key: cord-0721461-1hefq0na authors: Obayashi, Kazuko; Kodate, Naonori; Masuyama, Shigeru title: Assessing the Impact of an Original Soft Communicative Robot in a Nursing Home in Japan: Will Softness or Conversations Bring more Smiles to Older People? date: 2021-08-07 journal: Int J Soc Robot DOI: 10.1007/s12369-021-00815-4 sha: 5ec1ef927f8dbf16bfa423dd4fa40c72131efc5c doc_id: 721461 cord_uid: 1hefq0na It has been reported that robotics-aided care can contribute to enhancing older people’s social participation and quality of life in nursing homes, while simultaneously reducing the burden on care professionals at nighttime. Due to increasing demand for social care and the relative workforce shortage, it is likely that a greater number and variety of robots will be introduced and implemented in the future. While the benefits of applying robots and assistive technologies are recognized, the current limitations and weaknesses have also been identified. One of these is the difficulty associated with a user-centered design, involving older adults with impaired cognitive and sensory abilities in nursing homes. In order to overcome this challenge, a project was carried out to develop a soft and compact bedside communication robot with an input/output device, connected to existing technologies (e.g. monitoring camera, biological sensor). Drawing on the principle of gemba (deference to frontline professionals’ experience, expertise and skills), users’ feedback was reflected in the iterative steps of robot development. The original soft and communicative robot was introduced and its effectiveness was tested by measuring older people’s reactions and changes in their behaviors and engagement levels. The article reports the development process and results of a small-scale evaluation study, comparing the impact of this original soft-type robot with and without its communicative functions. The human–robot interactions were captured on video, and the analysis revealed that while communicative robots reduced the psychosocial burden on older adults, positive emotional, verbal, visual and behavioral engagement was generated with the help of the non-verbal plush toy. The global population aged 60 years or over reached 962 million in 2017, more than twice the population reported in 1980. The pace of aging is accelerating across the world and the number of older people is projected to reach more than 2 billion by 2050. Accordingly, the number of people affected by dementia has been on the increase, and the current number (approximately 47 million, as of 2018) is estimated to reach 75 million by 2030 [1] . The World Health Organization (WHO) published a report in 2015 [2] , which outlined a framework for action to foster healthy aging based on the new concept of functional ability. In order to support healthy aging and meet the increasing demand (e.g. shortage of workforce), manufacturers and policymakers across the globe are emphasizing development and production of more efficient and effective assistive technologies (ATs) including care robots. Furthermore, the recent global pandemic has affected the lives of older people particularly harshly, and reinforced the view that building capacity in health systems by utilizing technologies should be one of the public policy priorities. Research shows that older people's acceptance of care robots can be affected by the robot's appearance, materials and functionalities [3, 4] . It is suggested that even younger users prefer soft and cuddly robots over those with a cold, hard shell [5] . However, thus far usability, user-friendliness and outcome evaluation of these care robots have rarely been tested from the end-users' point of view [6] . Moreover, in long-term care settings, it is not uncommon that some older users cannot verbally express their needs. To address these weaknesses, a user-centered design has been introduced for socially assistive robots (SARs) and promoted in some countries [7] [8] [9] . Globally, the Responsible Research and Innovation (RRI) framework has been proposed and used [9] . In Japan in recent years, users' needs have been explored and identified by several key players such as the Japanese Association of Occupational Therapists and the Association of Technical Aids (established as a public interest incorporated foundation in 1987) [10] . These efforts are primarily targeted at understanding the 'needs' of users (care recipients and caregivers) so that manufacturers, funding bodies and care providers can develop their 'seeds' and deliver technological aids. Despite these efforts, user involvement is still a rare occurrence. This is partly because hurdles to a participatory approach remain high when the users include older adults with impaired ability. Under such circumstances, caregivers' inputs as proxy for older people's voice become highly important and valuable, although it should be acknowledged that this is not the same as public and patient involvement. In the Japanese context, with the concept of gemba ('actual site/field' in