This scoping review was conducted in accordance with the Joanna Briggs Institute guidelines for scoping research and reported following the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) [14] guidelines. The literature search was performed in January and February 2025. The study protocol was registered on the Open Science Framework platform (registration number: 10.17605/OSF.IO/CTPF4).

The search encompassed publications from January 2015 to December 2024, using a predefined search string. The search string for the PubMed database is provided in Textbox 1. The search strings for 4 other databases can be found in Multimedia Appendix 1. An initial search was performed in PubMed, SSCI, CINAHL, IEEE, and Cochrane Library, which informed the development of a comprehensive search strategy subsequently applied across all sources. Final searches were conducted in the 5 databases. After removing duplicates and screening titles and abstracts, full-text reviews were conducted.

Articles published in English in peer-reviewed journals that focused on smart walkers and independent walking users were included. Exclusion criteria encompassed secondary analyses (eg, reviews), smart walking aids with 2 wheels or fewer, or studies where the term “walker” referred to nonsmart devices.

The Rayyan (Rayyan Systems Inc) systematic review management software was used for study selection. Four independent reviewers (HB, MW, SS, and NS) initially screened titles and abstracts to exclude articles that did not meet the inclusion criteria. Subsequently, full-text screening was performed independently by 2 of the aforementioned reviewers. Discrepancies were resolved through discussion with a third reviewer until consensus was achieved.

Data from each included study were extracted independently by HB, MW, NS, and SS using a structured data extraction framework. Any conflicts were resolved through discussion with the principal investigator (NS). The reviewers’ professional backgrounds include doctoral (PhD) degrees in physiotherapy and health sciences, with experience in health technologies related to mobility in older patients, as well as physiotherapy care and therapy for older patients. They also have research experience and have conducted studies on health technologies that address mobility. Furthermore, the researchers are sports scientists and nursing professionals with expertise in geriatric care for older clients and 10 years of experience in clinical gait analysis.

Data extracted from each publication included general publication details (year, author, journal, and country), study population and setting, and characteristics of smart walkers (hardware, software, and technological features such as sensors and feedback mechanisms), as well as technological features (sensors and feedback mechanisms). These data are presented in the Results section.

As per the scoping review methodology, the quality of included studies was not appraised.

An initial pool of 800 articles was identified for screening (duplicates removed n=59, 7.4%). After title and abstract screening, 666 (83.3%) articles were excluded based on the predefined criteria related to theme and publication period. A total of 75 (9.4%) records were sought for retrieval. Although the original authors were contacted, 2 (0.3%) studies could not be included due to missing full texts. The remaining 73 (9.1%) articles underwent full-text review, leading to the exclusion of 29 (3.6%) studies due to study design, theme of investigation, and design of object or period. Ultimately, 44 (5.5%) articles met the inclusion criteria. The entire selection process is summarized in the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flowchart [15] (Figure 1).

Of the 800 articles screened, 44 (5.5%) met the predefined inclusion criteria. Geographically, the research was predominantly conducted in Europe (n=19, 43.2%; Figure 2).

Most of these studies were research reports (n=36, 81.8%). Sample sizes varied considerably across the included studies, ranging from a minimum of 1 participant to a maximum of 43 participants. A total of 5 (11.4%) studies did not report the number of participants. Most evaluations of smart walkers were conducted exclusively or partially in asymptomatic populations (n=29, 65.9%), with half (n=22, 50%) of the studies involving younger adults. A total of 14 (31.8%) studies involved older adults (6 studies without specifying age group and 1 without any person; Table 1).

The studies were conducted across various settings. Investigations involving smart walkers were most frequently conducted in laboratory-based environments (n=30, 68.2%).

Four groups of participants were evident in the articles included: symptomatic children or younger adults (n=5, 11.4%), asymptomatic younger adults (n=19, 43.2%), symptomatic older adults (n=8, 18.2%), and asymptomatic older adults (n=9, 20.4%). Several articles did not specify the age (n=6, 13.6%) or health status (n=5, 11.4%) of participants. A total of 4 (9.1%) articles provided no information on the age and health status. As there were studies with multiple participant groups, the number of participant groups exceeded the number of articles. The group of participants consisting of asymptomatic younger adults was the most frequently studied group across the articles (n=19, 43.2%; Table 1, Figure 3).

