Wearable haptic-feedback navigational assistance for people with dementia: Preliminary assessment

2.1 Decline of spatial navigation due to ageing and AD progression

The associative factor between aging and spatial navigation is well recognized. Typically, wayfinding deficits increased exponentially with the age and the severity of the AD [12, 15]. This association is usually assessed by investigating the spatial/route navigation tasks. For instance, due to this wayfinding deficit, older adults tend to navigate only in the familiar routes or environments [16, 17], to avoid getting lost. Also, in navigating a route, older adults always took longer to reaching the target destinations and made more mistakes as compared to younger adults [18, 9].

While this is the case, spatial cognition and performance are worsened, as the cognitive domain is deteriorated in AD patients. In fact, spatial disorientation is diagnosed early in the development of AD, which is during the Mild Cognitive Impairment (MCI) stage [19]. Although not every MCI case ends up with full-blown dementia, the possibility to develop it is still high [20]. Number of studies found that adults with dementia were significantly slower in motion search tasks [21], unable to sense movement [22] and also reduced control of spatial focus [23] due to deficits in the working memory.

2.2 Sensory impairments effects the sense of direction

Older adults with AD alongside with cognitive and behavioral changes also manifest the sensory changes as they grow older. In particular, the sensory decline indirectly affects their spatial cognition and visuospatial representation that associate with the sense of direction [17]. Whilst, cognitive mapping [24] constructs the association between sensory acuity to the oriented search [25, 10].

Loss of color sensitivity and reduced depth perception and contrast sensitivity [26, 27] are the main issues in ageing vision. In the same way, older adults with or without AD usually show mild-to-moderate hearing impairment [28]. And yet, deficits in central auditory processing capacity and cognitive decline weaken the communication process [29], as well as visuospatial and auditory information in AD [30]. These sensory changes resulted in poor spatial navigation skill and wayfinding abilities in general.

Similar to vision and hearing, tactile acuity is progressively impaired due to aging [31] which contributes to the age-related impairment of sensorimotor and cognitive abilities. However, as proven by the assessment of speeded object identification task, individuals with AD maintained intact haptic priming, despite being highly impaired in recognition performances [32].

These pieces of information are feasible in designing the new devices using other sensing abilities, like haptic perception itself. This is the fundamental indication that encourages us to explore further on its unique qualities. Also, the conducted literature analysis did not elicit such an approach, which deviously certifies the innovativeness of our conceptual design.

2.3 The needs for alternative form of navigational assistances

Current navigational devices took advantage of the advance technological applications and its rapid ubiquity. The commercial map and route-based navigations that use Global Positioning System (GPS), is a very well established technique to meet the current wayfinding issues. Typically, this method of navigation supports predominantly on visual interactions and the common additional feature is the integration of auditory support, i.e. speech instructions or sound cues. Unfortunately for the older adults with AD, their most obvious sensory declines are both visions and hearings. Hence, these modalities may hinder the effective wayfinding.

The use of haptic/tactile stimulus as an output on feedback could be a good substitute to the current wayfinding method for individuals with AD, mainly due to the abovementioned limitations. In addition, since the spatial disorientation and sensory changes worsen the navigability in older adults with AD, the ideal navigational assistances should apparently be designed to support the remaining abilities.

Haptic devices (or haptic interfaces) are mechanical devices that mediate the communication between users and computers. In human-computer interaction (HCI), haptic feedback means both tactile and force feedbacks. Tactile, or touch feedback is the term applied to the sensations felt by the skin, and this feedback allows the users to feel things such as the texture of surfaces, temperature and vibration [33, 34].

2.4 Haptic/tactile-feedback for navigation

Presently, there is a growing body of research on the use of haptic/tactile modalities in the domain of mobile human computer interactions. This unique modality has been explored in a variety of applications such as the haptic-feedback system to sense of virtual objects [35], in surgical tasks [36], designing the human-computer interfaces [37], and for navigational instructions [38, 39, 40, 41]. Nonetheless, for wayfinding or navigational purposes, the existing body of works often focuses on individuals with severe visual impairment or blind people, and not specifically for the persons with cognitive impairment.