Japanese), professional expertise, experience and skills have been highly valued. Originating in the manufacturing industry, gemba has been adopted in healthcare, alongside the more widely-known concept of kaizen ('improvement' in Japanese), even outside Japan [10, 11] . Guided by this bottom-up process and principle, frontline care professionals proactively identified the needs of care recipients, incorporated changes into the development of SARs, and received feedback from the care recipients. We hypothesized that by making SARs more user-friendly and accessible, better outcomes and a higher quality of life would result. In order to examine this hypothesis, research was conducted to assess the impact of using such SARs on older people. Concerning the outcome evaluation of using existing SARs, previous studies highlighted positive impacts on older people [12] [13] [14] [15] [16] [17] [18] [19] [20] . Examples from the previous studies include the positive impact of SARs on older people's activity and social participation [4, 21] and on older adults with moder-ate dementia [22, 23] . When using a soft tactile robot, the non-verbal type can elicit reactions in older people with advanced dementia [22] . On the other hand, when using verbal type SARs, research indicated that other features such as input/output devices and sound volume need adjustments and improvements in order for the care system to utilize the equipment effectively [22] . Further development of soft and cuddly SARs and the evaluation of their impact on older people in residential care settings was deemed necessary to advance the science behind the effectiveness of these care robots. Against this background, this project was launched to develop and assess a bedside communication soft robot with an input/output device, connected to a monitoring camera and biological sensor, while reflecting the needs of direct users (care professionals and care recipients). Given our previous research, the purpose was to design and develop a soft robot [23] that can be used both as a non-verbal and verbal companion plush toy, as well as a socially interactive communication tool. The effectiveness of the SAR produced was also tested. The article reports the development process of this robot named Mon-chan (a monster-looking doll bringing harmonious communications among users), and presents results of an evaluation study conducted in the nursing home, exploring future research agendas. The development process started from users' perspectives (older persons and care professionals), extracting the core issues identified with the previous setup and using various assistive technologies (a monitoring camera and communication robots, in particular). The gemba principle drove this process with frontline care professionals representing the voice of care recipients, particularly those who are not physically mobile enough to reach and press the nurse call button when they need assistance. In Japan, there are different types of care facilities. The facility selected for this study is a special nursing home for older people. There are 30 residents (25 female, 5 male; 86.8 ±6.8 years old). From the group of care professionals, 4 senior staff members in the nursing home participated in the design process of the robot and the data collection. The years of experience vary from 7 to 20 (11 on average) years. In order to ensure objectivity, the analysis was conducted by three researchers who were not involved in the collection of data. Ethics approval was granted by the Social Welfare Corporation Tokyo Sacramental Ethics Committee (TS 2019-004). Consent was sought from each participant and their family. The research was conducted between October 2019 and March 2020. Most conventional communication robots have a hard shell which gives a cold impression for both residents and staff. As suggested by previous research, the sense of touch is a critical element for older adults' engagement with social robots [5, 6, 16, [24] [25] [26] . The hard shell also meant that physical deployment is highly restricted in the context of nursing homes, as it requires a solid surface, such as a bedside It was hoped that a work system, aided by ATs, could send the warning signs or visual information wherever care professionals are, so that they can arrive in time before problems such as falls occur. In the past, there were some weaknesses in the ICT environment issues, such as time lag via the Cloud and malfunctions. User-centered design dictates that priority should be placed on development of a communication robot that meets their needs. The overall concept therefore was to create a soft shell SAR with an input/output device which is connected to the nurse call button and the monitoring camera. Figure 1 describes the overall concept and the development process. The system requirements include the capability of (i) remotely monitoring and sending signals to care staff when help is needed; (ii) watching over older people around the bed; and (iii) providing older people with a