Next Article in Journal
Tracking and Recognition of Multiple Human Targets Moving in a Wireless Pyroelectric Infrared Sensor Network
Next Article in Special Issue
Experience in Evaluating AAL Solutions in Living Labs
Previous Article in Journal
Angular Rate Optimal Design for the Rotary Strapdown Inertial Navigation System
Previous Article in Special Issue
WiSPH: A Wireless Sensor Network-Based Home Care Monitoring System
Article Menu

Export Article

Sensors 2014, 14(4), 7181-7208;

Smartphone-Based Solutions for Fall Detection and Prevention: Challenges and Open Issues
Department of Biomedical Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia
Centre for Innovation in Medical Engineering (CIME), Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia
Department of Computer Science & Engineering, Chittagong University of Engineering & Technology, Chittagong 4349, Bangladesh
Department of Medicine, Faculty of Medicine, University of Malaya, 50603 Kuala Lumpur, Malaysia
Author to whom correspondence should be addressed.
Received: 4 March 2014; in revised form: 10 April 2014 / Accepted: 11 April 2014 / Published: 22 April 2014


: This paper presents a state-of-the-art survey of smartphone (SP)-based solutions for fall detection and prevention. Falls are considered as major health hazards for both the elderly and people with neurodegenerative diseases. To mitigate the adverse consequences of falling, a great deal of research has been conducted, mainly focused on two different approaches, namely, fall detection and fall prevention. Required hardware for both fall detection and prevention are also available in SPs. Consequently, researchers' interest in finding SP-based solutions has increased dramatically over recent years. To the best of our knowledge, there has been no published review on SP-based fall detection and prevention. Thus in this paper, we present the taxonomy for SP-based fall detection and prevention solutions and systematic comparisons of existing studies. We have also identified three challenges and three open issues for future research, after reviewing the existing articles. Our time series analysis demonstrates a trend towards the integration of external sensing units with SPs for improvement in usability of the systems.
fall detection; fall prevention; smartphone; ubiquitous computing; pervasive computing; elderly

1. Introduction

Falls are defined as the inadvertent settling down of a body on the ground, floor or other lower level. The prevalence of falls is very common among the elderly and increases with age. The World Health Organization (WHO) reported that 28%–35% of people aged 65 years and above fall each year and the rate increases to 32%–42% for those over 70 years of age [1]. Those who are vulnerable to falls also include those suffering from neurological diseases (e.g., epilepsy and dementia), which commonly occur in older people. Individuals with epilepsy fall during seizure events due to loss of consciousness [2], while those with dementia are two to three times more likely to fall than individuals without cognitive impairment [3]. Living alone itself increases the risk of falls for community elders [4]. Falls can potentially cause severe physical injuries such as disabling fractures [5], and can reduce the independence of older individuals through dramatic psychological consequences [6]. If protective measures cannot be taken in the near future, the number of falls induced injuries is anticipated to double by 2030 [7].

Hence, early detection and treatment of falls are key strategies to be employed in reducing fall- related injuries and preventing their consequences, which include long laying periods (remaining on the floor for prolonged periods after a fall) leading to an increased risk of pneumonias, pressure ulcers and even death. The use of assistive devices for fall detection and prevention will help reduce its future burdens by preventing injurious falls, reducing the risk of long laying periods and admissions to nursing homes. Insights gained from research in this area by industry and academics will assist community, public health leaders and health care professionals in developing more efficacious intervention strategies to prevent or reduce falls, and its associated psychological, physical and economical consequences.

