An Integrated Intelligent Home Environment for the Provision of Health, Nutrition and Mobility Services to the Elderly
Final Deliverable
D2.7: MOBISERVUser Acceptance Criteria
(Issue 1)
Date of delivery: Dec 20th 2011
Contributing Partners: UWE, SMH
Date: 20th Dec 2011 / Version: Issue 1 ver6
D2.7: MOBISERV User Acceptance Criteria Report – Issue 11/53
Document Control
Title: / D2.7: MOBISERVUser Acceptance CriteriaProject: / MOBISERV (FP7 248434)
Nature: / Report / Dissemination Level: Restricted until publication in journal
Authors: / UWE, SMH
Origin: / UWE
Doc ID: / MOBISERV D2.7 Issue 1 v6
Amendment History
Version / Date / Author / Description/Commentsv0.1 / 2011-09-01 / UWE / First Version
v2 / 2011-09-12 / UWE / Findings from Embodiment Workshops
v3 / 2011-10-10 / UWE / Findings from Survey
v4 / 2011-11-29 / UWE, SMH / Findings from Scenario-focussed Workshops
V4 / 2011-12-1 / UWE, SMH / Draft sent
V5 / 2011-12-16 / UWE, SMH / Final version
V6 / 2011-12-20 / UWE, SMH / Responses to Internal Moderation (SYSTEMA) incorporated
Table of contents
Executive Summary
1Introduction
1.1Scope of the study
1.2Aims and Objectives
1.3Intended Audience
2Review of the literature
2.1Summary
3Methodology for primary research
3.1Focus Groups Workshops
3.1.1Embodiment workshops
3.1.2Scenario-focussed workshops
3.2Survey
4Results of the Embodiment Workshops
4.1Discussions on the nature of an Ideal Robot
4.1.1Functionality
4.1.1.1Memory associated tasks
4.1.1.2Cleaning tasks
4.1.1.3Assistive tasks
4.1.2Behaviour and Appearance
4.2Discussions on the nature of a Nightmare Robot
4.2.1Behaviour and Appearance
4.2.2Loss of Control and Reliability
4.3Discussions prompted by a documentary on robot development
4.3.1Initial Feedback, impressions and opinions
4.3.2Cue card discussion and comments
4.3.2.1Cue Card A
4.3.2.2Cue Card B
4.3.2.3Cue Card C
4.3.2.4Cue Card D
4.3.2.5Cue Card E
5Results of the Scenario-Focussed Workshops
5.1User Acceptance
5.2Embodiment
5.2.1Behaviour and appearance
6Results of the survey and discussions
6.1Some observations from the questionnaires
6.2New robot functions and characteristics drawn from the discussions
7Conclusion and Discussion
8References
9Appendix 1 - Questionnaire responses
Table of Figures
Figure 1 Eight different shaped robots images used in Broadbent et al.’s study
Figure 2 Hopis and In-touch Telemedicine Robot from the Broadbent et al. study
Figure 3 Robot A Care-O-bot II Fraunhofer IPA
Figure 4 Robot B (PaPeRo 2005From NEC)
Figure 5 Robot C Seeker by David Shinsel
Figure 6 Robot D ApriAttenda by Toshiba
Figure 7 Robot F, RIBA "Robot for Interactive Body Assistance" Institute of Physical and Chemical Research, Japan, and Tokai Rubber Industries, Ltd.
List of Tables
Table 1 Breakdown of participants by type of workshop
Glossary
Term / ExplanationMOBISERV / An Integrated Intelligent Home Environment for the Provision of Health, Nutrition and Mobility Services to the Elderly
PRU / Portable Robotic Unit
WHSU / Wearable Health Support Unit
Executive Summary
This document, D2.7: MOBISERV User Acceptance Criteria Report - Issue 1, discusses a range of issues which have been identified as being significant for user acceptance of the MOBISERV system as it is conceptualised and initially implemented, in particular, those relating to the robot which provides the primary interface to the system and the smart garments.
It should be noted that the first MOBISERV prototype (which was due in M21) has not been available for this study, only initial versions of individual components of the MOBISERV system were shown to the participants. As such, this user acceptance criteria report is made available as Issue 1, and Issue 2 will be made available in month 35 when the users have had adequate opportunities to experience the MOBISERV system and technology first-hand, and thus make better informed judgements of what is acceptable and what isnot to them.
The findings in this report are based on review of recent literature, workshop discussions and questionnaires with a range of primary and secondary users.
In summary, the criteria for user acceptance are as follows:
- High level of intuitiveness for interfaces and interactions with high technical performance and reliability
Successful interaction is based on knowledge of context – if a robot is perceived to be “intelligent” it will be expected to interact “intelligently” with appropriate feedback to user interactivity. When this is not present, or not easily discernable, then frustration can ensue, as well as a feeling of failure. It is natural for people to assume that if the robot did not respond as expected, then the mistake was theirs for not having communicated correctly, which can result in low self-efficacy, whereas the poor response is invariably due to the poor quality of voice recognition and touch-screen performance.