voice-controlled alternative method to a nurse call; and (iv) small, compact and audible robot with soft and warm appearance, which can be securely hung over the bed-rail without hindering its vertical movement. The first two requirements ((i) and (ii)) were fulfilled by the existing devices, which monitor the person with a bedside infrared camera. In case of emergencies such as falls, alerts were sent to the central nursing station as well as care professionals on duty. For the third requirement (iii), a voice-controlled alternative method was considered at the beginning. However, the idea of developing and installing a two-way communication system into the robot had to be rejected for reasons of cost and time. Instead, the team decided to rely on the existing nurse call system, and this was activated when the alert was received by care staff, so that the care recipient in need of help could communicate verbally. The main focus of this project was narrowed down to fulfil (iv). From basic design to development of the interface design, the iterative, PDCA (plan, do, check, do, act) cycle method was applied (Fig. 2 ). The processes for the robot (exterior) design and internal functional design were separated, and the latter proceeded with the development of verbal communication functions (e.g. system design and scenarios for conversations) and non-verbal aspects. Much attention was paid to the human-robot interface design, with consideration of its appearance and voice functionality. At each stage, discussions were held between manufacturers and frontline care professionals. As a result, softer plush fabrics were chosen, and the SAR made more compact, yet more stable on the bedrail using larger paws, though it was not feasible to add a variety of voice types. Table 1 shows the changes made to the design and functions of the SAR after the iterative process of consultations. One of the key aspects of the SAR design was the set of conversation scenarios, which were developed based on daily care routines (see Table 2 ). Messages about meals, wake-up, and bedtime (the timing of wake-up and bedtime messages were set individually) was input for all residents. A conversation scenario about recreation and toileting was tailored for each person, reflecting their lifestyles and daily patterns. In addition, the team enabled the scenario to be adjusted for each person, based on previous recordings captured by the monitoring camera. This was particularly important for conversations taking place during the night, when residents wake up. When the movement of a resident in her room is detected, the signal sent from the monitoring sensor (installed in each room) instructs the SAR's server to speak, and based on the created scenario, recorded voice and sound synthesis are transferred via the SIP server, and the SAR speaks to As previously mentioned, Mon-chan is a communication robot that has soft appearance and pleasant touch, resembling a non-verbal pet-type robot. When it is used by the bedside, it is also linked with an infrared monitoring camera and a biological sensor. For this evaluation, we employed two methods (the interRAI and Video Coding Protocol-Incorporating Observed Emotion (VC-IOE)). Verbal Mon-chan was used at all times in each resident's room for all participants, while Mon-chan, in both verbal and non-verbal mode, was tested when the latter method was used to observe each resident's personal interactions with Mon-chan. The evaluation took place for 4 weeks during the months of February and March 2020. The videorecording and observation sessions were held in the fourth week. The interRAI is one of the standardized assessment instruments for those who receive care. These scales measure physical abilities (activities of daily living (ADLs)), cognitive impairment and quality of life. The assessment items are organized into sections dealing with issues such as cognitive patterns, communication and hearing patterns, and physical functionality [27] [28] [29] . The self-rated or assessor-rated inter-RAI scores can be used as proxy for older people's Quality of Life (QoL). Prior to the introduction of Mon-chan, we collated residents' scores for two categories (E: Mood and Behavior; F: Psychosocial well-being) [30] , and compared them with those collected 4 weeks after the introduction of Mon-chan ( Table 4 ). The higher the scores are, the larger the burden is on mood, behavior and psychosocial well-being. The results are presented below (Sect. 3.2). The other method, VC-IOE was devised by Jones et al. [32] for observing short-term emotional changes when older persons use the pet-type robot PARO. The list of observation items include six categories "Emotional engagement", "Verbal engagement", "Visual engagement", "Behavioral engagement", "Collective engagement" and "Agitation"). This coding framework was adapted and translated into Japanese [ [32] , p.380], and used for the current study ( Table 5 ). The participants were asked to spend time with Mon-chan individually during their recreation, and