For instance, Zöllner et al. [38] conceptualized a mobile navigational assistance with Microsoft Kinect and optical marker tracking to help the indoor navigation of individuals with visual impairments. They created a belt embedded with vibrotactile outputs that were used to detect obstacles during wayfinding tasks.

Similarly, another study by Ertan et al. [39] described a wearable navigation system using haptic directional display integrated with a vest. The system gives haptic signals to the users’ back in the course of navigation. In a different study, Mann et al. [40] presented a blind navigation system with a Kinect 3D sensor range camera and a vibrotactile helmet. The vibrating actuators embedded inside the helmet convert depth information into haptic feedback, to allow the users in identifying depths for collision avoidances.

Closer to our goal is done by Grierson et al. [41] who investigated the applicability of tactile signals to assist the wayfinding of persons with dementia. In this study, they developed a wearable belt with vibrating motors. Participants were asked to navigate a series of routes within a hospital with the assistance of vibrotactile signals. This is where our design concept initiated but with a twist. We found that it is important to highlight the potentials of haptic/tactile stimuli to assist or improve the wayfinding of individuals with cognitive impairments to serve their different needs and issues.

Another essential point is most of the navigational assistances described above are the wearable technologies/devices. This is most probably because designing devices to be wearable improve the practicality of handling and operating [42, 43]. Furthermore, haptic/tactile feedback as forms of signals provides a simple, yet a promising form of directional cues that allows users to concentrate on the surrounding with other senses (vision and hearing) during wayfinding [44]. Besides, the vibrotactile signals that created the haptic simulation are less disruptive as compared to the auditory instructions, which is a suitable substitute for continuous feedback [38].

Technological interventions that utilize haptic/tactile modality to assist wayfinding have shown positive results, though most of them were meant for visually impaired and blind people. Furthermore, most of the existing wayfinding intervention strategies, which include the above-mentioned studies [38, 39, 41] focus on indoor navigation. This means, further research on navigational assistance for outdoor wayfinding purposes are highly recommended. The conceptual design that we propose employs the hybrid approach between; (1) the unique advantage of haptic/tactile, (2) device wearability, and (3) outdoor navigational strategies.

4.1 Method and participants

A survey was conducted as a preliminary assessment to investigate the feasibility of the design concept in terms of: (1) acceptability of intervention, (2) usability on users, (3) wearability of the device, and (4) setting suitability of interventions. There are several strategies to assess a design though the inclusion of users in the design and development process as suggested by [48]. Accordingly, background interviews and questionnaires are usually conducted at the beginning of the design project to obtain data associated to the needs and expectations of users. Meanwhile, although this intervention concerns individuals with AD, attaining feedbacks from them could be insubstantial because this survey requires a clear justification of preferences and definite reasoning. Thus, the target respondents of the survey were; (1) caregivers and (2) clinical/medical experts of AD and dementia.

We postulate that these respondents could be the best communication mediators and information pro- viders who represent AD patients for the questions and issues related to dementia in general. This is because, they recognize the distinctive behavioral and cognitive changes of the AD patients and have the most knowledge base about patients’ caregiving need [45].

In addition, the ethical issues related to the appropriateness of this proposed intervention towards the target users have been adequately considered before we administered the survey. This ethical consideration was initiated with the advice from dementia experts we are currently working with. For this reason, there was no need for the questions related to the ethical issues for this particular assessment. However, in the future and with an extended population that includes the participation of AD subjects, this matter will certainly be taken into account.

The survey was distributed online to the selected AD and dementia institutions in the United Kingdom, Italy and Malaysia based on the availability, recommendations and the established networking. It was made of both close-ended and open-ended questions, but the respondents could also add their additional comments where necessary. Before the questions began, we provided the detailed description of the proposed navigational assistance concept. Here, we explained the approach and reason of the choice of a haptic system through the drawings as previously presented in the Conceptual Design section. We supposed that the description was necessary since the proposed concept is unlike the ordinary or existing navigational devices. The description also explains the technical terms used that might not be familiar by the respondents. At the end of the questionnaire but before the completion of this assessment, the respondents were asked to state their full understanding of the proposed concept based on their general technological knowledge.