This past decade alone has seen a tremendous amount of research in the development of assistive devices for fall management. Researchers and industry mainly focus on two automatic fall management strategies namely, its detection and prevention. Typically fall detection systems help the elderly and their caregivers avoid the consequences of long laying periods by detecting falls, triggering notification alarms, sending messages and calling for help as soon as falls occur. Fall prevention systems are usually based on the assessment of the medical and behavioral histories of users in order to predict the possible risk of falls. Most of these fall management technologies consist of three common functional units: a sensing/data-acquisition unit, processing unit and communication unit. The accelerometer, gyroscope and camera are the most frequently used sensors in SPs, while Bluetooth and Wireless Fidelity (Wi-Fi) technologies are typically used for communication purposes. Various microcontrollers and wirelessly connected desktops or laptops are usually used for feature extraction and classification from the sensors' output signals. SP-based fall detection and prevention is attracting growing interest among researchers as state-of-the-art SPs come with built-in kinematic sensors (such as tri-axis accelerometers, gyroscopes, and magnetic sensors), high performance microprocessors, advance communication facilities (e.g., Wi-Fi and Bluetooth) and other sensors (such as camera, proximity sensor and microphone) [8]. In a recent survey, Igual et al. [7] have shown a new trend towards the integration of fall detection into SPs.

A variety of dedicated tools and methods have been proposed for fall management, but none of these solutions is universally accepted [9]. The SP however, is a very good candidate as this technology is widely accepted in daily life [10]. SPs are also more integrated than a dedicated monitoring device which reduces rejection due to the device's poor aesthetic value and intrusiveness [11]. For these and many other reasons, the number of studies on SP-based fall management has increased steadily in recent years. Currently, to the best of our knowledge, there has been no published review specifically on SP-based fall detection and prevention systems. Although, there are some relevant review articles [7,12,13], there are none that focus exclusively on SP-based fall detection and prevention systems.

This paper provides a comprehensive and integrative literature review of SP-based fall detection and prevention systems. The usability and overview of the general architecture of SP for fall management with several new dimensions including a comprehensive taxonomy of the SP-based fall management systems is presented. A critical analysis of the methods proposed so far and a comparison of their features, strengths and weaknesses is made. This includes the identification of the issues and challenges found with the SP-based fall management systems.

Throughout this paper, the terms fall prediction and fall prevention are used interchangeably because SP-based fall prevention systems attempt to prevent falls by predicting the imminent fall events. Unless otherwise stated, accelerometer and gyroscope represent tri-axial-accelerometer and tri-axial-gyroscope respectively. A SP is a combination of a normal mobile phone and a Personal Digital Assistant (PDA) [14]. Ordinary mobile phones and PDAs have less functionality than SPs and cannot be considered as SPs. Therefore, PDA or pocket Personal Computer (PC)-based [15,16] and ordinary mobile phone-based [17] solutions are excluded from our comparative study.

This paper is organized in five sections: Section 2 discusses the basic architecture and taxonomy of SP-based fall detection and prevention systems. A comparative analysis of the reviewed articles is provided in Section 3, illustrated by tables and graphs. Section 4 highlights the challenges of the SP- based solutions and also discusses some open issues. Finally, the concluding part—Section 5—points out important observations and areas that need further research.

2. SP Based Fall Detection and Prevention

Although a fall detection system was first introduced by Hormann in the early 1970s [18,19], the history of SP-based fall detection is far shorter. The first smartphone (“Simon”) was first introduced by IBM in 1993 [20] and subsequently, various sensors useful for human activity monitoring were integrated into SPs. Hansen et al. [21] used the SP camera for the first time in 2005 for fall detection. The SP is also used for fall prevention [22], but instead of active fall prevention, most of the solutions proposed were based on standard falls risk assessment tests Timed Up and Go (TUG) and Get Up and Go (GUG).

2.1. Basic Architecture

Fall detection and fall prevention systems have the same basic architecture as shown in Figure 1. Both systems follow three common phases of operation: sense, analysis and communication. The basic difference between the two systems lies in their analysis phase with differences in their feature extraction and classification algorithms. Fall detection systems try to detect the occurrence of fall events accurately by extracting the features from the acquired output signal(s)/data of the sensor(s) and then identifying fall events from other activities of daily living (ADL). On the other hand, fall prevention systems attempt to predict fall events early by analysing the outputs of the sensors. Data/signal acquisition, feature extraction and classification, and communication for notification are the necessary steps needed for both fall detection and prevention systems. The number and type of sensors and notification techniques however, vary from system to system (some examples are shown in Figure 1). In conventional systems, discrete hardware components are used for the implementation of each unit, whereas in SP-based systems, all required units may already be in-built within a state-of-the-art SP.

