Sage Journals: Discover world-class research

Abstract

Virtual coaching has emerged as a promising solution to extend independent living for older adults. A virtual coach system is an always-attentive personalized system that continuously monitors user’s activity and surroundings and delivers interventions – that is, intentional messages – in the appropriate moment. This article presents a survey of different approaches in virtual coaching for older adults, from the less technically supported tools to the latest developments and future avenues for research. It focuses on the technical aspects, especially on software architectures, user interaction and coaching personalization. Nevertheless, some aspects from the fields of personality/social psychology are also presented in the context of coaching strategies. Coaching is considered holistically, including matters such as physical and cognitive training, nutrition, social interaction and mood.

Keywords

coaching intervention coaching strategy context-awareness interface modality self-monitoring

Introduction

Most of the previous attempts to define virtual coach systems (VCSs)¹ have been so generic that many systems, devices or programmes could be considered. We share Siewiorek’s² more demanding VCS definition, presented in the abstract of the article. This work contributes by:

Collecting the state-of-the-art of technological and methodological approaches for a VCS for older adults. Information about the systematic review can be found in the supplementary material.

Presenting the design space of a VCS (Supplemental Figure 1).

Proposing implementable novel approaches and categorizing them within the proposed design space (Supplemental Figure 2).

Considering validation as a crucial part in the development of a VCS. Last section reviews the different methodologies.

VSC components

This article contributes with a proposal of the design space of VCSs. The purpose is to provide a common design space to refer to in the forthcoming publications.

This analysis started following Ding et al.’s¹ work, which defined the components of virtual coaching interventions (Supplemental Figure 3). According to Siewiorek’s² definition, a VCS must be able to deliver interventions. Ding et al.¹ defined four components for virtual coaching interventions: self-monitoring, context-awareness, interface modality and coaching strategy. These components respond to the questions of what message to deliver and when and how to intervene the user.

This article compared Ding et al.’s¹ components with literature, to confirm and validate their presence within the design space of a VCS. Bibliography reviews (Supplemental Tables 2 and 4 to 7) confirm components’ presence in the design space of a VCS. Nevertheless, the bibliography compiled in this article (Supplemental Table 6) points out the convenience of adding a fifth component to the design space. It will be called coaching intervention, and it defines the purpose of the intervention – that is, suggesting an activity, reminding something and so on. Supplemental Figure 1 represents the design space of VCSs defined and proposed in this work, as well as the dependencies between the components. VCS’s five components are:

Self-monitoring: sensor-based technology for automatic user and environmental data collection.

Context-awareness: computational component that infers contextual information from user-related data, extracted by self-monitoring and by integrating additional information through external databases.

Interface modality: interaction method between the user and the VCS.

Coaching interventions: short-term actions of the virtual coach following a coaching strategy, for example, reminding medicine intake or suggesting physical training exercises.

Coaching strategy: guidelines for optimal user-tailored coaching, dynamically adapting to user’s need, performance and progress. That is, it personalizes and adapts the coaching message considering the contextual information, which includes self-monitoring data, and the interfaces and coaching interventions available.

Self-monitoring

This component uses sensor-based technology to automatically collect data from the user and her or his environment, objectively and comfortably.

Self-monitoring has some limitations for subjective data collection. Self-report data is the best choice in these cases. However, if possible, self-monitoring should be used for data extraction (see Supplemental Table 3). Information and communications technology (ICT) platform can be used for the evaluation of self-monitoring devices.³

Sensor-based technology can be integrated into a wearable device or placed in an environment’s surroundings creating a sensorized environment. This article follows a technological main categorization, instead of user-centred.⁴ Here, a wide variety of sensors are considered. Supplemental Table 2 compiles those used in VCSs.

Wearable devices

Wearable devices or wearables refer to electronic devices that can be worn. They are mostly small, lightweight and discreet, pretending to be an accessory, for example, wristband,⁵ or being hidden under clothes, for example, chest belt.⁶ There are different packaging formats available to address diverse user preferences, including headsets.^7,8

Older adults could perceive an added value in their use.⁴ However, wearables’ major feature is their ability to connect to the Internet and permit network-device data exchanging – Internet of things. Despite their benefits, wearables’ usage implies some drawbacks, such as user’s discomfort while wearing, recharging, data privacy concerns, indirect data access, black-box-like algorithms, data standardization issues and unverifiable wearable usage.⁴

This section categorizes wearables, depending on their purpose of usage, into affective, cognitive, physical and clinical. The different categories of wearables may overlap since they all monitor physiological metrics in the end.

Affective wearables

Affective wearables can recognize user’s affective patterns, for example, expressions of emotion, a strained voice and a change in autonomic nervous system activity.⁹

Application examples include lifelogging technology¹⁰ and estimation of stress,¹¹ sleep and activity levels.¹²

Cognitive wearables

Cognitive wearables measure user’s cognition, the mental process of understanding through thought, experience and senses.

Automatic measurement of the cognitive capacity is a difficult task. Cognitive Kit¹³ combines self-monitoring with self-reporting, with positive results in major depressive disorder cases.¹⁴

Physical wearables

Physical wearables¹⁵ measure physical activity through direct activity data extraction, vital signs’ analysis or both.^16,17 Some are smart watch,¹⁸ smartphone’s accelerometers monitor daily physical activity of older adults,¹⁹ localization through global positioning system (GPS),²⁰ Bluetooth, Wi-Fi or Ultra-Wideband.²¹ A promising solution approach is Wi-Fi fingerprinting.²²

For validation of extracted data, ActiGraph GT3X is the gold standard, being capable of monitoring daily life activities of different age groups.¹⁹

Clinical wearables

These wearables provide clinically relevant information, usable for diagnostic or predictive purposes. Applications for older adults can describe user’s well-being through, for example, sleep quality,¹² body balance or frailty.²³

Body balance information allows evaluating older adults’ ability to carry instrumental activities of daily living and developing fall detection and prevention responses.²⁴

Frailty serves in older adults as a prognostic factor of disabilities, institutionalization/hospitalization and mortality.²⁵ Technological approaches focus on Fried’s clinical test²⁶ to measure frailty through gait parameters, physical activity, strength and balance.^25,27 Cognitive frailty should be measured, as well.²⁸

Sensorized environments

Sensorized environments are spaces where sensors have been deployed, for example, house. There are two types: general purpose and smart furniture and objects.

General purpose home sensors

They gather four data types: physical, clinical, affective and cognitive.

Physical data gathering

It is mainly carried out through computer vision approaches. Microsoft Kinect²⁹ tracks user’s presence and motion. Kinect’s manufacturing has been abandoned but its detection capacity has demonstrated that similar alternative technology can be developed, for example, OpenSim.³⁰

Acoustic sounds retrieved through a wireless sensor network provided footstep location.^31,32

Some research focuses on fall detection.²⁴ Gait pattern variations can be correlated to physical, mental or cognitive health variations.^30,31

Clinical data gathering

Approaches can sense frailty, pollution, smell, epileptic seizure and so on.^25,33
–35

Affective data gathering

Emotion recognition capacity is vital for an empathic VCS. Humans express emotions through verbal – speech and writing – and non-verbal signals – hand and body gestures and facial expressions.^36
–38 Emotion recognition accuracy will improve when exploiting data from various expression ways.³⁹

Accompanying people’s face identification improves tracking older adults’ social life.⁴⁰

Cognitive data gathering

Older adults can suffer a transitional stage between normal cognitive aging and dementia, called mild cognitive impairment. It can hinder the completion of complex daily tasks.⁴¹ Dawadi et al.⁴¹ used infrared motion detectors and magnetic door sensors, extracted 35 activity-related features and used them to feed the machine learning algorithms.

