Keywords

1 Introduction

Enhancing mobility in walk-up apartments, apartments that lack lifts in multi-storey structures, is essential for supporting aging in place [19, 29]. The challenge of stair navigating in such apartments may increase fall risks in older adults [8], which leads to severe consequences such as fractures, soft tissue damage, and even death [25], potentially restricting their ability to live independently, decreasing their willingness to go out [30], and may even lead to psychological issues [4]. Therefore, smart technology for such issue of aging in place has become a prominent focus in human-computer interaction [7, 23]. Deficiencies in lower limb muscle strength and balance contribute to the high occurrence of falls in older adults [13], especially during weight-bearing walking and stair navigation [27] (see Fig. 1).

Fig. 1.
figure 1

Weight bearing in older adults’ daily routine (Image courtesy of authors)

Full size image

Therefore, improving body balance and reducing the load on lower limb muscles while navigating stairs can enhance the mobility and safety of older adults. This study aims to prototype and evaluate a walking assistive device for older adults to carry objects and support them during stair navigation, thereby increasing their balance.

2 Related Work

Most factors related to falls in older adults during stair navigation can be divided into the built environment and physiological condition [11]. Regarding the built environment, the steps’ size, the staircase’s lighting, and other physical conditions in staircases may contribute to falls among older adults. Physiological conditions, such as decreased muscle mass, strength, bone density, and balance, can also lead to falls in older adults [5, 11, 22]. The related work in this paper mainly classifies devices assisting older adults in navigating stairs (Table 1).

Table 1. Devices helping elderly people ascending and descending stairs
Full size table
Fig. 2.
figure 2

Images of devices in Table 1

Full size image

  • Built environment. Devices related to built environment issues, such as lifts (see Fig. 2(a)) and stair lifts(see Fig. 2(b)(c)) [2, 6], fundamentally change how people move between floors. Lifts, while efficient, require significant modifications to the existing structure of apartment buildings and are costly to install. As a cost-effective alternative to lifts, stair lifts are relatively easy to install, but stairlifts’ main body and track occupy space within the staircase [24], which impede evacuation routes and potentially pose a security hazard. Therefore, there is a clear need for this research to focus on designing a device that is both smaller in scale and lower in cost, effectively addressing these limitations.

  • Physiological condition. Devices designed to address the weakened physiological conditions of older adults encompass sliding auxiliary bars, exoskeletons, walkers, and stair-climbing wheelchairs (see Fig. 2(d)(e)(f)(g)) [3, 10, 16, 17]. The sliding auxiliary bar [1, 26], exoskeletons, and walkers could be independently used by older adults, while stair-climbing wheelchairs require professional assistance. None of these devices consider scenarios where older people use stairs by themselves with belongings, which is a frequent and practical aspect of their daily lives. Furthermore, there are the health benefits of stair usage under safe conditions, including enhancement of limb muscle strength [18], the cardio-pulmonary function [20], and reduction in metabolic syndrome risks [28].

Consequently, the walking assistive device in this study should have the following characteristics:

a.:

cost-effective and small-scale;

b.:

load-bearing;

c.:

stable support during stair navigation.

3 Prototype of StairMate

The comprehensive system of the walking assistive device, namely StairMate, comprises three subsystems (see Fig. 3): main section, power section, and balance section.

Fig. 3.
figure 3

Design of StairMate (a) and its prototype(b)

Full size image

The main section serves as the primary interface for elderly users, including four modules: aluminum frame, load-bearing unit, auxiliary handle, and signal input. Using the handrail as the sliding track, the aluminum frame allows for adjustment to different handrail sizes. The load-bearing module carries items, and the auxiliary handle, integrated with the aluminum frame, provides stable support and can be folded to save space. The Signal input module includes obstacle avoidance (IR sensor), direction (single control switch), and start-stop signals (crash sensor). The obstacle avoidance signal at both ends of the aluminum frame halts movement if an obstacle is detected. The direction signal corresponds to the two gears of the single control switch, and the start-stop signal is controlled by a crash sensor built into the auxiliary handle.