2.1.1. Phase 1: Sense

This is the first phase of any fall detection and prevention system and in this phase, appropriate physical quantities are sensed or measured using suitable sensors. Modern SPs come with various built-in sensors and that is one of the vital reasons for choosing SPs as an alternative of conventional fall detection and prevention tools [9]. Moreover, the users of SP-based systems are more likely to carry SP (with built-in sensors) throughout the day since mobile phones are seen as indispensable in daily living. This is in contrast to the users of the conventional systems who may forget to wear the special microsensors [17]. Many types of sensors are now available for SPs. These include accelerometers, gyroscopes, temperature sensors and magnetic field sensors [2325]. These sensors are used in various ways in SP-based solutions. Some solutions use only one of the abovementioned SP sensors for fall detection or prediction [26,27]. According to our survey, the tri-axial accelerometer is the most used sensor for SP-based fall detection and prevention. SP-based solutions can use combinations of two or more SP sensors during this sensing phase [22,28]. Some solutions use both SP sensors and external sensors for detection and prediction of falls events [29,30]. It is also possible to use SPs for analysis and/or communication but not for sensing [31,32]. An uncommon type of solution was proposed by Hansen et al. [21]. They used a SP for sensing only, and external systems to perform the analysis and communication tasks.

2.1.2. Phase 2: Analysis

After measuring the physical quantities by using sensors, obtained signals/data should be analysed. In this phase, the significant features are extracted from the sensor's outputs and preliminary decisions are made by classifying and analysing those extracted features. Most SP-based solutions, especially solutions for fall detection, use a Threshold-Based Algorithm (TBA). The most vital reason for choosing TBAs is that these algorithms are less complex and hence require the lowest computational power [9], which helps to reduce battery power consumption [33]. In order to make preliminary decisions about a potential fall event, these algorithms usually compare the sensor's output(s) with predefined threshold value(s). Threshold-based algorithms may use more than one threshold [27] and threshold value(s) could be predefined (fixed) or adaptive. It should be noted that the adaptive threshold values are not calculated dynamically while using the system. Instead, users introduce some physiological data and the system obtains the corresponding threshold that is not re-calculated during the system operation. The algorithm proposed in [34] uses an adaptive threshold which changes with user-provided parameters such as: height, weight and level of activity.

As mentioned earlier, most solutions employ the tri-axial accelerometer for sensing which measure simultaneous accelerations in three orthogonal directions. Threshold-based algorithms use these acceleration values for calculating Signal Magnitude Vector by using the following relation:

Signal Magnitude Vector = | A x | 2 + | A y | 2 + | A z | 2
where Ax, Ay, and Az represent tri-axial accelerometer signals of the x, y, and z-axis respectively. If the value of signal magnitude vector for a particular incident exceeds a predefined threshold value, then the algorithm primarily identifies that incident as a fall event. To make the final decision, algorithms usually depend on the next communication phase.

The processing power of SP processors has increased dramatically over the past few years. The computational power of the latest SPs has become comparable to that of former workstations [35] and, thus, even complex machine learning and statistical classification algorithms for fall detection and prevention can easily be implemented in SPs [36]. Zhao et al. [37] implemented three machine learning algorithms, namely C4.5, Decision Tree (DT) [38], Naïve Bayes (NB) Classifier [39] and Support Vector Machine (SVM) [40], on SPs and compared their recognition accuracy. He and Li [8] employed a combined algorithm of Fisher's Discriminant Ratio (FDR) criterion and J3 criterion [41] for fall detection. Majumder et al. [22] applied Hjorth mobility and complexity [42] for classifying gait and hence developed a fall prevention system. Some solutions [21,43] include external sensors and processing units, using the SP for sensing and/or communicating with the users and/or their caregivers.

