepper	3	making the informed consent form (3h)

Week 4

Group member	Total hours	Tasks
Amélie van Rossum	9,5	Tutor meeting (0,5h), group meeting (1,5h), robotics lab (2h), trying pepper robot (2h), programming robot dance moves level 1 (2,5h), interview & transcript (1h)
Annika Valkering	9,5	Tutor meeting (0,5h), group meeting (1,5h), robotics lab (2h), interview Vitalis caregiver (2h), ERB form approval (0,5h), programming dialog (2h), interview & transcript (1h)
Koen Jaartsveld	7	Tutor meeting (0,5h), group meeting (1,5h), robotics lab (2h), interview Vitalis caregiver (2h), Write methods for interview study (1h)
Lisa van Nuland	11,5	Tutor meeting (0,5h), group meeting (1,5h), robotics lab (2h), programming robot dance moves level 3 (4h), interview & transcript 1 (1h), programming robot dance moves level 2 (2h), interview 2 (0,5h)
Max te Brake	6	Tutor meeting (0,5h), group meeting (1,5h), robotics lab (2h), programming dialog (2h)
Carmen Knepper	4	Tutor meeting (0,5h), group meeting (1,5h), robotics lab (2h)

Week 5

Group member	Total hours	Tasks
Amélie van Rossum	6,5	Tutor meeting (0,5h), testing with Pepper (3h), group meeting (1,5h), working on thematic analysis (1,5h)
Annika Valkering	8	Tutor meeting (0,5h), testing with Pepper (3h), group meeting (1,5h), interview with an elderly volunteer (1h), transcription and analysis of interview (1h), adjust ERB form for application (1h)
Koen Jaartsveld	7	Tutor meeting (0,5h), testing with Pepper (3h), group meeting (1,5h), Transcribe Interview (2h)
Lisa van Nuland	8	Testing with Pepper (3h), interview transcript 2 (0,5h), group meeting (1,5h), integrate insights from interviews in dances & fix level 1 (3h)
Max te Brake	7	Testing with Pepper (3h), group meeting (1,5h), interview & transcribing (1,5h), pepper dialog (1h)
Carmen Knepper	5	Tutor meeting (0,5h), testing with Pepper (3h), group meeting (1,5h)

Week 6

Group member	Total hours	Tasks
Amélie van Rossum		Thematic analysis (7h), writing thematic analysis in report (..)
Annika Valkering		Tutor meeting (0,5h), group meeting (1,5h)
Koen Jaartsveld		Tutor meeting (0,5h), group meeting (1,5h)
Lisa van Nuland		Tutor meeting (0,5h), group meeting (1,5h), brainstorming user tests and writing user testing (4h)
Max te Brake		Tutor meeting (0,5h), group meeting (1,5h)
Carmen Knepper

Week 7

Group member	Total hours	Tasks
Amélie van Rossum
Annika Valkering
Koen Jaartsveld
Lisa van Nuland
Max te Brake
Carmen Knepper

Total hours spend during the course

Group member	Total hours	Result
Amélie van Rossum
Annika Valkering
Koen Jaartsveld
Lisa van Nuland
Max te Brake
Carmen Knepper

Appendix

Appendix A: Interview Questions

How happy do you feel on average?

How physically fit do you feel on average?

How mentally fit do you feel on average?

Do you think you exercise/move enough?

What kind of movement/exercises do you do weekly?

Would you like to move more? Why or why not?

What kind of music do you like to listen to?

Which song always makes you happy?

What kind of music do you particularly dislike?

Do you think you’d rather dance for the company or for the exercise?

How much exercise would you like to do during a dance session?

If a robot was able to give you the instructions and dance with you, would that change your interest in participating?

If you could personalize the difficulty of your dance through the robot, would that motivate you to participate?

Interview transcripts

https://docs.google.com/document/d/1sdkABqFiG8jitQSJzctcGgyFTwXAWUYwBxXxz9bMH-c/edit?usp=sharing

https://docs.google.com/document/d/1VQ1OGbs9HguFlQWO55uGvIwSDurT2Wc5g2WVnD3qnXU/edit?usp=sharing

https://docs.google.com/document/d/1selfyb1_eLcO6M-QFypnGHlYixZAw9VL5tCD3TmzTuQ/edit?usp=sharing

https://tuenl-my.sharepoint.com/:w:/g/personal/k_a_h_jaartsveld_student_tue_nl/EflDvNcFfIVAnX6bAyK9ATsB5NTyirscBlSv_V96ppg5OQ?e=WPtZwX

https://tuenl-my.sharepoint.com/:w:/g/personal/m_a_j_t_brake_student_tue_nl/EQ_abn6tumBBo7o2FVN0zccBPJgs0ffTK6O3bvLtz87pbA?e=GOy5VR

Appendix B: Pre- and post mood survey

Motivation for survey choice

Various surveys can be used to assess mood. However, for this specific study, we need a survey that is accessible, sensitive to changes in mood, and not too long to avoid fatigue. Taking this into account, PANAS seems the best choice since it is a quick yet comprehensive measure of mood changes before and after the dancing intervention . Other surveys to assess mood are POMS or BRUMS, but these are longer and assess multiple mood dimensions . At the same time, PANAS focusses specifically on positive and negative affect which makes it ideal for detecting shifts in mood resulting from a dancing activity. Furthermore, PANAS is easier to complete than POMS and more precise than the non-verbal SAM scale .