Selker et al.⁴² developed a system for cognitive decline assessment through walking times, walking speed and activity amount.

Smart furniture and objects

These are employed to gather user’s both physical – activity level – and clinical information, for example, sleep, nutrition, snoring, heartbeat and oximetry.

Smart furniture

Detected smart furniture includes mostly tables, then chairs and finally individual cases of mat, mattress, floor, shelf, sink and board.

Smart tables’ application involves nutritional intake monitoring,⁴³ multiple objects’ identification and tracking⁴⁴ and monitoring the use of the table.⁴²

Smart chairs monitor stress levels – via heart and breathing rates⁴⁵ – or activity level – through body weight, body temperature, heart rate, blood oxygen saturation and blood pressure.⁴⁶

Smart mats and mattresses monitor sleep quality through respiratory rate and heart frequency.⁴⁷ Andoh et al.⁴⁸ adds snoring and body movements.

The smart floor⁴⁸ tracks location, the smart shelf⁴⁹ monitors in-stock products, the smart sink⁵⁰ controls temperature and water flow and the magnetic board⁵¹ tracks physical activity accessories’ usage.

Smart objects

Smart objects are varied such as toothbrush and mirror,⁵² cups,^53,54 a bottle,⁵⁵ a spoon⁵⁶ and a walking stick.⁵⁷

Context-awareness

Context is crucial in VCSs. First, it can reduce the complexity of the problem to solve, by decreasing user choices and reducing the search space. Second, it allows adapting to the user’s state.

Context is captured through self-monitoring and external databases – weather, electronic health record (her) and so on. Context is based on information about the current status of entities – people, places, things or devices – in the environment, including the location, time, activities and preferences of each entity.⁵⁸ Eventually, it is proven that a user’s responsiveness to an intervention highly depends on her or his current tasks, in addition to other factors such as emotional state and the modality of the intervention.⁵⁸ Context-awareness means that one can use contextual information.⁵⁹ Context-aware systems are characterized by their capacity to extract, interpret and use contextual information for information and services’ provision and to adapt its functionality to the current context of use.^58,59 Abowd et al.⁵⁸ defined the core context-aware features that context-aware applications may support (1) presentation of information and services to the user, (2) automatic execution of a service and (3) tagging context to information for later retrieval.

Main components of a context-aware system are:

Context providers and context-aware services⁶⁰: context providers acquire data from different sensors through direct sensor access, middleware or context servers.⁶¹ Context-aware services provide relevant services based on acquired contextual data. The concept of service integration can be constructed in a service-oriented architecture.^62,63

Context-aware computing: a ubiquitous software component that adapts the system’s functions depending on the context information.⁶⁴

Context modelling: the formulation of a system that interprets and understands several system nuances.⁶⁵ Nykänen and Rodriguez⁶¹ detected the major strategies for context modelling, including key-value, object-oriented and ontology-based models. There is not a commonly agreed standard for modelling, but ontology-based models seem to offer many desirable properties such as dealing with incomplete or partially understood information and domain-independent modelling.⁶⁶ Standard representation of context data in context-aware systems can be built upon semantic web metadata standards, such as Resource Description Framework (RDF) and OWL.⁶⁷ There are many modelling approaches: SNN,⁶¹ SOCAM,⁶⁵ ASSL⁶⁵ and Process Mining.⁶⁵

Despite the technological advancements and research results, the realization and the role of complex context-aware computing are not clear in application ecosystem(s).⁶¹

Literature collects different attempts to define context.⁵⁸ Loureiro and Rodrigues⁶⁸ provided the easiest definition where context refers to any information that characterizes the situation of an entity – that is, a person, place, or object relevant to the user–applications interaction. For Loureiro and Rodrigues,⁶⁸ the primary context pieces for characterizing a situation of an entity are location – where, identity – who is the older adult with, activity – what and time – when. Primary context types act as indices into other sources of contextual information to find secondary context pieces for that same entity, for example, email address. Secondaries are obtained by combining different primary context types and/or integrating external database information, for example, weather. Supplemental Table 4 shows how context has been used in VCSs.

Interface modality

Interface modalities define the interaction path between the user and the system. Information is presented to users through interface devices and user interfaces (UIs). The presentation method strongly impacts the communication interaction and may affect the usability and user’s adherence towards the system.

Following review provides a broad point of view about interface modalities. Supplemental Table 5 collects which and how have been employed in VCSs.

The UI is the space where bidirectional human–machine interactions occur: from human to machine (H2M), for example, to control the system or from machine to human (M2H), for example, to provide information to the user. UIs are supported by interface devices. These can be varied, for example, smartphone and VR headsets, providing different possibilities to adapt to user preferences and system requirements.

Older adults’ adherence will increase if VCS’s interface modality is intuitive. Interface devices’ usage intuitivity can increase by integrating context-aware computing techniques. For example, handheld interface devices as smartphones adjust the UI design in real time to the orientation of the visualization screen. Natural user interfaces (NUIs) are the intuitive UI option. NUIs replace device-based control by gestural movements, promoting direct, intuitive and natural interaction ways.⁶⁸ Multi-touch and gestural computing are among the most often used technologies for implementing NUIs.⁶⁸ Speech and handwriting are additional examples. Part of our natural interaction manner with our environment is through senses: sight, hearing, touch, smell and taste. Based on the NUI concept, UIs can be categorized as vision-based graphical UIs (GUIs), audio-based voice UIs (VUIs) and sound UIs (SUIs), tactile UIs (TUIs), olfactory UIs and gustatory UIs.

GUIs provide an immediate visual feedback of user’s actions, facilitating the learning stage. GUIs’ advantages’ exploitation strongly depends on the interface device employed. Innovative GUIs based on VR, AR or holograms may require obtrusive interface devices: headsets, smart glasses or projectors. So, tangible UIs⁶⁹ are recommended.

VUIs consist of the voice application system, in charge of allowing speech-based interactions: Mayer et al.⁷⁰ showed that participants perform better when the voice is human rather than machine generated.

SUIs rely on human’s capacity to extract passively information from sounds and associate it to actions. For example, when we hear the alarm bell in a building, we could relate it to a fire evacuation. Bickmore et al.⁷¹ recommends not using impolite audio alert tones. In contrast, text warning messages can be easily ignored.⁷²

TUIs centre on haptic feedback, which include pressure, vibration, heat and pain⁷³: thermal stimulation for nervous system responsiveness quantification⁷⁴ and sensory substitution.⁷⁵ Sensory substitution is useful for people suffering from balance disorders and loss/reduction of visual, vestibular and somatosensory information. Smartphones might also serve this purpose.

Olfactory and gustatory UIs may be the most difficult and complex ones to develop since humans depend on chemicals to experiment these senses. Research on these two interface modalities was boosted by VR applications, looking for innovative ways to enhance immersion feelings. Olfactory sense stimulation can be used for warning interventions, for example, food burning, or lifelogging applications,⁷⁶ gustatory UIs to stimulate appetite in older adults.⁷⁷

Middleware interfaces are not within the scope of this review. Some commercial and open source approaches for IoT devices are OpenIoT,⁷⁸ MiddleWhere,⁷⁹ Hydra,⁸⁰ FiWare⁸¹ and Oracle Fusion Middleware.⁸²

Animated agents

They provide a mixed UI modality. Animated agents create stronger user-system bonds, increasing adherence. Often, they act as virtual coaches.