2.1.3. Phase 3: Communication

Depending on the sensor's responses from the first phase, preliminary detection or prediction of falls events is performed by algorithms in the second phase. Whenever a SP-based solution detects or predicts a fall event, it communicates with the user of the system and/or caregivers. Most fall detection solutions carry out this communication phase in two steps. In the first step, the system attempts to obtain feedback from the user by verifying the preliminary decision and thus improve the sensitivity of the system. The second step depends on the user's response. If the user actively rejects the suspected fall, then the system restarts. Otherwise, a notification is sent to caregivers to ask for immediate assistance. Some systems may not wait for user's feedback and will immediately convey an alert message to the caregiver [44,45]. Rather than requesting feedback, fall prevention systems generally alert the users about their imminent fall. Moreover, instead of alerting the users, fall prevention systems can also activate other assistive systems (e.g., wearable airbag [16,4648], intelligent walker [49,50], intelligent cane [51,52], intelligent shoe [53], etc.) for protecting the user from the adverse effects of falling.

User's feedback can be collected automatically by analyzing the sensor's output. For example, the algorithm proposed by Sposaro and Tyson [34] generates the final decision by automatically analyzing the difference in position-data before and after the suspected fall event. Other systems demand manual feedback from the user. Requests for the user's feedback can be attempted by using the external speakers on the phone and requesting a vocal or keypad response from the user [21]. Combinations of alarm systems and graphical user interface of SPs are also used for collecting the feedback of the user [9,54]. After requesting a response from the user, the system waits for a pre-defined period (typically ≤ 1 min). If the user does not respond within that time, the system will consider the event as a fall. Fall detection systems may fail to detect a real fall event automatically. In such cases, some systems provide help (or panic) buttons and thus allow users to seek help manually [55].

Smartphone-based systems generate several types of notifications to seek help from caregivers or for forewarning the users about an imminent fall such as audible alarms [56], vibrations [22], Short Message Service (SMS) [34,43,57], Multimedia Messaging Service (MMS) [8,27], and even automatic voice calls [21,57]. E-mails and Twitter messaging have also been described [2]. Notification messages may contain information on time [27], Global Positioning System (GPS) location (coordinates) [27,29,57], and location map [2,26,58]. SP-based solutions can also support streaming of phone data from microphones and cameras for further analysis of the situation [21].

2.2. Taxonomy

This section presents a detailed taxonomy of SP-based fall detection and prediction systems with respect to the three different phases of operation: sense, analyze and communicate. Here we focus on the categorization of various attributes/aspects of SP-based solutions for fall detection and prevention. The aim of this taxonomy is to provide a complete reflection of the properties of existing as well as possible SP-based solutions. The correctness and completeness of the taxonomy will be reflected upon in Section 3.

Figure 2 illustrates the taxonomy of SP-based fall detection and prevention technologies based on their sensing mechanism and sensor placement. Existing solutions are represented with rectangles, while rounded rectangles represent possible solutions that have not previously been reported to identify areas for future research. SP-based solutions can be categorized into two types: context-aware and body worn. With context-aware systems, the user should not wear any sensor or system. Sensors are placed in the surrounding and the user can move freely, but within the catchment areas of the sensors. Though, the main advantage of context-aware systems is that the person does not need to wear any special device, their operation is limited to those places where the sensors have been previously deployed [59]. No such SP-based context-aware solution has been found. All the SP-based solutions, proposed so far, are body worn systems and users are required to keep their SPs close to their body. This type of solution can be further classified according to the existence of external sensor(s)/system(s) and the placement of the SP.

Smartphone-based solutions can also be categorized on the basis of algorithms used in the analysis phase. Figure 3 presents the taxonomy of SP-based fall detection and prevention algorithms. Due to the lower processing capacity and low energy storage capacity of batteries in SP compared to desktop or laptop computers, SP-based solutions mostly use TBAs for the detection or prediction of falls events. Machine learning algorithms are also attracting research interest because of the improved processing and battery capacities of newer, high-end, SPs.