their 3-min interactions were video-taped with their permission (Fig. 4) . A pair of care professionals observed the interactions and took notes. This was carried out partly to test individual responses to Mon-chan, which is designed to be both a nonverbal (cuddly toy) and a verbal robot. Mon-chan in both modes was examined. Three researchers (two care professionals and one external researcher) subsequently analyzed the recorded videos, using the VC-IOE method (Japanese version). The disagreements between coders were adjusted The overall interRAI scores for the two categories were 11.9 ± 9.7 before the introduction of Mon-chan. Four weeks later, the scores changed to 10.2±7.9, which was statistically significant ( p < 0.01). At the individual level, 57 percent of participants' scores improved while 20 percent remained the same and 23 percent worsened. In particular, E1 "Indicators of possible depressed, anxious or sad mood" showed a marked improvement ( p < 0.005), while E3 "behavior and mood" showed some improvements ( p = 0.08). Looking at each item, E1b (persistent anger with self or others, p = 0.03), E1d (repetitive health complaints, p = 0.02), E1e (repetitive anxious complaints/concerns (non-health related), p = 0.02), E1h (recurrent statements that something terrible is about to happen, p = 0.03), these individual items for category E also indicated statistically significant improvement ( p < 0.05). Furthermore, positive improvements were also found in E3b (Verbal abuse, p = 0.08), E3e (Inappropriate public sexual behavior or public disrobing, p = 0.08) and E3g (Absconding or at risk of absconding, p = 0.08). On the contrary, F4 (Length of time alone during the day, p = 0.39), F5 (Willingness to initiate or participate, p = 0.4) and F6 (Instability of interpersonal relationship was statistically significant, p = 0.33) did not show any change. Overall the psychosocial burden was reduced, and it can be argued that Mon-chan had a positive effect on the quality of life of the participants. For this study, each item was given a score of − 2 (worse), − 1 (bad), 0 (no change), + 1 (good), + 2 (improvement) by the coder, observing the three-minute clip. Although it was a short-term change, a positive stimulating effect was observed for the categories ("Emotional engagement", "Verbal engagement", "Visual engagement", "Behavioral engagement"). Figure 5 shows the comparison of total scores in each category between verbal and non-verbal Mon-chan. The results demonstrate particularly positive engagement with the (non-verbal) Mon-chan doll compared to that with the verbal Mon-chan robot. However there was no statistically significant difference. The combined counts of positive scores ("good" and "improvement") accounted for more than half (55.2%) for non-verbal Mon-chan, and 42.0% for verbal Mon-chan (Fig. 5) . We further investigated the data for nonverbal Mon-chan's effect on the participants with dementia. Focusing on non-verbal responses, we used the Mini-Mental State Exam (MMSE) scores to measure the level of dementia, as it was deemed pertinent to examine whether any patterns were detected between those scores and the results from the VC-IOE method. The MMSE is a 30-point test used to measure the degree of cognitive ability. The test includes orientation to time and place, short-term memory, attention and ability to solve problems, language and comprehension and motor skills, and the greater the scores are, the greater the ability is [31] . A mildly positive linear regression was observed (Fig. 6) , suggesting that there is a possible posi-tive pattern depending on the level of dementia. People with dementia can potentially make positive use of the Mon-chan doll. As previous studies show, even soft plush toys without communicative functions can have therapeutic impact on older people with impaired cognitive ability, with the assistance of carers. The research team aimed to develop an SAR which is userfriendly, is appealing, and contributes to better and safer care delivery. The aim was to overcome formerly observed weaknesses of SARs and draw on the views of users (residents and care professionals working on site) in the nursing home. The result was the production of Mon-chan, which was designed, developed, introduced and evaluated on site. User-centered design still remains relatively uncommon in the domain of robotics design and development for care of older people, particularly for SARs in Japan [6] [7] [8] . This can be attributed to several factors, including the difficulty of having a regular dialogue or forum where developers, care recipients and care professionals can meet and discuss their needs and preferences. Moreover, there has been a scarcity of research evidence to support the arguments that these SARs can improve quality of life for older adults. Because of this, stakeholders and users can be skeptical and take a cautious approach [33] . In the last few years, more research results are being published [4, [17] [18] [19] [20] , and this situation is likely to