Positive And Negative Affect Schedule (PANAS) Survey

Instructions: Below is a list of words that describe different feelings and emotions. Please indicate to what extent you feel this way right now, at this moment. Circle/select the number that best represents your feeling for each word.

Feeling/Emotion	Very Slightly or Not at All (1)	A Little (2)	Moderately (3)	Quite a Bit (4)	Extremely (5)
Interested	1	2	3	4	5
Excited	1	2	3	4	5
Strong	1	2	3	4	5
Enthusiastic	1	2	3	4	5
Proud	1	2	3	4	5
Alert	1	2	3	4	5
Inspired	1	2	3	4	5
Determined	1	2	3	4	5
Attentive	1	2	3	4	5
Active	1	2	3	4	5
Distressed	1	2	3	4	5
Upset	1	2	3	4	5
Guilty	1	2	3	4	5
Scared	1	2	3	4	5
Hostile	1	2	3	4	5
Irritable	1	2	3	4	5
Ashamed	1	2	3	4	5
Nervous	1	2	3	4	5
Jittery	1	2	3	4	5
Afraid	1	2	3	4	5

Probably our user group will be more proficient in Dutch, therefore we have included a Dutch version as well.

Instructies: Hieronder staat een lijst met woorden die verschillende gevoelens en emoties beschrijven. Geef aan in welke mate u zich op dit moment zo voelt. Omcirkel/selecteer het nummer dat uw gevoel het beste weergeeft.

Gevoel/Emotie	Helemaal niet (1)	Een beetje (2)	Matig (3)	Vrij veel (4)	Heel veel (5)
Geïnteresseerd	1	2	3	4	5
Opgewonden	1	2	3	4	5
Sterk	1	2	3	4	5
Enthousiast	1	2	3	4	5
Trots	1	2	3	4	5
Alarmerend	1	2	3	4	5
Geïnspireerd	1	2	3	4	5
Vastberaden	1	2	3	4	5
Aandachtig	1	2	3	4	5
Actief	1	2	3	4	5
Gekwetst	1	2	3	4	5
Van streek	1	2	3	4	5
Schuldig	1	2	3	4	5
Bang	1	2	3	4	5
Vijandig	1	2	3	4	5
Geïrriteerd	1	2	3	4	5
Beschaamd	1	2	3	4	5
Nerveus	1	2	3	4	5
Rusteloos	1	2	3	4	5
Angstig	1	2	3	4	5

Afterwards the scoring works as follows:

- Positive Affect Score: add the score on items 1, 3, 5, 9, 10, 12, 14, 16, 17 and 19. This score can range from 10-50, where higher scores represent higher levels of positive affect (mean score is 33.3).

- Negative Affect Score: add the score on items 2, 4, 6, 7, 8, 11, 13, 15, 18, and 20. This score also ranges from 10-50, where lower scores represent lower levels of negative affect (mean score is 17.4).

For our research, this survey will be filled in before the dancing intervention and after and the change between the positive and negative affect score between these moments indicates the effect of the dancing intervention on users' mood.

Appendix C: Dancing Robots

NAO robots

https://www.youtube.com/watch?v=SZhE7HBT4oI

https://www.youtube.com/watch?v=2laujomh0JY

Pepper robots

https://www.youtube.com/watch?v=fndpS3Ak_WU

https://www.youtube.com/watch?v=UQhEjrXWTBA

https://www.youtube.com/watch?v=YVGVy_YnKlc

Appendix D: ERB and consent form

ERB form: https://tuenl-my.sharepoint.com/:w:/g/personal/c_knepper_student_tue_nl/Ebwn8EsVO-VOnJrDHDaFt5EB6NbsoTDwfWU3h7y5mE3IIQ?e=pp9hTI

Informed consent form user study: https://tuenl-my.sharepoint.com/:w:/g/personal/c_knepper_student_tue_nl/EWU7SbciNjBBsT3VYa0zJEUBh1_BHnR4Ko8iuh3M3fVHqQ?e=Dq89Zc

Informed consent form interview: https://tuenl-my.sharepoint.com/:w:/g/personal/k_a_h_jaartsveld_student_tue_nl/ESgMzjc1xYxAkGic0vLdfToBnBeQQkE-UMP2vlw8w0-m9A?e=LOFqO4

Appendix E: Thematic Analysis

Thematic Analysis: The Future Elderly

Thematic Analysis: The Current Elderly