Existing and potential SP-based fall detection and prevention systems communicate with the users, caregivers or assistive systems by sending alert signals, obtaining user or system feedback or activating assistive systems. The taxonomy of communication patterns in SP-based fall detection and prevention is shown in Figure 4. Rectangles and rounded rectangles hold the same meaning as in Figure 2. Detection systems communicate with the users to obtain feedback, whereas prediction systems communicate to alert them about their possible forthcoming falls. Prediction systems are only concerned with pre-fall data, but detection systems deal with pre-fall, post-fall and intermediate data. Finally, detection systems notify caregivers of fall events and ask for help, whereas prediction systems attempt to prevent impending falls with the help of other assistive systems. Some SP-based solutions require external sensing units that may or may not have built-in processors. These external units may transmit either raw data or results after primary analysis. No article has been found, that uses assistive system and/or external processing unit for implementing SP-based fall prevention solution.

3. Comparative Analysis

In the reviewed articles, the authors commonly report their main objective (detection/preventing), usability (sensor placement & type), the SP operating systems, algorithm novelty, efficiency (sensitivity and specificity) and notification techniques. For comparison we focused on those features, which are inevitable or have comparatively more variants. Other features have been discussed separately. This section compares existing works based on their functional and architectural properties and quantitative properties.

We included journal articles and conference proceedings published on SP-based fall detection and fall prevention. Advanced Boolean searches are conducted, with no time limit, in MDPI, IEEE Xplore, PubMed, Web of Knowledge and Google Scholar with the search condition: “Find articles with all the words {keyword1 AND keyword2} anywhere in the article”. The keyword “smartphone” is always inserted as keyword1 with any one of the other three keywords: “fall detection”, “fall prevention” and “fall prediction”. Each keyword is inserted within double quotation marks and two keywords are separated by a Boolean operator AND. Additional articles are identified from the cross-referencing from these articles. A total of 578 articles are matched our search criteria. Among these articles, 51 articles included some experimental results or pioneering investigations on SP-based solutions for fall detection and fall prevention and are selected for further review. The remaining articles were excluded as they have used these keywords for other purposes such as, use of their proposed systems, references, and examples.

3.1. Functional and Architectural Comparison

Common built-in sensors of recent SPs and their corresponding functions are shown in Table 1. Examples of fall detection and prevention or related solutions (SP-based or non-SP-based), which use similar dedicated sensors, are also included, to identify potential new areas for research.

3.1.1. SP-Only Systems

Depending on the uses and placement of sensors the SP-based solutions are categorized into two major categories: context-aware systems and body-worn systems (see Figure 2). Table 2 summarizes and compares the important features of existing SP only systems. In this table the articles are organized chronologically.

3.1.2. Smartphones with Other External Systems

Table 2 shows that most of SP-only systems demand fixed placement of SPs, but this is considered as a usability constraint, because not all people carry their mobile phones in a fixed position [31]. Moreover, sensors in SPs usually have much lower resolutions than dedicated sensors [33]. Body-worn systems can also use external sensing and processing units together with SPs to overcome these two constraints. Some of these external units are used only for sensing or measuring physical quantities [31,32]. These units will transmit raw data to the SP, and then the SP will perform feature extraction, classification and notification tasks. External units can also perform the feature extraction and classification tasks with the help of attached microcontrollers. Such units will communicate with the SP for the communication step. Moreover, these external units will minimize the computational load and wireless communication burden of the SP and reduce battery consumption. External components, which are used in various SP-based fall detection and prevention solutions, are listed in Table 3.

Features of SP-based fall detection and prevention solutions, which employ external system(s) along with SPs, are summarized in Table 4. Smartphones with other external systems can be subcategorised, based on three phases of operations, into four types as shown in Figure 2. Such solutions can utilize SP for all of the three phases of operations while employing external units for the sensing phase only. It is also possible to use SPs for only the sensing or communication phases, but such systems must use external microcontrollers for analysis. If the SP is only used for the sensing phase, then for acquiring less ambiguous signals, it is important to firmly attach the SP at a fixed position of the user's body, but not all users like to carry their SPs in a fixed location. In order to overcome this constraint, some solutions utilize SPs for both analysis and communication phase and an external sensor for the sense phase. Since the SP is mainly a communication device, using SPs for analysis phase only or for both sensing and analysis phases is not a better solution. Moreover, using SPs for sensing and communication phase is also an impractical solution, because that will demand excessive wireless communication and thus consume excessive battery power. We therefore omit the latter three options from our taxonomy and Table 4 also supports our decision.