Agents’ embodiment varies from geometrical shapes to human-like characters, including holograms.⁸³ Agents’ appearance and expressiveness relies on user preferences. For example, a 3D human agent with blinking eyes is rated as more intelligent than geometric shapes or caricatures.⁸⁴

Animated agents can have non-verbal and/or verbal conversational capacity. Non-verbal is expressed through facial expressions and gestures. Non-verbal expressiveness and appearance define emotional and relational skills, increasing adherence.⁸⁵

Verbal animated agents express text-based or voice-based messages. A talking face increases user adherence.⁸⁶

Regarding non-verbal communication, opinions are varied: Moreno et al.⁸⁷ states that non-verbal communication does not improve UIs and Atkinson⁸⁸ defends that non-verbal presentation enhances learning.

Coaching interventions

This section describes a new component proposal, not considered by Ding et al.¹ The literature review compiled in VCS state-of-the-art (see Supplemental Table 6) and this section provides a formulation for this contribution. Coaching interventions refer to short-term actions delivered through intentional messages. They define the purpose of the message, for example, reminding something.

User’s interruption level and intervention’s urgency level can be estimated through activity-related contextual information.⁸⁹ That is, there is no need of delivering a medicine-intake reminder while the user is talking on the phone, for instance. The intervention could be postponed until the user hangs up the phone.

Based on literature analysis, this article classifies interventions as reminders, warnings, recommendations, motivationals and teaching. That is, VCS’s behaviour has two main functions: assisting – through reminder, warning or recommendation interventions – and empathizing with the user – via motivational interventions.

Reminders

Reminders reinforce the cognitive abilities of users. Older adults with cognitive impairment such as memory loss would be benefitted by an adaptive software-based personal reminder system. Some reminder approaches go beyond the context-based reminder delivery by estimating parameters such as forgetting probability of the user and the reasons for forgetting.⁸⁶ Zhou et al.⁸⁹ reviewed different factors that could influence the design and framework selection for reminder systems. Reminders are classified into two groups⁹⁰:

Scheduling aids: for personalized daily fixed time activities’ prompting. Different information can be used for triggering delivery⁸⁹: time, location, activity and complex context. An application example is a medicine intake reminder system,⁹¹ which tells when and how to take the medicine.

Instructional cueing aids: for subjects’ guidance through sequential processes that once were known but are currently forgotten.

This article contributes with an additional category:

3. Activity accomplishment aids: for accomplishment of forgotten initiated daily tasks that remain unfinished. Unlike scheduling aids, activities are not subject to fixed times. Application examples include a smart reminder system (SRS) for in-home activities⁸⁶ and a gate reminder,⁹² a home system in the front door that screens things users need to take before leaving home.

Warnings

Warnings announce detected precarious situations, for example, when the oven remains turned on after use. For example, Boulos et al.⁹³ developed a wearable that communicates with caregivers in case of fall detection.

Recommendations

Recommendation interventions offer suggestions concerning a non-urgent task that can cause, somehow, a negative impact on the user.

When defining these interventions, conclusions extracted from a workshop can be useful⁹⁴: (1) delivery time must be adaptable to older adults’ routine activities’ time and (2) older adults require different recommendation types depending on the daytime. Hammer et al.⁵¹ differs at least two types: daily routine-targeted and stimulators. First ones suggest the best time and/or place for a daily activity task. Stimulators boost activities out of the daily agenda, for example, meditation exercising. For computational implementation, tools such as LensKit recommender toolkit⁹⁰ and Drools⁹⁵ could be used.⁹⁶

Motivationals

Motivational interventions foster positive behaviour by cheering and guiding the user to overcome challenging situations.

Behaviour change inducement requires the employment of evidence- and psychology-based behaviour change strategies, discussed in the next section. This explains why motivation is categorized according to the self-determination theory (SDT)⁹⁷:

Extrinsic: drives behaviour changes through external rewards, for example, scoring or social recognition. Gamification or serious games are an easy example where external rewards are used. During implementation, it is relevant distinguishing different reward types, for example, there are eight types in gamification,⁹⁸ and rewarding methods, for example, De Caravalho and Ishitani⁹⁹ compiled motivation guidelines for serious games for older adults. Older adults’ preferences should be also contemplated. For example, Herpich et al.⁹⁴ collected preferences about rewarding the accomplishment of lifestyle recommendations.

Intrinsic: refers to behaviour change driven by internal rewards. For example, the satisfaction of taking self-care. P-creativity¹⁰⁰ is an innovative implementation method. It employs curiosity in novel and valuable concepts/designs/artifacts as a motivational engine.

Grace et al.¹⁰⁰ developed an intrinsic motivation-boosting engine in a framework called personalized curiosity engine (PQE).

Concluding, intrinsic motivation is the core point for healthy aging promotion.

Teaching

These teach users new skills. Older adults may face new challenges while they age. Learning how to face them could be a key factor to ensure their well-being. Skills could include stressors’ management and education in training exercises or clinical awareness.

Coaching strategy

Coaching strategy states the plan of actions or guidelines to maximize user’s potential¹⁰¹ and offer context-aware personalized coaching. The guidelines are grounded on contextual information, coaching interventions and available interfaces. This component can be identified in any VCS. Nevertheless, the implementation methods are diverse. This section focuses on the reasoning and decision-making ability requirements of VCSs. These can be provided through four strategies: clinical decision support systems (CDSSs), ecological momentary assessment (EMA) and intervention (EMI), behaviour change theories (BCTs) and models (BCMs) and clinical guidelines.

CDSS, EMA and EMI are more technology-centred. EMA and EMI are complementary, so usually they are not independently understood or explained. Remaining two strategies focus on action plan guidelines’ definition. These strategies provide adaptivity to the VCS. EMA allows assessing the user in near real time. BCTs, BCMs and the clinical guidelines are employed to define the knowledge and evidence-based action plans. A CDSS is the decision-making engine, which provides a recommendation considering user-related data and the possible action plans. By last, EMI permits intervening on the older adults near to real time.

The four strategies can be implemented into the same system. Supplemental Figure 4 reflects the relationship between them. Supplemental Table 7 summarizes strategies in VCSs.

Clinical decision support systems

A CDSS is a type of computerized information system for supporting decision-making tasks.¹⁰² According to Sim et al.,¹⁰³ a CDSS is a software for aiding in clinical decision-making processes: it matches computationally patient’s characteristics with computerized clinical knowledge base, to provide tailored recommendations or assessments to the user. That is, a CDSS relies on its clinical knowledge and reasoning ability. This converts CDSSs into a safe, reliable and efficient approach for in-home self-care. Pursuing an empathic and encouraging virtual coach, the CDSS may help the coach intervene using the appropriate intervention and interface depending on the context. For example, use a softer voice when the user is sad or remind picking up the laundry through visual interfaces if the user is talking on the phone. In addition, CDSSs may provide personalized healthcare recommendation interventions to the user, for example, physical exercises. Some articles omit ‘clinical’ and use DSS.

Holsapple¹⁰⁴ presented a generic architecture for CDSSs. Marschollek¹⁰⁵ presented a state-of-the-art in CDSSs for in-home medical healthcare or patient support. It defined a set of requirements and properties to assess and categorize them. We consider that some requirements should be added: real-world data (RWD)-based training, big data management capacity and considering user preferences and context.