The included articles described various types of sensors and functions that constitute the extent of smart technology integrated into modern walkers. The sensors used were fundamental to the operation of these intelligent walkers. They encompassed a range of sensor types, including camera systems (2D cameras, depth cameras, and infrared cameras), optical sensors that used laser lights (2D laser scanners and light detection and ranging [LiDAR]), force and load sensors, ultrasound sensors, and inertial measurement units (IMUs; Table 2).

Among the sensor modalities reported, camera- and LiDAR-based sensors were the most prevalent, accounting for half of the implementations. Camera sensors were described either as standard 2D cameras (n=11, 25%) [10,17,23,24,27,28,31,33,42,44,56] or depth cameras. Depth cameras were used for multiple purposes, including gait analysis (n=3, 6.8%) [16,41,55] and intention recognition (n=3, 6.8%) [23,42,52], by capturing gaze direction. Additionally, depth cameras were used for collision avoidance through object detection (n=4, 9.1%) [17,23,25,42], and 1 (2.3%) study specifically described the use of depth cameras for fall prediction purposes [24]. LiDAR-based sensors can be categorized according to their primary functions: gait analysis (n=11, 25%), collision detection (n=9, 20.5%), and navigation (n=5, 11.4%).

In addition, load sensors (n=10, 22.7%) and ultrasonic sensors (n=11, 25%) were among the most frequently cited sensor modalities in the literature. Furthermore, load and force sensors were typically integrated into the handlebars of smart walkers (n=10, 22.7%) [10,17,18,23,39,44,45,47,48,56]. These sensors were described as “the simplest approach to onboard sensor-based gait analysis” [47] and were used to assess the user’s balance during ambulation [17]. Zhao et al [39] further detailed handgrip pressure sensors that can potentially prevent falls by detecting abnormal pressure or rapid pressure changes, indicating a fall risk. For instance, the smart walker developed by Ballesteros et al [18] incorporated force sensors in the handlebars alongside encoders to measure wheel rotation.

Beyond handlebar-mounted sensors, force sensors were also embedded elsewhere in the walker’s structure, such as in the frame, forearm supports, or chest pads (n=7, 15.9%) [16,32,46,51-54]. Sensors placed within the frame measured the user’s weight load, providing critical data for calculating body weight support (n=3, 6.8%) [46,51,52]. Additionally, Machado et al [54] and Jiménez et al [36] reported on force sensors located in forearm supports, which were used to infer the user’s intention to move, thereby enabling more responsive and adaptive assistance.

Ultrasound sensors were described in a quarter of the articles (n=11, 25%) and were used for collision avoidance [10,13,27,28,30,31,33,41,44,45,48]. IMUs were described with a similar frequency (n=12, 27.3%) [10,26,27,30,33,36,39,44,45,54-56]. An analysis of their functions revealed that they were mainly used to determine position and orientation, measure accelerations, record angular velocities, or support odometry sensors.

With regard to sensory-based feedback generated, the following distinctions can be reported: user feedback, which provides relevant information on, for example, navigation details or collision warning during the use of the smart walker (Table 3), or observer feedback, which, for example, provides valuable additional information such as gait-related or system-relevant measurement data (Table 4) to the observer or therapist.

This scoping review was conducted to provide a comprehensive overview of the current application settings, sensors, and functionalities of smart walkers.

Various sensors, such as LiDAR, depth camera, and 2D laser range finder, which could be used for a clinical assessment of the gait pattern or for measuring balance, have been integrated into smart walkers. Analysis of the intended use of sensors revealed a wide variety of applications. For instance, IMUs were mainly used to determine position and orientation, measure system accelerations, record angular velocities, and support odometry. The gait analysis functionality and the validity of the measured parameters depend on the sensor types used and their placement on the smart walker.