3.2. Quantitative Analysis

This section presents some statistical and time series analysis based on the articles that have been compared in Tables 2 and 4. The most important feature, that is not included in these articles, is the performance or the correctness of the reviewed solutions. More than half of the articles [2,21,26,27,29,32,34,43,45,54,55,57,65,70,71,73,74,77,79,82,85,89,91,9395] do not declare the performance/accuracy of their systems, because these articles present very preliminary investigations on SP-based fall detection and fall prevention. The remaining articles, included in Table 5, discussed the performance of their proposed solutions but there were major differences between the evaluation techniques. Moreover, their test results were obtained by analysing simulated falls events, not true falls.

The existing solutions tried to detect and classify the falls events, risk of falls and other normal ADLs accurately. Usually, the performance of such solutions is examined based on the sensitivity, specificity and total accuracy [97]. Some articles [64,87] measured the performance of their proposed systems in a different way. They used the performance parameters: precision and recall [80] Some other articles measured the accuracy of their proposed systems, simply by finding the ratio of number of correctly identified cases and the total number of cases [58,92]. Same as fall detection systems, standard approach for describing accuracy of fall prevention systems has been through sensitivity (proportion of fallers correctly classified as high fall risk) and specificity (proportion of nonfallers correctly classified as low fall risk) [98]. Table 5 summarizes the declared performances of the SP based fall detection and prevention solutions.

Fifty-one SP-based solutions are compared in Tables 2 and 4 and forty-one (80%) solutions used SP with the Android operating system. The Android platform is preferred [8,33,83] as it is an open source framework designed for mobile devices [34,78,89]. Other SP operating systems which have been used in fall detection and prevention solutions include iOS (8%) [22], Symbian OS (6%) [64] and Windows Mobile (4%) [57]. One paper (2%) did not report the SP operating system they used.

The accelerometer was used in all the reviewed solutions and the GPS receiver is the second most commonly used sensor (42%) followed by the gyroscope. In addition we have performed a time series analysis on SP based fall detection and prevention solutions and the outcome is shown in Figure 5. This line chart shows a comparison of the numbers of studies on SP-only solutions with other solutions having a combination of SP and external devices. In the past few years, though the number of studies on SP-only solutions are higher than those of other SP based solutions, the use of external devices in SP based fall detection and prevention systems is increasing gradually.

4. Discussion

Various benefits of using the SP as a pervasive fall management system have already been discussed [28]. Despite all these benefits, SP-based systems do face some critical challenges with certain issues remaining open to further research. Based on our extensive literature review, these challenges and open issues in SP-based fall management systems have been identified. This section presents the most relevant ones.

4.1. Challenges

4.1.1. Quality of SP Sensors

It remains doubtful whether the qualities of built-in SP sensors in existing SPs are adequate to produce fall detection and prevention systems with acceptable performance. The SP sensor that is used by all SP-only solutions is the accelerometer and the usual dynamic ranges of these built-in accelerometers are insufficient for accurate fall incident detection [31]. Acceptable dynamic ranges for accelerometers from ±4 g to ±16 g have been mentioned in previous publications (where, 1 g = 9.8 m/s2) [31,33,99]. Smartphones typically contain accelerometers with dynamic ranges of ±2 g or less [33], but higher dynamic ranges can be found in high-end SPs [81]. While choosing an SP for a particular application (fall detection or fall prevention) adequate attention should be paid to the quality of the sensors. Specifications of the sensors should satisfy the minimum requirements of the applications. Similar attention should be paid to all other SP sensors.

4.1.2. Energy Consumption and Battery Life

A major weakness of SP-based solutions is the limited battery life of SPs. Usually the battery life of an SP in normal use is about one day [33], but no SP battery will last more than a few hours with heavy usage [36,100]. The issue of energy consumption should therefore be considered when designing an SP-based system. The energy consumption or battery life of the SP is dependent on the number of sensors used [54], data sampling frequency [28,54], data recording time [75], features of algorithm [87] and mode (backend or frontend) of operation [26]. The battery life of a particular SP (Samsung Galaxy S II) was reduced from 30 h when only one sensor was used, to 16 h if three sensors were used simultaneously [54]. Majumder et al. [22] showed that an iPhone, which runs a machine learning algorithm, can run for at most 3 h with a fully charged battery. The battery life is also directly proportional to the recording time and activities of user [74].