Marschollek¹⁰⁵ did not consider these features: dynamism^102,103 and intelligence.¹⁰⁶ Dynamism refers to CDSS’s necessity to adapt to users’ dynamic behaviour and environment, that is, context-awareness. Dynamic changes, for example, mood, must be detected and updated to provide personalized recommendations. Intelligence pursues dealing with uncertain environments where new knowledge must be discovered through approximation, reasoning or intuition.

Domain-specific analysis is recommended as an optimal computation strategy. Concerning new knowledge generation, data mining (DM) and artificial intelligence (AI) techniques are very recurrent.¹⁰⁶ AI is involved with the intelligence of the CDSS while DM with both intelligence and dynamism.

An additional feature could be the inclusion of visual analytics (VA). VA has the advantage of providing simple data and/or pattern views, as well as supporting visually decision-making activities.¹⁰⁷

Ecological momentary assessment and intervention

They offer the capability to monitor variations in user’s everyday life and deliver behavioural responses in the appropriate context. EMA methods involve collecting real-time data, on repeated occasions and in context of daily life.¹⁰⁸ EMIs refer to delivering interventions for coping with everyday lives in real time and in the real world.¹⁰⁹

These techniques are ecologically valid because they are based on the real-world environment (RWE). EMA captures data directly from the RWE, through automatic approaches or manual self-report. EMI delivers interventions in the most appropriate moment of the subject’s everyday life.¹¹⁰ Data capture via EMA runs parallel to the use of EMI.¹⁰⁹ Momentary indicates they are both focused on the user’s current situation.¹¹⁰ In 2010, Heron and Smyth¹⁰⁹ stated that real-time assessment and intervention proceedings were considered separately. Nowadays, the combo of EMA and EMI has become an important research area. First, EMA’s results can inform and tailor the content and timing of EMI delivery, to be the most appropriate.¹⁰⁹ Heron and Smyth¹⁰⁹ states the importance of interventions with a certain level of tailoring. Second, EMA can trigger EMI if problematic data information is detected.¹⁰⁹ This helps addressing behavioural guidance, for example, self-care activities’ delivery or instructions for medication adherence, or other interventions, for example, relaxation stimuli, when most needed.¹⁰⁹

EMA and EMI implementation imply facing three main limitations¹¹⁰: (1) defining the set of questions to capture information data, (2) decide in which context to deliver the questions and (3) managing the EMI strategy.

Behaviour change theories and models

The strategy to change a person’s lifestyle consists on the application of evidence-based BCTs and BCMs. BCTs attempt to explain why behaviour changes by extracting the impact factors. BCMs address the behaviour change processes, so they allow predicting behaviours.

de Vries¹¹¹ referenced Lim and Dey¹¹² book, where they stated that 1700 constructs were used in 83 theories. de Vries¹¹¹ highlighted some famous health behaviour theories and models, such as health belief model (HBM),¹¹³ social cognitive theory (SCT),¹¹⁴ the theory of planned behaviour (TPB)¹¹⁵ and the transtheoretical model (TTM).¹¹⁶

Integration of behavioural theories can be seen as an advantage by some people, such as reinforcing a theory by overcoming its weak points and by adding new elements.¹¹¹ In integration, unlike in combination, elements can only be added if they offer a supplementary theoretical and empirical value and if their addition will lead to develop and test new hypotheses. de Vries¹¹¹ presented an interesting health behaviour theory integration approach named integration-change model, where all of the previous theories – HBM, SCT, TPB and TTM – are integrated.

BCT Taxonomy v1 (BCTTv1)¹¹⁷ might be helpful for implementation. This is a cross-domain, hierarchically structured taxonomy of 93 distinct BCTs with labels, definitions and examples.

Clinical guidelines

Clinical guidelines state the action programmes in clinical practice.

Only 1 out of 19 VCSs indicated using clinical guidelines.⁷² The VCS was about educating powered wheelchair users about the usage of seating functions, including prescribed personalized seating regimens.

Due to the lack of information about clinical guidelines in the VCS-related articles, Supplemental Table 8 proposes more implementable guidelines from different clinical knowledge domains: pharmacology, multiple comorbid diseases, diabetes mellitus, depression and falls.

Validation

This section and Supplemental Table 9 review the relevant key factors involved in validation. It includes information related to subject population – number of sample and demographic data – and validation methodologies including the use of control groups.

Four out of nineteen articles did not include information about validation.^118
–121 These are excluded from Supplemental Table 9.

Subject population

Studies’ subject population number varies considerably from 2 to 70. Blanson Henkemans et al.¹²² had 118 participants but only 35 completed the whole process. Kyriazakos et al.¹²³ did not indicate the number of participants, but only said that next study would be larger up to 200. So, without considering,¹²³ the average number of subjects is 32. Differences could be justified by the validation methodology followed. For instance, Wu et al.⁷² evaluated VCS’s usability with five participants. According to a study, 80 per cent of usability problems can be detected by the first four or five subjects.¹²⁴

General information about participants, for example, gender, age, ethnicity, education and so on, was extracted by demographic questionnaires in previous studies.^125
–127 The sample of responders might not be an accurate representation of the population to address.

Validation methodologies

Only a few validation protocols included control groups.^{125,128
–131} The rest of the studies analysed participants’ progression to evaluate the VCS.

Among validation methodologies, questionnaires are the most frequent approach – 12 out of 19. Some studies may have used more than one.^{123,126,132,133} Questionnaires can be grouped by whether they assess the VCS itself (Sullivan and Artino¹³⁴ in Asselin et al.,¹³⁵ Laugwitz et al.¹³⁶ and Davis¹³⁷ in Kyriazakos et al.¹²³), participants’ progress (Harris and D’Eon,¹³⁸ Cleeland and Ryan¹³⁹ and Hoehn and Yahr¹⁴⁰ for Parkinson in Ellis et al.,¹³² Folstein et al.¹⁴¹ in French et al.¹¹⁹ and Kyriazakos et al.¹²³ and Thomas et al.¹⁴² in Bickmore and Brown¹²⁶ and Lins and Carvalho¹⁴³ in Kyriazakos et al.¹²³) and participants’ subjective data (Horvath and Greenbargh¹⁴⁴ in Bickmore and Brown¹²⁶ and Lessiter et al.¹⁴⁵ in IJsselsteijn et al.¹³³).

Alternative approaches to questionnaires included interviews^123,146 and self-reported data usage^130,122,129 methods. There were also some objective data-gathering methods for validation. Objective data was used for physical capacity evaluation in Kyriazakos et al.¹²³ and Buttussi et al.¹²⁷

The wide variety of validation methodologies states that there is eventually not a standard validation protocol. Validation methods were designed for determined VCSs. Another open issue in validation could be related to the independent contribution of the self-monitoring approach and the VCS.

Discussion

Nowadays, there is a paucity of literature concerning VCSs – 19 articles, especially for older adults – 5 articles. However, an increase is expected in the coming years, thanks to the European H2020 programme¹⁴⁷ and the funded projects.^148
–153

The lack of a common design space could lead to misunderstandings between researchers. This article proposes a literature-based design space and defines a structured VCS framework, to provide common guidelines for posterior publications about this topic.

In order to situate researchers in the current state of VCSs, a state-of-the-art is provided. Information about the VCS components was extracted and presented in review tables. However, articles excluded for not fulfilling the definition used in this survey are referred to in the core of the section. In addition, validation approaches are analysed.