- Flexibility to enable personalisation of system features and settings
For a robot this means personalising gender, voice, accent, speed, nature and style of feedback/interaction e.g. the graphic user interface specifications. The personalisation would ensure that the system could be integrated into the persons’ particular social context, which they felt comfortable with. This customisation would also ensure that the system would cope with the users’ specific needs and limitations, both physical and cognitive.
- Offer assurance of operational safety
- Provide means for people to create their own engaging learning about the system benefits by controlling information privacy
A key part of acceptance is for older people to see the benefit of the system for themselves. This can be achieved by allowing them to privately build a picture of their own unshared data within the system, enable them to review it and share it with a carer when they are ready. In this way, older people can learn about the system features from an engaging example (i.e. their own data), rather than from some generic data.
- Enable fine grained privacy control
Sharing data is not simply a matter of yes or no, all or nothing. Higher-level abstract representations of health status are currently accepted as forms of monitoring by older people.
- Respect a person’s personal routine
Learning behavioural patterns, and based on this making informed decisions of when and where it is appropriate to interrupt and not interrupt the person and give context appropriate responses and help.
- Not undermine existing human contact
People are concerned that the robot will replace the current care and support they receive from human carers. The argument that there will not be adequate carers available in the future due to the vast numbers of older people in society is not well known and thus not adequate to justify the need for a robot. As such it is important to clarify the role of the MOBISERV system in aiding and assisting current carer models and thus enhancing the quality and level of service.
- Offer mixed format training and a phased introduction – to prevent a feeling of intrusion
Training, in the form of videos, instruction leaflets, pictorial cards, addressing different learning needs, and brief introductory interactions spaced over a period of time, would make the system seem more familiar to the person, making them more comfortable with having the system around.
- Provide a clearly defined case for how the MOBISERV system, and in particular the robot, will improve the quality of life, from a wellbeing, as well as social perspective
It is difficult for a person to psychologically accept that they actually require assistance – most people prefer to remain in denial. A robot assistant or companion could make the admission of need too real for people to accept, which could also be seen as a social stigma. The robot really needs to offer significant support and be provided as an alternative to care by unknown human carers and/or residential care.
- Involving all stakeholders in the design of the technology in a participatory manner
Engagement with all stakeholders within different levels and forms of participatory design will help to ensure that stakeholder needs are explicitly considered and addressed in the way the system is developed and the functionality shaped. It should be recognised there is some tension between the needs of primary users (e.g. privacy) and secondary users (e.g. monitoring). It is imperative for user acceptance that the needs of primary users are met above the needs of secondary users – after all, secondary users are dependent on primary users accepting it.
1Introduction
1.1Scope of the study
In developing the MOBISERV system and technology it is important to consider and define what aspects will make the technology acceptable to the users. Without the presence of the integrated system, users can only base their views on the design concept and initial components implementation and have to imagine the scenarios to be able to say what will, and what won’t, be acceptable to them. They can only do so on the basis of their past experiences, often shaped by the media – science fiction movies that they have seen and books they have read. The research team can offer them realistic scenarios and examples, and artefacts such as concept videos and existing technologies and initial version of individual components, however this will only provide a guide to prompt the discussion. A realistic understanding of user acceptance criteria can only emerge when users have experienced something tangible and understood fully the scope of what is proposed by interacting with it, to truly say what is acceptable and what isn’t.
This document therefore discusses a range of issues which have been identified as being significant for user acceptance of the MOBISERV system mainly as it is conceptualised(and less as it is initially implemented), in particular, those relating to the robot which provides the primary interface to the system and the smart garments.
It should be noted that as the first MOBISERV prototype (which was due in M21) has not been available for this study, only initial versions of individual components of the MOBISERV system were shown to the participants. As such this user acceptance criteria report is made available as Issue 1, and Issue 2 will be made available in month 35 when the users have had adequate opportunities to experience the MOBISERV system and technology first-hand, and thus make better informed judgements of what is and what isn’t acceptable to them.
The findings in this report are based on review of recent literature, workshop discussions and questionnaires with a range of primary and secondary users.
1.2Aims and Objectives
The purpose of the primary research for this study is to ascertain the criteria that will maximise the user acceptance of the MOBISERV system and components. To this end we aim to:
- Discover older people’s perceptions, expectations and impression of domestic care service robots and other MOBISERV assistive technologies
- Provide an opportunity for members of the target user groups to discuss both their’s and each other’s perceptions, expectations and impression of the MOBISERV system.
- Discover what potential functions and tasks members of the target user groups would expect such system to perform.
- Discover with members of the target user groups:1. what their individual ideal embodiment preferences and requirements are for a domestic care service robot, 2. What physical and functional properties would be ideal for other assistive technologies.
MOBISERVFP7 248434
D2.7: MOBISERV User Acceptance Criteria Report – Issue 11/53
1.3Intended Audience
This report provides key guidance for the consortium for shaping future developments of the technology as well as for considering aspects that will be significant for the exploitation and dissemination of the system as a whole and the various sub-components. In addition it will also guide the perspectives and agenda for public engagement and dissemination to promote further conversations and discussions about how this technology can be designed to be acceptable.
