Older adults are increasing in Europe[1]. Technology may play a role in older adults' health. Giving technology the ability to be physically manipulated can result in creating trust[2] . According to Spreicer [3], co-designing tangible user interface with elderly people can increase the chance of technology acceptance. NESTORE [4]is a virtual coach that comes in diferent forms of modalities/interfaces: a chatbot, a mobile application and a tangible coach. This article focuses on presenting one of the interfaces of NESTORE: the “tangible coach”. The tangible coach is an embodied conversational agent that seeks to empower older adults in becoming the co-producers of their wellbeing lifestyle through its tangible and its vocal interaction.
The well-being of older people will increasingly depend on technology [
NESTORE[
The NESTORE tangible coach consists of building trust in order to become a friend, empathy with older adults in order to become a companion while e coaching them for a healthier lifestyle[7]. the journey. In order to that we need to answer some problematic: • How can the physical design and direct manipulation through “touching” the e-coach builds a form of bonding and companionship with older adults? • How can the tangible coach vocal capability and tangible capability motivate older adults to become his or her friend and build trust? • How can tangible interaction with vocal capability improve or maintain the wellbeing of older adults?
The tangible coach has an oval shape, is white, and made of 3D printed plastics with a textile cover for a cozy and emotional feel to the touch, supposed to encourage tangible manipulation. The aim was to create a feeling of companionship (Q1). The tangible coach has its main face which is when the device is positioned on its side. This oval face is the main part of this body which looks like a human face. Users can communicate with it through its diferent light models. His human face expresses a specific act. Such acts are "the tangible coach who listens" or "the tangible thought of the coach" or "the tangible coach who speaks". These acts are represented by LED patterns. This type of interaction is what we will call expressive interaction, which means that the tangible coach will convey his data by expressing it with his interface(Q1). The tangible coach has two main states: the state of sleep and the state of wakefulness. To switch between these two states, the user consists of touching the tangible coach, and changing the position of the tangible coach. In fact, the sleep state is when the user puts the small base up and the large base on the table, as shown in Figure 1a). In the sleep state, the tangible coach does not interact with the user and cannot listen, think, or speak. During our co-design process with the user, we concluded that older people care a lot about data privacy and prefer the tangible coach not to listen to them all the time or talk to them without permission(Q2). In the awake state, that is to say when the device is positioned on its side as illustrated in Figure 1 b). As it seems, the tangible coach is awake, waiting for the user to call (the wake-up word) and start speaking. The awake state consists of three substates (or acts): the listening substate, the thinking sub-state, and the speaking sub-state, as illustrated in the diagram in Figure 3. Each sub-state is designed with LED patterns and colors to show the user what sub-state the device is in. In other words, when the tangible coach is listening (sub-listening state), the tangible coach shows an orange circle on its front base for 10 seconds. Then the trainer will start showing an orange and white colored LED pattern rotating around the front base circle indicating that he is thinking about an adequate answer depending on the question. Finally, once the tangible coach has their response ready, the LEDs will begin to flash orange in a circle shape while speaking the response. Figure 2 shows the LED model of the tangible coach during the listening sub-state (a) and the thinking sub-state (b).
The tangible coach is based on Google Cloud speech to text and text to speech [8]. It focuses on a natural conversation with users. The whole logic and scenario of the conversation are processed with RASA, a natural language processing framework that detects the intentions and intentions of a text. This will lead to imitating a human to human conversation. Plus, the turn of the conversation is two-sided. The user or the tangible coach can initiate the conversation. Voice interaction is used with the support of visual information displayed via LEDs. The LEDs generate suficient light that can provide the information anyway even if the user is not looking directly at the device, harnessing peripheral attention, allowing multitasking to be supported. Finally, the user can still talk to the device even if it is not in the line of sight and continue to interact with it while performing other activities. This kind of behavior aims to show the user that the system can be talked to and thus build a feeling of trust(Q2) which is found usually among friends. The conversational agent can cite the user’s planned and recommended activities, the user’s score with comments, notifications, and congratulatory messages (Q3).
In order to answer our research questions, the tangible coach is tested on a three-month basis.
For each problematic, we will evaluate the tangible coach through questionnaires or logging
through our system. Q.1 How can the physical design and direct manipulation through
“touching” the e-coach builds a form of bonding and companionship with older adults? To answer
Q.1, we will evaluate the use of this embodied tangible coach with the Companion Animal
Bonding Scale questionnaire [9] to assess the level of companionship the tangible coach has
given to the user. Q.2 How can the tangible coach vocal capability and tangible capability
motivate older adults to become his or her friend and build trust? To answer Q.2, we will assess the
friendship between the coach and the user by using the friendship scale questionnaire [10] and
by logging from the interface and the backend [
NESTORE tangible coach aims to use its tangible capability and vocal capability to create empathy, trust and coaching older adults to have a healthy lifestyle. The system will be tested in three European countries (Spain, Italy and the Netherlands). The questionnaires will show whether the sense of companionship, friendship and coaching are fulfilled.
This project is supported by the European Commission under the Horizon 2020 programme, SC1-PM-15-2017 – Personalised Medicine topic, through the project grant N.76964. The authors want to thank all the NESTORE Consortium. The opinions expressed in this paper are those of the authors and are not necessarily those of the project partners or the European Commission. [7] M. El Kamali, L. Angelini, M. Caon, D. Lalanne, O. Abou Khaled, E. Mugellini, An embodied and ubiquitous e-coach for accompanying older adults towards a better lifestyle (2020) 23–35. [8] Google cloud speech to text and text to speech, https://cloud.google.com, 2020. [9] R. H. Poresky, C. Hendrix, J. E. Mosier, M. L. Samuelson, The companion animal bonding scale: Internal reliability and construct validity, Psychological Reports 60 (1987) 743–746. [10] G. Hawthorne, Measuring social isolation in older adults: development and initial validation of the friendship scale, Social indicators research 77 (2006) 521–548. [11] S. Jowett, N. Ntoumanis, The coach–athlete relationship questionnaire (cart-q): Development and initial validation, Scandinavian journal of medicine & science in sports 14 (2004) 245–257.