Future work involves testing the applicability and extrapolation of the proposed design space in future VCSs^148
–154 and surveillance of their progress and discoveries. Main challenges include the definition of methodologies to select the right method to develop each design space component and interconnect them adequately.

Supplemental Material

Final_revision_SupplementaryMaterial_Survey_on_virtual_coaching_for_older_adults.cleaned – Supplemental material for Survey on virtual coaching for older adults

Supplemental material, Final_revision_SupplementaryMaterial_Survey_on_virtual_coaching_for_older_adults.cleaned for Survey on virtual coaching for older adults by Nerea Lete, Andoni Beristain and Alejandro García-Alonso in Health Informatics Journal

Supplemental Material

SupplementaryMaterial_Survey_on_virtual_coaching_for_older_adults.cleaned – Supplemental material for Survey on virtual coaching for older adults

Supplemental material, SupplementaryMaterial_Survey_on_virtual_coaching_for_older_adults.cleaned for Survey on virtual coaching for older adults by Nerea Lete, Andoni Beristain and Alejandro García-Alonso in Health Informatics Journal

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 769830.

ORCID iDs

Nerea Lete

Andoni Beristain

Alejandro García-Alonso

Supplemental material

Supplemental material for this article is available online.

References

Ding

Liu

Cooper

, et al. Virtual coach technology for supporting self-care. Phys Med Rehabil Clin N Am 2010; 21(1): 179–194.

Siewiorek

. Invited talk: virtual coaches in health care. In: IEEE/IFIP international conference on dependable systems and networks workshops (DSN 2012), Boston, MA, 25–28 June 2012, p. 1. New York: IEEE.

Maestro. Home, http://www.maestro-aal.eu/ (accessed 17 September 2018).

Cabrita

Op den Akker

Tabak

, et al. Persuasive technology to support active and healthy ageing: an exploration of past, present, and future. J Biomed Inform 2018; 84: 17–30.

Fitbit Charge HR^TM, https://www.fitbit.com/es/chargehr (accessed 28 March 2018).

Polar USA. Heart Rate Monitors, activity trackers and bike computers, https://www.polar.com/us-en (accessed 17 April 2018).

Linx IAS®. 125 Tech Park Dr and Rochester. Know your impact, https://linxias.com/ (accessed 13 March 2018).

Q30 Sports Science. Q-Collar, https://q30innovations.com/q-collar/

Picard

Healey

. Affective wearables. Pers Technol 1997; 1(4): 231–240.

10.

Lee

Dey

. Lifelogging memory appliance for people with episodic memory impairment. In: Proceedings of the 10th international conference on ubiquitous computing, Seoul, South Korea, 21–24 September 2008, pp. 44–53. New York: ACM.

11.

Spire. Spire Stone, https://spire.io/pages/stone (accessed 17 September 2018).

12.

Spire. Spire Health Tag, https://spire.io/pages/healthtag (accessed 19 September 2018).

13.

Cognition Kit, https://www.cognitionkit.com/ (accessed 17 September 2018).

14.

Cognition Kit. Takeda and Cognition Kit present results from digital wearable technology study in patients with major depressive disorder (MDD), https://www.cognitionkit.com/press/release-3/ (accessed 17 September 2018).

15.

Scalise

Cosoli

. Wearables for health and fitness: measurement characteristics and accuracy. In: 2018 IEEE international instrumentation and measurement technology conference (I2MTC), Houston, TX, 14–17 May 2018, pp. 1–6. New York: IEEE.

16.

WHOOP. WHOOP Strap 2.0, https://get.whoop.com/products/whoop-strap (accessed 12 July 2020).

17.

Knight

Schwirtz

Psomadelis

, et al. The design of the SensVest. Pers Ubiquitous Comput 2005; 9(1): 6–19.

18.

Belmonte-Fernández

Puertas-Cabedo

Torres-Sospedra

, et al. An indoor positioning system based on wearables for ambient-assisted living. Sensors 2016; 17(1): 36.

19.

Hammoud

Vermeulen

Jaber

, et al. Activity measurement for elderly population using smart phone: criterion validity in daily life. In: 2017 sensors networks smart and emerging technologies (SENSET), Beirut, Lebanon, 12–14 September 2017, pp. 1–4. New York: IEEE.

20.

Seshadri

Drummond

Craker

, et al. Wearable devices for sports: new integrated technologies allow coaches, physicians, and trainers to better understand the physical demands of athletes in real time. IEEE Pulse 2017; 8(1): 38–43.

21.

Nuaimi

Kamel

. A survey of indoor positioning systems and algorithms. In: 2011 international conference on innovations in information technology, Abu Dhabi, United Arab Emirates, 25–27 April 2011, pp. 185–190. New York: IEEE.

22.

Chan

SHG

. Wi-Fi fingerprint-based indoor positioning: recent advances and comparisons. IEEE Commun Surv Tutor 2016; 18(1): 466–490.

23.

Morley

Vellas

van Kan

, et al. Frailty consensus: a call to action. J Am Med Dir Assoc 2013; 14(6): 392–397.

24.

Silapasuphakornwong

Sookhanaphibarn

. A conceptual framework for an elder-supported smart home. In: 2017 IEEE 6th global conference on consumer electronics (GCCE), Nagoya, Japan, 24–27 October 2017, pp. 1–5. New York: IEEE.

25.

Dasenbrock

Heinks

Schwenk Bauer

MJM

. Technology-based measurements for screening, monitoring and preventing frailty. Z Gerontol Geriatr 2016; 49(7): 581–595.

26.

Weka. Home, https://www.cs.waikato.ac.nz/ml/weka/ (accessed 12 July 2020).

27.

Lee

Joseph

Enriquez

, et al. Toward using a smartwatch to monitor frailty in a hospital setting: using a single wrist-wearable sensor to assess frailty in bedbound inpatients. Gerontology 2018; 64(4): 389–400.

28.

Zhou

Lee

, et al. Motor planning error: toward measuring cognitive frailty in older adults using wearables. Sensors 2018; 18(3): 926.

29.

Han

Shao

, et al. Enhanced computer vision with Microsoft Kinect sensor: a review. IEEE T Cybern 2013; 43(5): 1318–1334.

30.

Delp

Anderson

Arnold

, et al. OpenSim: Open-source software to create and analyze dynamic simulations of movement. IEEE T Biomed Eng 2007; 54(11): 1940–1950.

31.

Van Den Broeck

. Footstep localization based on in-home microphone-array signals. In: Encarnação

Azevedo

Newell

, et al. (eds) AAATE 2013. Amsterdam: IOS Press, 2013, pp. 90–94.

32.

Van Den Broeck

Karsmakers

Van Hamme

, et al. Noise robust footstep location estimation using a wireless acoustic sensor network. J Ambient Intell Smart Environ 2016; 8(6): 665–679.

33.

Hâkansson

Ljungblad

Holmquist

. Capturing the invisible: designing context-aware photography. In: Proceedings of the 2003 conference on designing for user experiences, San Francisco, CA, 6–7 June 2003, pp. 1–4. New York: ACM.

34.

Loke

. Context-aware pervasive systems. Boston, MA: Auerbach Publications, 2006.

35.

Croonenborghs

Luca

Karsmakers

, et al. Healthcare decision support systems at home. In: Artificial intelligence applied to assistive technologies and smart environments: papers from the AAAI-14 workshop, Québec City, QC, Canada, 27 July 2014, pp. 9–10. Québec City, QC, Canada: AAAI.

36.