It should be noted that this report (Issue 1) is mainly based on the conceptualised MOBISERV system, rather than the actual integrated prototype (as this was not available) and Issue 2 of the report, which will be delivered at the end of the project (M35) will relate to the actual MOBISERV system.
2Review of the literature
This section consists of a brief literature review of studies conducted in order to discover and take into consideration stakeholder perspectives and acceptance criteria in regards to human-robot interaction within the home.
In one of the early studies regarding embodiments of assistive robots, Mataric (2005) argues that a robot’s physical presence and shared context with the end user is a fundamental and crucial area of design and development for assistive robots in order to provide a supportive role for older adults. This is due to the complex nature of the interaction that must take place between the end user and the machine in order to establish credible human-robot interaction. Mataric also argues that the same supportive role can be provided by disembodied solutions embedded within the user’s environment.
Mataric draws upon prior research that suggests that people will assign personality, emotions, intentions and objectives towards machines no matter how complex or simple they might be. This process is affected by various factors including an individual’s background or culture. In order to successful develop robotic solutions for a highly complex domain such as care for older people, it is necessary to take into account these issues so that the end user is confident and capable of interacting and engaging with the technology.
In order to demonstrate the importance of the role of embodiment in regards to assistive robots for older people, Mataric (2005) explored developing prototypes of embodied technology and compared their effectiveness against disembodied equivalents. The Clara assistant nurse robot described in the study is capable of locating a hospital room, bed and patient for the purpose of spirometry. This task is performed by nurses for patients recovering from heart surgery, who are required to undertake breathing tests in order to monitor regaining lung functionality and to avert infection. Clara is also required to perform several monitoring tasks associated with this function and interacts with patients via speech. Clara has been designed to describe the spirometry task to the patient, provide feedback, encouragement and report its findings to staff within a hospital. The robot is capable of performing these tasks amongst patients based on their preferred personality and mode of interaction in order to personalise the experience. Mataric also researches the benefits of non-contact socially assistive robots for post stroke rehabilitation therapy that includes monitoring and encouraging exercise as well as providing feedback. Results are positive with stroke patients increasing the amount of exercise as a result of interacting with the system.
Mataric argues that a greater understanding of the role of embodiment is required for the design and development of assistive robots for older people on the grounds that this user group may be more technologically disinclined compared to younger users. Furthermore Mataric also argues that more research is required with respect to different types of behaviour in relation to the embodiment, with different users groups. Again this is a crucial area of study in regards to older people in order to establish credible human-robot interaction for providing care service for this user group.
Related to the behaviour of assistive robots, the degree of autonomy and supporting roles that require the user to engage with the system in order to achieve a desired objective raises a variety of ethical issues that must be considered. These include vulnerability in the presence of an assistive robot.
Ongoing research conducted by Mataric (2005) also includes modelling empathy. Empathy is a key aid used by healthcare professionals in regards to providing care for older people. Again this raises ethical issues related to an older person’s vulnerability, and the extent to which they can be manipulated on account of cognitive decline.
Ina study aimed to identify useful tasks for robots to assist residents of a retirement village, and preferences for the appearance and features of healthcare robots, Broadbent et al. (2009) used a mixture of questionnaires and interviews.Two sets of questionnaires were issued to both staff and residents at retirement village and this was followed up by an interview to discuss preference for the robot’s colour, shape, design and size. The questionnaires focused upon which tasks a robot could be designed to assist with and general attitudes towards healthcare robots.
In general residents’ responses were more positive towards robots compared to staff. This information was gathered using a PANAS (positive and negative affect scale) scale. From a questionnaire, 30 tasks were prioritised from most useful to least useful using an average of the participants’ responses. Those tasks rated highest by the residents were detecting falls and calling for help, switching on and off appliances, cleaning, making phone calls to a doctor or nurse, lifting heavy things, monitoring the location of people, and reminders to take medications.
The results indicated that residents prioritized healthcare related functions while staff prioritized tasks related to their care roles. Participants were also asked to assess their preferred embodiment and shape of a robot using the images below:
Figure 1Eight different shaped robots images used in Broadbent et al.’s study
The participants showed a preference towards a robot with a screen held on the body compared to a screen on the head. There were no major differences between a humanoid or rounded, box shaped embodiments.
In regards to the height of the robot, 1.25m was the preferred height chosen by the participants. Both carers and residents stressed the need for the robot to be non-intrusive yet tall enough to perform bed related tasks. Bright colours were preferred with silver as the most popular followed by gold and white.Participants were also asked to choose between two robots for the following tasks: Medication reminding, companionship and blood pressure. Over 70% of the participants chose Hopis (left in the picture below) for companionship and over 70% of the participants chose In-touch (right in the picture below) for health related functions. Comments from the participants in regards to Hopis, revealed that the participants believed the older people would pay more attention to a soft touch “childlike” robot. Comments in regards to the In-touch robot indicated that the participants believed such a robot would be more capable at performing healthcare related functions.