Analysis of the included studies indicates that validation data for gait-related sensor technology are often missing or only partially reported. Only 8 (22.9%) of the 35 studies measuring gait-relevant data used reference systems such as the one used in the study by Werner et al [34], IMU-based systems (MTw Awinda, Xsens Technologies, B.V.) [48], or a marker-based camera system operating at 120 Hz (Vicon motion analysis system) [44]. These studies provided the most methodologically robust evidence regarding measurement accuracy and are best suited for comparison with established clinical measurement systems. An additional 17 (38.6%) studies reported some form of validation but without a reference standard, mostly in the form of internal error measures (eg, root mean square error, difference metrics, and trial comparisons). While these data provide valuable technical insights, they allow only limited conclusions regarding external validity. Some articles provided no information on the validation of gait sensors, which is partly due to the studies’ focus. In many studies, the primary objective was different, such as navigation, safety, human-robot interaction, or technology design, meaning that gait analysis was only of marginal or no relevance. Overall, there is a clear gap between technical innovation and methodological transparency: although smart walker systems increasingly integrate complex sensor sets, validation against established gold standards remains uncommon. This lack of consistent validation limits the transferability of findings to clinical practice.

This review demonstrates that smart walkers that have been investigated in laboratory-oriented settings and among user populations are not representative of real-world applications. Most research has taken place in controlled laboratory environments and primarily involves asymptomatic and/or younger users of smart walkers. In contrast, older adults (representing an increasing global aging population) are key drivers of the expanding market for walking aids and walkers [58]. It is undisputed that younger adults have fewer functional limitations, better balance abilities, and overall exhibit fewer fall-related risk factors than older adults. As noted by Maranesi et al [59], research in this field should move beyond focusing solely on the technical functionality of gait devices and walkers. Our results illustrate a gap in effectively incorporating these devices into relevant settings, such as clinical environments or the daily lives of end users. This underscores the need for a comprehensive understanding of the physical, social, and cognitive needs of older adults. In 2025, a funded research and development initiative was launched in Germany to develop a prototype of a smart walker. The SmartRoll project adopts a user-centered study design, systematically identifying the needs of patients, professional caregivers, and physiotherapists [60]. These insights are subsequently evaluated with the target population—patients on an acute geriatric ward—as part of an integrated research and development effort. Additionally, the project involves the integration and testing of sensors capable of performing gait analysis in this vulnerable group of older adults. This scoping review indicates that the search for smart walkers in the clinical setting implies their use for research purposes within clinical environments. The relevance of smart walkers in clinical contexts is driven by the urgent need to address age-related skill decline [26] and mobility impairment [45]. This paper is embedded in a study that examines the current relevance of deploying smart rollators in clinical settings.

Older individuals who rely on assistive devices often do so due to underlying health conditions or following adverse events such as falls. Consequently, it is not surprising that users of walkers exhibit a fourfold increased risk of falls compared to individuals walking without such aids [61]. This population is particularly vulnerable during acute hospital stays. The use of smart walkers equipped with gait analysis and balance measurement functions could potentially be used in hospital settings to assess fall risk among patients. Indicators of fall risk are primarily identified through parameters that suggest an unstable gait balance and gait pattern. These parameters include, but are not limited to, foot angle at contact, step length, maximum foot height, and a high SD in step characteristics [1].

It should be noted that research comparing 2-, 3-, and 4-wheeled walkers is limited. Two-wheeled walkers often need to be partially lifted, requiring strength and coordination and affecting gait parameters. As the poor maneuverability and lower lateral stability of 2-wheeled walkers may pose additional risk factors [62], they were excluded from this review.

The smart walkers described in the literature were predominantly tested in asymptomatic and young populations. Developing smart walkers that explicitly address the specific needs of older adults—through systematic requirements analyses—and iterative testing throughout the development process remains an ongoing challenge. Integrating these steps as core components of research and development projects is essential to enhance clinical relevance.