Varghese

Cherian

Kizhakkethottam

. Overview on emotion recognition system. In: 2015 international conference on soft-computing and networks security (ICSNS), Coimbatore, India, 25–27 February 2015, https://ieeexplore.ieee.org/document/7292443/ (accessed 28 May 2018).

37.

Mehta

Siddiqui

MFH

Javaid

. Facial emotion recognition: a survey and real-world user experiences in mixed reality. Sensors 2018; 18(2): 416.

38.

Khan

Zualkernan

. SensePods: a ZigBee-based tangible smart home interface. IEEE T Consum Electron 2018; 64(2): 145–152.

39.

Emerich

Lupu

Apatean

. Emotions recognition by speechand facial expressions analysis. In: 2009 17th European signal processing conference, Glasgow, 24–28 August 2009, pp. 1617–1621. New York: IEEE.

40.

Saravanan

Palanivel

Balasubramanian

. Facial expression based person authentication. Int J Comput Appl 2014; 94: 1–8.

41.

Dawadi

Cook

Schmitter-Edgecombe

, et al. Automated assessment of cognitive health using smart home technologies. Technol Health Care 2013; 21(4): 323–343.

42.

Selker

Arroyo

Burleson

. Chameleon tables: using context information in everyday objects. In: CHI ’02 extended abstracts on human factors in computing systems, Minneapolis, MN, 20–25 April 2002, pp. 580–581. New York: ACM.

43.

Chang

Liu

Chu

, et al. The diet-aware dining table: observing dietary behaviors over a tabletop surface. In: Pervasive computing, Dublin, 7–10 May 2006, pp. 366–382. Berlin: Springer.

44.

Steurer

Srivastava

. System design of smart table. In: Proceedings of the first IEEE international conference on pervasive computing and communications, 2003 (PerCom 2003), Fort Worth, TX, 26 March 2003, pp. 473–480. New York: IEEE.

45.

THK. Smart Sensing Chair, Argus Chair, http://www.thk.com/arguschair/en/ (accessed 28 August 2018).

46.

Industrial Technology Research Institute. Smart Health Chair, https://www.itri.org.tw/eng/Content/MsgPic01/Contents.aspx?SiteID=1&MmmID=620651711475751171&MSid=654720265076425445 (accessed 28 August 2018).

47.

2M Engineering – Sensor Product Development. Smart sensor mattress contactless measurement of presence, respiration & heart rate in bed, https://www.2mel.nl/projects/bedsense-bed-sensor-technology-platform/

48.

Andoh

Ishikawa

Kobayashi

, et al. Home health monitoring system in the sleep. In: SICE 2003 annual conference (IEEE Cat. No.03TH8734), Fukui, Japan, 4–6 August 2003, pp. 1223–1226. New York: IEEE.

49.

Metzger

Meyer

Fleisch

, et al. Weight-sensitive foam to monitor product availability on retail shelves. In: Pervasive computing, Toronto, ON, Canada, 13–16 May 2007, pp. 268–279. Berlin: Springer.

50.

Bonanni

Arroyo

Lee

C-H

, et al. Smart sinks: real-world opportunities for context-aware interaction. In: CHI ’05 extended abstracts on human factors in computing systems, Portland, OR, 2–7 April 2005, pp. 1232–1235. New York: ACM.

51.

Hammer

Seiderer

Andre

, et al. Design of a lifestyle recommender system for the elderly: requirement gatherings in Germany and Greece. In: Proceedings of the 8th ACM international conference on pervasive technologies related to assistive environments, Corfu, 1–3 July 2015, pp. 1–80. New York: ACM.

52.

Fujinami

Kawsar

. An experience with augmenting a mirror as a personal ambient display. In: Computer-human interaction, Seoul, Korea, 6–9 July 2008, pp. 183–192. Berlin: Springer.

53.

Beigl

Gellerson

Schmidt

. MediaCups: experience with design and use of computer-augmented everyday objects. Comput Netw 2001; 35(4): 401–409.

54.

Counter intelligence – mit media lab, https://www.media.mit.edu/ci/projects/chameleonmug.html (accessed 4 May 2018).

55.

Greenberg

. The context aware pill bottle and medication monitor. In: Sixth international conference on ubiquitous computing, Nottingham, 7–10 September 2004.

56.

Cheng

Bonanni

. Intelligent spoon, https://www.media.mit.edu/ci/projects/intelligentspoon.html (accessed 28 August 2018).

57.

Jordan

, et al. The SmartCane system: an assistive device for geriatrics. In: Proceedings of the ICST 3rd international conference on body area networks, Tempe, AZ, 13–15 March 2008, pp. 1–2. Brussels: ICST.

58.

Abowd

Dey

Brown

, et al. Towards a better understanding of context and context-awareness. In: Handheld and ubiquitous computing, Karlsruhe, 27–29 September 1999, pp. 304–307. Berlin: Springer.

59.

Hong

Suh

Kim

S-J

. Context-aware systems: a literature review and classification. Expert Syst Appl 2009; 36(4): 8509–8522.

60.

Mehra

. Context-aware computing: beyond search and location-based services. IEEE Internet Comput 2012; 16(2): 12–16.

61.

Nykänen

Rodriguez

. Problems in context-aware semantic computing. Int J Interact Mob Technol 2014; 8(3): 32–39.

62.

Natis

Schulte

. Introduction to service-oriented architecture. Stamford, CT: Gartner, 2003.

63.

Jih

Hsu

Tsai

. Context-aware service integration for elderly care in a smart environment, 2006, pp. 44–48, https://www.aaai.org/Papers/Workshops/2006/WS-06-12/WS06-12-010.pdf

64.

Chen

Kotz

. A survey of context-aware mobile computing research. Hanover, NH: Dartmouth College, 2000.

65.

Pileggi

Rivero-Rodriguez

Nykanen

. Towards traditional simulation models of context using process mining. In: 2015 7th international conference on computational intelligence, communication systems and networks, Riga, 3–5 June 2015, pp. 70–75. New York: IEEE.

66.

Chen

Finin

Joshi

. An ontology for context-aware pervasive computing environments. Knowl Eng Rev 2003; 18(3): 197–207.

67.

Smith

Welty

McGuiness

. OWL web ontology language guide, https://www.w3.org/TR/owl-guide/ (2004, accessed 18 September 2018).

68.

Loureiro

Rodrigues

. Multi-touch as a natural user interface for elders: a survey. In: 6th Iberian conference on information systems and technologies (CISTI 2011), Chaves, 15–18 June 2011, pp. 1–6. New York: IEEE.

69.

Antle

Wise

. Getting down to details: using theories of cognition and learning to inform tangible user interface design. Interact Comput 2013; 25(1): 1–20.

70.

Mayer

Sobko

Mautone

. Social cues in multimedia learning: role of speaker’s voice. J Educ Psychol 2003; 95: 419–425, https://www.researchgate.net/publication/232560209_Social_Cues_in_Multimedia_Learning_Role_of_Speaker%27s_Voice (accessed 14 March 2018).

71.

Bickmore

Mauer

Crespo

, et al. Persuasion, task interruption and health regimen adherence. In: Persuasive technology, Palo Alto, CA, 26–27 April 2007, pp. 1–11. Berlin: Springer.

72.

Y-K

Liu

Kelleher

, et al. Evaluating the usability of a smartphone virtual seating coach application for powered wheelchair users. Med Eng Phys 2016; 38(6): 569–575.

73.