Fig. 4.
figure 4

Signal transmission of the main section and power section

Full size image
Fig. 5.
figure 5

The load-bearing module

Full size image
Fig. 6.
figure 6

(a)Two states of the handle; (b)Usage of the handle

Full size image

In Fig. 4, we utilize a pair of BLE NANO boards integrated with Bluetooth 4.2 to facilitate signal transmission between the main and power sections. The BLE-NANO (master) is positioned in the main section, while the BLE-NANO (slave) is in the power section. The power section receives directional and start-stop signals from the main section and propels the main section to move vertically through the actuator. Enhancing StairMate’s stability, the balance section is situated under the handrail.

StairMate implements the following three features:

  • Cost and Scale: To achieve the purpose of cost-effectiveness and small scale, StairMate is elaborately designed and implemented by three subsystems. The cost primarily involves expenses and modifications to existing staircases, which are limited to 120 US dollars. To cover a 10-centimeter-wide handrail, the size of the main section is adjusted to 20*26*37 cm.

  • Load-bearing: To accommodate three common categories of items carried by older adults while navigating stairs: small personal belongings, shopping bags, and shopping carts, the inner side of the main section features mesh pockets, hooks, and L-shaped aluminum rods designed (see Fig. 5).

  • Stable Support: To provide support for users’ upper limbs, we integrated a handle, which is ergonomically designed and facilitates an easy single-handed grip for elderly individuals, on the stairway side of the main section. It folds to a width of 5cm when not in use (see Fig. 6(a)) and extends to 25cm during operation. During both ascent and descent, the handle remains horizontally positioned in front of the elderly user’s body (see Fig. 6(b)). It provides upward force during ascent, reducing effort for older adults’ lower limbs. It is a safety measure during descent, preventing forward-leaning and potential falls.

4 Evaluation

To assess StairMate’s efficacy in assisting elderly users during stair navigation, we conducted an experiment analyzing users’ physical fatigue and lower limb load. The prototype was installed at Tsinghua Shenzhen International Graduate School, with approval from the Ethics Committee of Artificial Intelligence at Tsinghua University (THU-03-2023-0012).

4.1 Participants and Experiment Setup

According to previous research [1, 12], 12 elderly participants (7 females, 5 males, mean age = 64.67 ± 3.03 years old, height = 1.63 ± 0.07 m, weight = 105.00 ± 23.11 g) were recruited. All participants had basic literacy skills and experience with smart devices. During recruitment and the experiment, they reported being in good physical condition without injuries affecting walking or navigating stairs.

Fig. 7.
figure 7

The procedure of the experiment

Full size image
Fig. 8.
figure 8

Images of participants in three rounds of the experiment (participants’ faces obscured to protect their privacy)

Full size image

4.2 Procedure and Experimental Design

The whole experiment is divided into three steps (see Fig. 7):

  • Step A: Pre-Experiment. Before the experiment, each participant was briefed on the procedure by the researcher, signed a consent form, and completed a basic information survey. Participants were instructed to complete tasks at an appropriate pace, with the option to withdraw if they felt uncomfortable.

  • Step B: Experiment. The experiment comprised three rounds of tasks (see Fig. 8) to evaluate participants’ physical fatigue and lower limb load.

    • Round 1: Participants ascended and descended 10 steps without any assistive device or load.

    • Round 2: Participants repeated the task from Round 1 but carried a 5kg load.

    • Round 3: Participants hung a 5kg shopping bag on the prototype’s hook and used the walking assistive device for stair ascending and descending.

    We adopted physiological measuring equipment, including sensors for heart rate (HR) and surface electromyography (sEMG) (see Fig. 9(a)). The experiment began with a 5-minute seated rest to measure resting HR, with a minimum 5-minute rest between rounds to restore HR to resting levels. According to Geiger et al. ’s experimental research [9, 15], we chose three muscles contributing to stair navigation for the measurement: Anterior Tibialis, Gastrocnemius, and Rectus Femoris (see Fig. 9(b)). The decrease in sEMG activity will directly reflect the assistance of the prototype during stair navigation for older adults.

  • Step C: Post-Experiment. After the experiment, the researcher conducted semi-structured interviews with participants to gather insights on their experiences using StairMate and obtain suggestions for future development.