improve. There are also some studies highlighting (potential) users' perceptions and attitudes towards care robots [34, 35] . Ethical issues have also been examined and considered by researchers and bodies such as the IEEE [36] [37] [38] . Furthermore, SARs can only be utilized fully when care recipients and care professionals find them both user-friendly and useful. The application of ATs in this area needs careful consideration of users' needs and workforce development implications. Successful implementation requires multi-level thinking (from user experience and organizational processes to policy and industry context) and the understanding that technology-supported work is cooperative and embedded in organizational routines [39] . For this study, we also assessed the changes in interRAI scores over the 4-week period when Mon-chan was used. The results indicated that mood and behavior of the participants improved significantly, while psychosocial aspects remained stable. Simultaneously, using the VC-IOE observation framework, the findings showed positive engagement of the participants, irrespective of the difference in type (verbal vs. non-verbal). Non-verbal use of Mon-chan exceeded the expectations of this study, generating very positive engagement in emotional, verbal, visual and behavioral aspects (Fig. 5) . While the mechanism through which the technologydriven system improved residents' quality of life requires future study, it could be argued that the reflection of their voices in the design process of Mon-chan contributed to their engagement. On the other hand, the positive impact of Mon-chan with communicative functions was not clear, considering the relatively lower positive scores than those for non-verbal Mon-chan (Fig. 4) . Although the improved sound quality of Mon-chan's verbal function was noted, the interactions between care recipients and verbal Mon-chan did not yield positive behavioral or collective engagement. Therefore, as indicated by the previous study [40] , the impact of verbal SARs and non-verbal plush toys on older adults with dementia is inconclusive. Further research is needed to examine the relationship and improve the communication functionalities. It is also worth investigating how we can accurately measure the impact of these SARs and ATs on the quality of life of older people. There are a variety of assessment tools such as the WHO's International Classification of Functioning, Disability and Health (ICF), Dementia Behavior Disturbance Scale-13 (DBD-13), Cohen-Manfield Agitation Inventory (CMAI), Neuropsychiatric Inventory with Caregiver Distress Scale (NPI-D), with the interRAI being merely one of them. For our study, inter-RAI was deemed highly effective and accurate in capturing the changes in residents' mood and behavior, though this merits further examination. The limitations of this study include the small sample size and the absence of a longitudinal study design to examine the impact of a longer period of SAR use on care recipients, care professionals and the nursing home. The analysis of contributing factors such as organizational context, physical environment, and the impact of unanticipated disruptions to daily care routines (e.g. Covid19 pandemic) is also worth considering, although it is beyond the scope of this paper. Nonetheless, the study demonstrated the significance of care professionals' involvement in the design and development of an original soft social robot as a way of translating the voice of older adults with impaired cognitive and sensory ability. By assessing the impact, and despite some inconclusive results, the production loop of SARs was closed, which fulfilled the principle of gemba. Whether care professionals' involvement in SARs design, production and evaluation can lead to greater job satisfaction and a better work environment also needs further research. However, this study is well-aligned with the progress towards the RRI framework [9] , with greater public and patient involvement. Future research into and development of care robots need to examine the connection between process and outcomes. This study outlined the user-centered design and evaluation of a socially assistive robot within the context of a nursing home in Japan. The collaborative developmental process among different stakeholders-manufacturers, care professionals, care recipients and researchers, improved quality of life among residents. The original SAR "Mon-chan" was connected to existing ATs (monitoring camera, nurse call, and biomedical sensor), which contributed to safety and enhanced quality of care. This safety function, in addition to the socially assistive one, provides another layer of security for care recipients and care professionals. As the new coronavirus infectious disease has highlighted, there is now a new focus on the introduction of telemedicine, and applications of remote monitoring and care delivery models as countermeasures for the spread of the virus and the protection of older people in nursing care facilities. The importance of user-oriented design and development of robotics-aided care for the improvement of a care system will continue to be highlighted in the future. The authors declare that they have no conflict of interest. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecomm ons.org/licenses/by/4.0/. World report on ageing and health. World Health Organization Acceptance of healthcare robots for the older population: review and future directions Socially assistive robots and their potential in enhancing older people's activity and social participation Softness, warmth, and responsiveness improve robot hugs Companion robots for older people: importance of user-centred design demonstrated through observations and focus groups comparing preferences of older people and roboticists in South West England Designing socially assistive robots for Alzheimer's disease and related dementia patients and their caregivers: where we are and where we are headed Ethics of healthcare robotics: towards responsible research and innovation Using Gemba boards to facilitate evidence-based practice in critical care Going to the Gemba: identifying opportunities for improvement in radiology Information and communication technology systems to improve quality of life and safety of Alzheimer's disease patients: a multicenter international survey Comparison of verbal and emotional responses of elderly people with mild/moderate dementia and those with severe dementia in responses to seal robot Effectiveness of robot PARO in intramural psychogeriatric care: a multicenter quasi-experimental study Use of a robotic seal as a therapeutic tool to improve dementia symptoms: a cluster-randomized controlled trial The multi-modal interface of robot-era multi-robot services tailored for the elderly Supporting human autonomy in a robot-assisted medication sorting task The multi-modal interface of robotera multi-robot services tailored for the elderly She had a smile on her face as wide as the great Australian bite": a qualitative examination of family perceptions of a therapeutic robot and a plush toy Enhancing older people's activity and participation with socially assistive robots: a multicentre quasi-experimental study using the ICF framework Can connected technologies improve sleep quality and safety of older adults and care-givers? An evaluation study of sleep monitors and communicative robots at a residential care home in Japan Measuring the impact of age, gender and dementia on communication-robot interventions in residential care homes Measuring impacts of introducing communicative robots with infrared radiation monitoring system on workload of night shifts in nursing care facilities Factors affecting the acceptability of social robots by older adults including people with dementia or cognitive impairment: a literature review The benefits of and barriers to using a social robot PARO in care settings: a scoping review Technical Committee for Soft Robotics A good life in old age? Monitoring and improving quality in long-term care. OECD/European Commission, Ch. Using interRAI assessment systems to measure and maintain quality of long-term care inter-RAI home care quality indicators The inter-RAI suite of mental health assessment instruments: an integrated system for the continuum of care interRAI hoshiki gaido bukku: kea puran sakusei/shitsu no kanri/kango deno katsuyo (interrai method guidebook: care planning, quality control and utilization in nursing An examination of psychometric properties of the mini-mental status examination and the standardized mini-mental status examination: implications for clinical practice Assessing engagement in people with dementia: a new approach to assessment using video analysis Robots vs migrants? Reconfiguring the future of Japanese institutional eldercare The use of care robots in aged care: a systematic review of argumentbased ethics literature Exploring perceptions towards home-care robots for older people in Finland, Ireland, and Japan: a comparative questionnaire study Assistive technologies for people with dementia: ethical considerations A value-driven eldercare robot: virtual and physical instantiations of a casesupported principle-based behavior paradigm Ethically aligned design: a vision for prioritizing human wellbeing with autonomous and intelligent systems SCALS: a fourth-generation study of assisted living technologies in their organisational, social, political and policy context The adoption of care robots in home care-a survey on the attitudes of Finnish home care personnel Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations Acknowledgements This study was in part financially supported by The Japan Keirin Autorace Foundation and The Toyota Foundation (D18-ST-0005). The authors wish to thank all the participants, funding bodies, A.I.Viewlife Co. Ltd., NTT Data Corporation, Trendmaster Ltd. and Yukai Engineering Inc., and staff members who provided help. We would particularly like to thank Hiromasa Kondo, Yoshimi Okamoto, Yoko Ishii, Takahiro Nonoda and Mark Linne.