Kortum

. HCI beyond the GUI: design for haptic, speech, olfactory, and other nontraditional interfaces. San Francisco, CA: Morgan Kaufmann, 2008.

74.

Silva

Restivo

Gabriel

. Haptic device demo using temperature feedback. In: 2013 2nd experiment@ international conference (expat’13), Coimbra, 18–20 September 2013, pp. 172–173. New York: IEEE.

75.

Wall

Wrisley

Oddsson

. Vibrotactile feedback of mediolateral trunk tilt or foot pressure increases locomotor performance in healthy older adults – a pilot study. In: 2012 annual international conference of the IEEE engineering in medicine and biology society, San Diego, CA, 28 August–1 September 2012, pp. 6145–6148. New York: IEEE.

76.

Byun

H-G

Lee

H-R

Kim

J-D

, et al. Implementation of olfactory interaction between images and smells. In: 2012 IEEE sensors, Taipei, Taiwan, 28–31 October 2012, pp. 1–4. New York: IEEE.

77.

Karunanayaka

Johari

Hariri

, et al. New thermal taste actuation technology for future multisensory virtual reality and Internet. IEEE T Vis Comput Graph 2018; 24(4): 1496–1505.

78.

OpenIoT. Open Source cloud solution for the Internet of Things, http://www.openiot.eu (accessed 18 September 2018).

79.

Ranganathan

Al-Muhtadi

Chetan

, et al. MiddleWhere: a middleware for location awareness in ubiquitous computing applications. In: Proceedings of the 5th ACM/IFIP/USENIX international conference on middleware, Toronto, ON, Canada, 18–22 October 2004, pp. 397–416. Berlin; Heidelberg: Springer.

80.

Eisenhauer

Rosengren

Antolin

. A development platform for integrating wireless devices and sensors into ambient intelligence systems. In: 2009 6th IEEE annual communications society conference on sensor, mesh and ad hoc communications and networks workshops, Rome, 22–26 June 2009, pp. 1–3. New York: IEEE.

81.

FIWARE. The open source platform for our smart digital future, https://www.fiware.org/ (accessed 18 September 2018).

82.

Oracle. Oracle fusion middleware, https://www.oracle.com/middleware/ (accessed 20 December 2018).

83.

Gatebox, https://gatebox.ai/home/ (accessed 17 September 2018).

84.

King

Ohya

. The representation of agents: anthropomorphism, agency, and intelligence. In: Conference companion on human factors in computing systems, Vancouver, BC, Canada, 13–18 April 1996, pp. 289–290. New York: ACM.

85.

Bickmore

Picard

. Establishing and maintaining long-term human–computer relationships. ACM Trans Comput-Hum Interact 2005; 12(2): 293–327.

86.

Chaminda

Klyuev

Naruse

. A smart reminder system for complex human activities. In: 2012 14th international conference on advanced communication technology (ICACT), Pyeongchang, South Korea, 19–22 February 2012, pp. 235–240. New York: IEEE.

87.

Moreno

Mayer

Spires

, et al. The case for social agency in computer-based teaching: do students learn more deeply when they interact with animated pedagogical agents. Cogn Instr 2001; 19(2): 177–213.

88.

Atkinson

. Optimizing learning from examples using animated pedagogical agents. J Educ Psychol 2002; 94(2): 416–427.

89.

Zhou

Chu

C-H

, et al. A context-aware reminder system for elders based on fuzzy linguistic approach. Expert Syst Appl 2012; 39(10): 9411–9419.

90.

LensKit recommender toolkit, http://lenskit.org/ (accessed 17 April 2018).

91.

MedSignal. Medication monitoring, medication management systems, http://www.medsignals.com/ (accessed 17 September 2018).

92.

Kim

Park

, et al. Gate reminder: a design case of a smart reminder. In: Proceedings of the 5th conference on designing interactive systems: processes, practices, methods, and techniques, Cambridge, MA, 1–4 August 2004, pp. 81–90. New York: ACM.

93.

Boulos

MNK

Rocha

Martins

, et al. CAALYX: a new generation of location-based services in healthcare. Int J Health Geogr 2007; 6: 9.

94.

Herpich

Rist

André

. Co-designing a recommender system for the elderly. In: Proceedings of the 10th international conference on pervasive technologies related to assistive environments, Island of Rhodes, 21–23 June 2017, pp. 52–55. New York: ACM.

95.

Drools. Drools: business rules management system (Java^TM, Open Source), http://drools.org/ (accessed 17 April 2018).

96.

Pessemier

Dooms

Martens

. Context-aware recommendations through context and activity recognition in a mobile environment. Multimed Tools Appl 2014; 72(3): 2925–2948.

97.

Ryan

Deci

. Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. Am Psychol 2000; 55(1): 68–78.

98.

Hao

Chuen-Tsai

. Game reward systems: gaming experiences and social meanings. In: 5th international conference on digital research association: think design play, Utrecht, 14–17 September 2011.

99.

De Caravalho

RNS

Ishitani

. Motivational factors for mobile serious games for elderly users. In: Proceedings of XI SB games, Brasilia, Brazil, 2–4 November 2012.

100.

Grace

Maher

Mohseni

, et al. Encouraging p-creative behaviour with computational curiosity. In: International conference on computational creativity, Atlanta, GA, 19–23 June 2017.

101.

Ajdukovic

Cajvert

Judy

, et al. ECVision. A European glossary of supervision and coaching, https://www.oevs.or.at/wp-content/uploads/2013/12/ECVision_Glossary.pdf (2014, accessed 4 December 2018).

102.

Power

Sharda

Burstein

. Decision support systems. In: Cooper

(ed.) Wiley encyclopedia of management. Chichester: Wiley, 2015, pp. 1–4.

103.

Sim

Gorman

Greenes

, et al. Clinical decision support systems for the practice of evidence-based medicine. J Am Med Inform Assoc 2001; 8(6): 527–534.

104.

Holsapple

. DSS architecture and types. In: Burstein

Holsapple

(eds) Handbook on decision support systems 1: basic themes. Berlin; Heidelberg: Springer, 2008, pp. 163–189.

105.

Marschollek

. Decision support at home (DS@HOME) – system architectures and requirements. BMC Med Inform Decis Mak 2012; 12: 43, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3464181/ (accessed 13 March 2018).

106.

Sellak

Ouhbi

Frikh

. Towards an intelligent decision support system for renewable energy management. In: 2015 15th international conference on intelligent systems design and applications (ISDA), Marrakech, Morocco, 14–15 December 2015, pp. 319–324. New York: IEEE.

107.

Ellouzi

Ltifi

Ben Ayed

. New multi-agent architecture of visual intelligent decision support systems application in the medical field. In: 2015 IEEE/ACS 12th international conference of computer systems and applications (AICCSA), Marrakech, Morocco, 17–20 November 2015, pp. 1–8. New York: IEEE.

108.

Shiffman

Stone

Hufford

. Ecological momentary assessment. Annu Rev Clin Psychol 2008; 4: 1–32.

109.

Heron

Smyth

. Ecological momentary interventions: incorporating mobile technology into psychosocial and health behavior treatments. Br J Health Psychol 2010; 15(1): 1–39.

110.

Álvarez

Murua

Artetxe

, et al. A platform for user empowerment through self ecological momentary assessment/intervention. In: 5th EAI international conference on wireless mobile communication and healthcare – MobiHealth 2015, London, 14–16 October 2015.

111.

de Vries

. An integrated approach for understanding health behavior; the I-Change model as an example. Psychol Behav Sci Int J 2017; 2(2): 555585.