Fig. 9.
figure 9

Sensors’ placements (a) and measured muscles (b)

Full size image

4.3 Data Analysis Method

Relative HR effectively quantified individual physical fatigue, minimizing variations between individuals [12]. Lower relative HR indicated lower exercise intensity and reduced fatigue. We adopted equation (1) for calculating relative HR (\(HR_{r}\)), where \(HR_{a}\) (bmp) represents real-time HR, and \(HR_{c}\) (bmp) represents the participant’s resting HR:

$$\begin{aligned} HR_{r} = \frac{HR_{a}}{HR_{c}} \end{aligned}$$
(1)

sEMG activity directly reflects the effectiveness of StairMate in assisting stair navigation. The prototype enhances the balance by lightening the load on lower limbs during ascent and descent. Visual charts and statistical methods were used for analysis, including the Kruskal-Wallis H Test [14] and Spearman’s rank correlation coefficient [21].

5 Results and Discussion

5.1 Results of Main Experiment

Relative HR’s Changes. In three rounds, Fig. 10 compares relative HR during ascending and descending progress.

  • In Rounds 1 and 2 (without assistance), relative HR increases with the progress, indicating participants’ fatigue increases when ascending and descending stairs.

  • In Round 2 (bearing load), both the magnitude and value of relative HR exceed those in Round 1, indicating more significant exercise intensity and fatigue.

  • In Round 3 (with assistance), participants experience a rapid increase in relative HR midway through the process, exceeding the level in Round 2. However, the relative HR subsequently slows down and stabilizes, revealing a distinctive pattern.

Fig. 10.
figure 10

Changes in Average relative HR during (a) ascending progress and (b) descending progress

Full size image

Relative HR’s Kruskal-Wallis H Tests. To discern distinct trends in relative HR across three experimental rounds, the ascending and descending process was divided into five stages: Stage 1 (0–20\(\%\)), Stage 2 (20\(\%\)–40\(\%\)), Stage 3 (40\(\%\)–60\(\%\)), Stage 4 (60\(\%\)-80\(\%\)), and Stage 5 (80\(\%\)–100\(\%\)). The Kruskal-Wallis H test revealed significant variations in relative HR across all stages of stair ascent and descent for elderly participants in three experimental rounds (Ascending Progress: 2170.54 < H < 5238.95, p < 0.001; Descending Progress: 2205.26 < H < 5332.44, p < 0.001). These differences primarily stem from distinctions in each experimental task, surpassing individual participant variances.

Fig. 11.
figure 11

Average relative HR in different stages during (a)(b)(c) ascending progress and (d)(e)(f) descending progress

Full size image

Boxplots further highlight noticeable differences in participants’ relative HR among the three experimental rounds across the five stages (see Fig. 11). Relative HR in Round 3 exhibits the largest variance in Stage 1 and 2, and the highest median in Stage 2 and 3. However, since Stage 3, relative HR in Round 3 remains stable or declines slightly, indicating the impact of StairMate on the HR of elderly individuals during stair navigation.

sEMG’ S Changes of Muscles. sEMG data were processed to illustrate the smooth envelope of sEMG activity across the three experimental rounds for three muscles: Tibialis Anterior, Gastrocnemius, and Rectus Femoris. In Fig. 12, the comparison of sEMG activity during three rounds of stair-related tasks, both ascending and descending, reveals a notable reduction in the load on selected muscles in Round 3 compared to Round 1.

  • During ascending progress in Round 3, the reduction of activity level: 75.99 ± 11.63\(\%\) for tibialis anterior, 74.90 ± 9.77\(\%\) for gastrocnemius, and 63.49 ± 15.06\(\%\) for rectus femoris.

  • During descending progress in Round 3, the reduction of activity level: 75.22 ± 11.72\(\%\) for tibialis anterior, 67.48 ± 9.76\(\%\) for gastrocnemius, and 59.61 ± 11.45\(\%\) for rectus femoris.

Fig. 12.
figure 12

Changes in sEMG activity of (a)(b) Tibialis Anterior, (c)(d) Gastrocnemius, and (e)(f) Rectus Femoris during stair ascent and descent. Shaded areas represent 95\(\%\) confidence intervals

Full size image

The reduction indicates a significant increase in balance during stair navigation when using StairMate, a sentiment widely acknowledged by participants.

Relationship Between Relative HR and sEMG. This study employs Spearman’s correlation coefficient for analysis (see Fig. 13) to validate the impact of increased relative HR on sEMG activity and explore potential reasons.