In order to justify the specific objectives of this preliminary assessment, the questions are subcategorized into six segments; (1) demographic of respondents, (2) acceptability, (3) wearability, (4) setting suitability, (5) usability, and (6) general concept. Table 1 shows these segments and its questions, as well as the scores and the descriptive analysis.

Forty two respondents ( F = 23; M = 19) responded to the online questionnaire. The age of the participants ranged from above 65 years old ( n = 4) and the youngest is between 22 to 34 years old ( n = 11). Most of the respondents were professionals and consisted of therapist/neuropsychologist/medical doctor/ student/researcher in dementia & AD. It was followed closely by AD & Dementia caregivers. Some of the caregivers were the family members or the relatives of the persons with AD. Nonetheless, the feedbacks from the family members or the relatives were only considered if only they are the professionally-certified as dementia caregivers or if they have the working experience at the nursing homes. This is to avoid the ambiguities of the validity for this preliminary assessment.

According to the UCD principles, we considered that even the AD patients are the primary users of the device, but the overall intervention – in assisting the wayfinding of persons with dementia, includes a bigger stakeholder structure that comprises of patients’ caregivers, family members, nursing homes staffs and managers etc. Everyone who directly or indirectly involves has an active role in the device use and its management. However, even if the collected data had polarized initial results, the assessment on real users (AD patients) by means of Usability Testing will be performed soon after the development of the functioning prototype. Results presented in the next section comprehend both answers to closed questions and open comments.

The importance to identify if the proposed conceptual design of navigational assistance is acceptable by the users (persons with AD) is highly appropriate. In fact, it is one of the primary concerns of the investigation, before the design concept may be forwarded into the next stages; prototyping, usability testing and implementation. Specifically for this segment, we need to understand how the new proposed navigational assistance is received by the users, positively or negatively. As previously mentioned, none of the persons with AD took part in this preliminary assessment, since they might not be able to clearly opinionate their perceived views. Having said this, the feedbacks from the caregivers and the AD experts are already a convincing dataset and may determine the success of this proposal in its initial stage.

Firstly, we asked if the persons with AD would use this new navigational device as well as if they can cope with the continuous use to get familiarized with the concept, since it is a new form of navigational assistance. The quantitative data confidently shows that the respondents received the proposal well, with most of the respondents agreed that this concept is practical to be used by the target users.

The following questions directly looked into the ‘acceptability’ of this proposed design concept. Once again, most of the respondents (62.3%) rated acceptable and only 4.8% rated highly unacceptable. The further qualitative analysis show that the high percentage of acceptance might be due to the proposed concept provides the uncomplicated features. The simple features are clearly meant to minimize their confusion caused by the spatial disorientation and navigational disabilities, especially during the complex wayfinding tasks.

In terms of wearability, the respondents were first required to choose the most preferred positions to place the device or to trigger the haptic/tactile stimuli. The body-part positions are based on suggestions by Gemperle [45], and with possible integration on underwear. Results show that both waist and shoulder shared the same highest score (26.2%), while thighs and other positions were the least favored (9.5% and 16.7% respectively).

In this segment, respondents were also asked to rate (on the scale of 1 to 5, with 1 represents the minimum score of agreement and 5 for the maximum) if it is appropriate to integrate or attach the device onto users’ clothes or underwear. This is purposely for the practicality of use, as previously discussed. Positive feedbacks were gathered for this specific question, with 31.0% of the respondents rated 4/5 and 26.2% rated 5/5. Results indicate the possibility to explore more on the wearable aspect of the navigational device. From the proposed clothes or underwear, respondents mostly preferred the singlet (35.7%) and secondly the briefs (26.2%) to be integrated or attached with the device. These selections validate the most preferred body-part positions as justified earlier – singlet is suitable for shoulder, while the brief is appropriate for waist-length and abdominal area.