112.

Lim

Dey

. Toolkit to support intelligibility in context-aware applications. In: Proceedings of the 12th ACM international conference on ubiquitous computing, Copenhagen, 26–29 September 2010, pp. 13–22. New York: ACM.

113.

Janz

Becker

. The health belief model: a decade later. Health Educ Q 1984; 11(1): 1–47.

114.

Bandura

. Social foundations of thought and action: a social cognitive theory. Englewood Cliffs, NJ: Prentice-Hall, 1986.

115.

Ajzen

. The theory of planned behavior. Organ Behav Hum Decis Process 1991; 50(2): 179–211.

116.

Prochaska

DiClemente

. Stages and processes of self-change of smoking: toward an integrative model of change. J Consult Clin Psychol 1983; 51(3): 390–395.

117.

Michie

Richardson

Johnston

, et al. The behavior change technique taxonomy (v1) of 93 hierarchically clustered techniques: building an international consensus for the reporting of behavior change interventions. Ann Behav Med 2013; 46(1): 81–95.

118.

Pollack

Brown

Colbry

, et al. Autominder: an intelligent cognitive orthotic system for people with memory impairment. Robot Auton Syst 2003; 44(3): 273–282.

119.

French

Tyamagundlu

Siewiorek

, et al. Towards a virtual coach for manual wheelchair users. In: 2008 12th IEEE international symposium on wearable computers, Pittsburgh, PA, 28 September–1 October 2008, pp. 77–80. New York: IEEE.

120.

Klein

MCA

Manzoor

Middelweerd

, et al. Encouraging physical activity via a personalized mobile system. IEEE Internet Comput 2015; 19(4): 20–27.

121.

Vales-Alonso

Chaves-Diéguez

López-Matencio

, et al. SAETA: a smart coaching assistant for professional volleyball training. IEEE Trans Syst Man Cybern 2015; 45(8): 1138–1150.

122.

Blanson Henkemans

van der Boog

PJM

Lindenberg

, et al. An online lifestyle diary with a persuasive computer assistant providing feedback on self-management. Technol Health Care 2009; 17(3): 253–267.

123.

Kyriazakos

Valentini

Cesario Zachariae

. FORECAST – a cloud-based personalized intelligent virtual coaching platform for the well-being of cancer patients. Clin Transl Radiat Oncol 2018; 8: 50–59.

124.

Virzi

. Refining the test phase of usability evaluation: how many subjects is enough? Hum Factors 1992; 34(4): 457–468.

125.

McDonald

Walsh

Vergara

, et al. Effect of a virtual pain coach on pain management discussions: a pilot study. Pain Manag Nurs 2013; 14(4): 200–209.

126.

Bickmore

Mauer Brown

. Context awareness in a handheld exercise agent. Pervasive Mob Comput 2009; 5(3): 226–235.

127.

Buttussi

Chittaro

Nadalutti

. Bringing mobile guides and fitness activities together: a solution based on an embodied virtual trainer. In: Proceedings of the 8th conference on human–computer interaction with mobile devices and services, Helsinki, 12–15 September 2006, pp. 29–36. New York: ACM.

128.

Mihailidis

Boger

Craig

, et al. The COACH prompting system to assist older adults with dementia through handwashing: an efficacy study. BMC Geriatr 2008; 8: 28.

129.

Beun

Fitrianie

Griffioen-Both

, et al. Talk and Tools: the best of both worlds in mobile user interfaces for E-coaching. Pers Ubiquitous Comput 2017; 21(4): 661–674.

130.

Hudlicka

. Virtual training and coaching of health behavior: example from mindfulness meditation training. Patient Educ Couns 2013; 92(2): 160–166.

131.

Watson

Bickmore

Cange

, et al. An internet-based virtual coach to promote physical activity adherence in overweight adults: randomized controlled trial. J Med Internet Res 2012; 14(1): e1.

132.

Ellis

Latham

DeAngelis

, et al. Feasibility of a virtual exercise coach to promote walking in community-dwelling persons with Parkinson disease. Am J Phys Med Rehabil 2013; 92(6): 472–481; quiz 482–485.

133.

IJsselsteijn

Kort

YAWd

Westerink

, et al. Virtual fitness: stimulating exercise behavior through media technology. Presence 2006; 15(6): 688–698.

134.

Sullivan

Artino

Jr.

Analyzing and interpreting data from Likert-type scales. J Grad Med Educ 2013; 5(4): 541–542.

135.

Asselin

Ortiz

Pui

, et al. Implementation and evaluation of the personal wellness coach. In: 25th IEEE international conference on distributed computing systems workshops, Columbus, OH, 6–10 June 2005, pp. 529–535. New York: IEEE.

136.

Laugwitz

Held

Schrepp

. Construction and evaluation of a user experience questionnaire. In: HCI and usability for education and work, Graz, 20–21 November 2008, pp. 63–76. Berlin: Springer.

137.

Davis

. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Q 1989; 13(3): 319–340.

138.

Harris

D’Eon

. Psychometric properties of the Beck Depression Inventory-second edition (BDI-II) in individuals with chronic pain. Pain 2008; 137(3): 609–622.

139.

Cleeland

Ryan

. Pain assessment: global use of the Brief Pain Inventory. Ann Acad Med Singap 1994; 23(2): 129–138.

140.

Hoehn

Yahr

. Parkinsonism: onset, progression, and mortality 1967. Neurology 2001; 57(10 Suppl. 3): S11–S26.

141.

Folstein

McHugh

. ‘Mini-mental state’: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12(3): 189–198.

142.

Thomas

Reading

Shephard

. Revision of the Physical Activity Readiness Questionnaire (PAR-Q). Can J Sport Sci 199; 217(4): 338–345.

143.

Lins

Carvalho

. SF-36 total score as a single measure of health-related quality of life: scoping review. SAGE Open Med 2016; 4: 1–12.

144.

Horvath

Greenbargh

. Development and validation of the working alliance inventory. J Couns Psychol 1989; 36(2): 223–233.

145.

Lessiter

Freeman

Keogh

, et al. A cross-media presence questionnaire: the ITC-sense of presence inventory. Presence 2001; 10(3): 282–297.

146.

Albaina

Visser

van der Mast

CAPG

, et al. Flowie: a persuasive virtual coach to motivate elderly individuals to walk. In: 2009 3rd international conference on pervasive computing technologies for healthcare, London, 1–3 April 2009, pp. 1–7. Brussels: ICST.

147.

Personalised coaching for well-being and care of people as they age, http://ec.europa.eu/research/participants/portal/desktop/en/opportunities/h2020/topics/sc1-pm-15-2017.html (accessed 20 December 2018).

148.

CAPTAIN (Coach Assistant via Projected and Tangible Interface), http://captain-eu.org/ (accessed 6 June 2018).

149.

vCare project, https://vcare-project.eu/

150.

Empathic Project, http://www.empathic-project.eu/ (accessed 19 June 2018).

151.

Nestore Project, https://nestore-coach.eu/home (accessed 17 September 2018).

152.

SAAM (Supporting Active Ageing through Multimodal coaching), https://www.saam2020.eu/ (accessed 17 September 2018).

153.

Council of Coaches, http://council-of-coaches.eu/ (accessed 17 September 2018).

154.

Konstantinidis

Petsani

Conti

, et al. A new approach for ageing at home: the CAPTAIN system. Stud Health Technol Inform 2019; 264: 1704–1705.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.53 MB

0.50 MB