  • According to Fig. 13(a)(b)(c)(d), sEMG activity of each muscle shows a significant correlation with relative HR during the ascending and descending progress in Round 1 and Round 2 (without assistance), with values of Spearman correlation coefficients exceeding 0.05 (p < 0.001).

  • According to Fig. 13(e)(f), in Round 3, Spearman correlation coefficients between each muscle’s sEMG activity and relative HR are substantially lower compared to Rounds 1 and 2. This shows that using StairMate significantly weakens the association between HR and sEMG.

Fig. 13.
figure 13

Spearman correlation coefficient between relative HR and average sEMG activity in three experimental rounds: (a) Round 1-Ascending, (b) Round 1-Descending, (c) Round 2-Ascending, (d) Round 2-Descending, (e) Round 3-Ascending, and (f) Round 3-Descending

Full size image

5.2 Discussion

  • Novel Device’s Impact on Older Adults’ HR Changes. The experimental results indicated an initial significant rise in relative heart rate (HR), followed by a stabilization phase. Using the newly developed walking assistive device might impose a psychological burden on the elderly, increasing their nervousness and concentration compared to regular tasks in Rounds 1 and 2. As the duration of use extends, participants gradually adapt, and their relative HR stabilizes.

  • Data Collection of Older Adults’ Movement. In this study, we measured muscles’ sEMG activity by sEMG sensors. However, some participants felt wearing unusual sensors discomforting, affecting their natural stair movements. Employing discreet methods such as the Vicon motion capture system could minimize interference, ensuring better user comfort and data accuracy.

6 Limitations and Prospects

Regarding the prototype of StairMate in the experiment, we conducted semi-structured interviews between researchers and participants to gain more scientific insights of the limitations and prospects of this research (Table 2).

Target User. 25\(\%\) of the participants in the experiment deemed the walking assistive device suitable for individuals aged 75 and above.

Capability.

Speed and Auxiliary Handle. Among the participants, 83\(\%\) found the StairMate’s movement speed to be slow and recommended the inclusion of a speed adjustment feature. Conversely, the remaining participants favored the slower speed, citing its role in reducing anxiety. Furthermore, 16\(\%\) of the participants expressed a wish for a longer handle that could be held with both hands, and all participants voiced a desire to integrate voice control functionality into the device.

Multi-modal Interactions. All participants preferred adding the voice control function while maintaining the current physical buttons because this will both simplify the interaction process and ensure the security of use.

Cognitive Load. All participants agreed that the functionality and interface of the StairMate were clear and easily comprehensible.

Social Attributes

Publicity. 83\(\%\) of the participants believed it could be permanently installed in residential staircases, while the rest of the participants suggested it could be further developed as a personal product.

Security. Only one participant expressed concerns about potential risks, especially if children mistake it for a toy.

Cost. All participants highlighted the significant cost advantage and a shared desire to minimize its impact on the living environment.

Table 2. Results of semi-structured interviews
Full size table

7 Conclusion

As preliminary research, this work effectively explores the feasibility and effectiveness of StairMate in assisting elderly individuals to navigate stairs. StairMate, featuring an innovative support mechanism and actuating system, tackles the issues related to cost and size, and it also addresses the need to carry objects while enhancing the balance of older adults.

The prototype of StairMate underwent a small-scale user experiment with elderly participants on a real staircase. Findings revealed an initial increase in participants’ heart rate (HR) when using StairMate, followed by stabilization or even a decrease in HR (p < 0.001). This contrasts with the upward trend observed in the HR of elderly individuals during stair navigation without the assistive device. Measurement of surface electromyography (sEMG) on participants’ lower limbs demonstrated a reduction of over 65\(\%\) in lower limb muscle activity during stair ascent and 55\(\%\) during descent with StairMate (p < 0.001). Additionally, StairMate attenuated the potential impact of HR changes on lower limb muscle activity during stair navigation. Based on semi-structured interviews, we gained suggestions for future improvements, including participants’ age, adjustable speed, handle length and height, integration of voice control features, and additional user instructions. The outcomes of this research significantly contribute to improving the balance of older adults during stair navigation, thereby enhancing accessibility in walk-up apartments.