From the usability viewpoint, the respondents were required to rate (using the same scale of 1 to 5) the unhelpfulness/helpfulness of the proposed concept of navigation. Most of them (62.3%) rated the device as ‘helpful’ which indicates its usefulness for wayfinding. Next, they were asked if they preferred the navigational assistance to be a standalone device. Standalone here means the device works on its own without the integration of another device, such as mobile phone. For this, the majority (45.2%) approved the concept of standalone device.

However, when they were asked if additional features such as visual and auditory instructions were needed, 42.9% were unsure of this, and 33.3% prefer the additional features. The following question rectified this disagreement. Accordingly, given the choices of additional modalities: [1] speech instruction, [2] visual instruction, and [3] both speech and visual instructions, mostly agreed that the essential additional features to the device are; speech instructions (40.5%) and both speech and visual instructions (also 40.5%).

In terms of environmental setting suitability, the majority agreed that this intervention strategy is best to be implemented mainly in an outdoor environment. Again, from the scale of 1 to 5, 35.7% rated 4/5 and 33.3% rated 5/5. For the outdoor wayfinding purposes, respondents agreed that the proposed device is useful to facilitate the users to navigate; (1) to the nearest parks, markets, hospitals, etc., (2) to use the public transportations, and (3) socializing inside community. Thus, as many other outdoor wayfinding interventions [49, 50, 51], the data suggest that the respondents approved that this outdoor navigation device is more appropriate to be used within the neighborhoods they live in.

Furthermore, most of the respondents recommended that the users should not be allowed to travel alone too far with any navigational assistance devices. Given the proposed range of allowed travel distances, 31.0% chose less than 1 km, 26.2% chose the range of 2 to 4 km, 23.8% preferred 5 to 6 km, and none of them chose the longer range of distance (above 7 km). The preferred short distance is appropriate for the users’ safety, by means of reducing their risk being in an open space unattended for a long period of time. This is therefore justified that the proposed device should be designed to be used within their neighborhood that only involves the normal daily routines/activities and covers not so wide travel range (between 1 to 4 km).

There are some important concerns raised from the further analysis of the corresponding qualitative data. These concerns are important to be considered to improve the proposed design concept. In terms of its functionality, even the quantitative analysis positively demonstrated that the device is acceptable and practical to be used, introducing a new form of navigational assistance itself could be a challenge especially for the target users; AD patients. However in this case, the qualitative analysis suggests that if the issues of familiarity and training can be resolved, the proposed concept can be successfully implemented. Here, many of the respondents recommended to provide a manual in a simple lay language to the caregivers and therapists, who will get familiarized with the device before they can train the real users.

Another important point is the design of the wearable device itself. The selection of suitable materials to be used for the wearable device should consider the appropriateness of long hours of use, as well as of different environmental and contextual climates. Here, comfortability and unobtrusive are the key criteria to be considered. This is to avoid the post-implementation issues faced by the users, possibly due to the skin sensitivity and differential climate conditions. In addition, giving a discreet look at the design is crucial because the too noticeable device may cause discomfort and stigmatizing to the users.

OPEN IN VIEWER

Finally, as many of them agreed, this intervention should be targeting to the individuals in the early stage of cognitive impairment. The level of consciousness and cognitive function in AD patients are worsen as the disease progresses [52, 2], which affect their spatial disorientation and the ability to navigate simultaneously [7, 11]. This specific concern is raised because those in the severe stage might not be able to perceive the haptic stimuli as forms of wayfinding signals. This apparently approved one of the main research objectives, which is to provide a navigational assistance to the AD patients who are not yet institutionalized. Typically, persons with AD who still live at home are in the early to moderate stage of the disease [53].

From all the above results, we may suggest that the respondents found the device relevant to be implemented towards the persons with AD based on three main reasons: (1) the safety aspect when the patients wearing/using it is ensured, (2) its simple functionality that applies to individuals in both early and moderate stages of AD and MCI, and (3) its unobtrusiveness and non-stigmatizing aspects that could augment a high acceptance by the users.