While choosing the right algorithm, care should be taken to incorporate a minimal number of features, fewer features would decrease the usage of processor and would save energy [87]. Experimental results of [26] shows that the consumption rate of the battery per hour for foreground execution mode and background execution mode are 2.5% and 2.25% respectively. However, energy saving measures could adversely affect accuracy and usability.

4.1.3. SP Placement and Usability Issues

Smartphone-based fall detection and prevention systems are mostly designed for older people and individuals with neurodegenerative disorders. However, the acceptability of these solutions among older individuals has been suggested as a limiting factor [31]. People with intellectual disabilities also face great difficulty using the complicated interfaces of modern SP-based applications [101,102]. A recent study has revealed the myth that older people avoid new technologies is a fallacy [103]. Older people have been found to be willing to accept new technologies to support their independence and safety [104]. The older person may also prefer to have a single phone with self-contained fall detection functionality than to wear a separate fall detection device [22].

As mentioned earlier, all SP-only solutions use the accelerometer as a sensor which requires fixed placement of the SP. Various fixed positions on the body have been proposed, such as: the shirt pocket [73], waist [44] and trouser pocket [70]. This requirement limits the usability of SP-based solutions because not everyone caries their SP in a fixed position [31] and it may be difficult to convince them to do so [105]. In order to overcome this obstacle, researchers have proposed the use of external body-worn sensors in combination with SPs. This solution is also not accepted universally because these external devices expose the frailty of the user [33] and many users forget to put on such external devices [106]. Therefore, while designing new SP based solution, SP placement and usability issue should be handled carefully.

4.2. Open Issues

4.2.1. SP Based Context-Aware Fall Detection and Prevention

Context-aware fall detection and prevention systems use sensors deployed in the environment to detect or predict falls. The main advantage of such systems is that the user does not need to wear any special device on his or her body [59]. Due to this advantage, several context-aware fall detection and prevention solutions using various conventional external systems have been proposed [62,69,107109]. No previous report has been found in our literature search on SP-based context-aware solutions. Existing SP based solutions are body-worn type, but at home, users usually do not carry SPs on their bodies, so those SP based solutions are not suitable for home environments. Users should depend on separate conventional context-aware solutions at home. In this context, single SP based solution having both body-worn and context-aware modes of operations would be a better alternative to using separate solutions for indoor and outdoor protection. Such a SP-based solution may run in body-worn mode and context-aware mode when the user goes outside and comes back home, respectively. Automatic switching between two modes of operations is also possible.

The taxonomy of such SP-based systems is shown in Figure 2. Han et al. [110] have proposed a multimodal approach which utilizes the set of embedded sensors (accelerometer, audio tool, GPS, Wi-Fi, etc.) on smartphones in order to recognize eight different user contexts, such as walking, jogging, riding on a bus, or taking the subway. Although this system does not recognize fall events, it provides feasible support for SP-based context-aware fall detection and prevention solution. The sensors that are used frequently in traditional context-aware systems are cameras, infrared sensors, microphones and pressure sensors. Most of these sensors are also available in modern SPs. Moreover the computational and processing capacities of SPs are continuously improving. Therefore it is highly possible to use SPs for context-aware fall detection and prevention. For small monitoring area, such as a single room, context-aware system may require a single sensor. Such single sensor (e.g., camera) based context-aware system can be completely replaced with SP-only context-aware system. In that case, SP should be kept at the place (e.g., wall mounted holder) of that sensor during its context-aware mode of operation. It should be noted that we have proposed this novel concept of SP-based context-aware system based on our own observations.

4.2.2. Smartphones with Other Assistive Devices for Fall Prevention

Smartphone-based fall prevention is comparatively less explored with respect to SP-based fall detection. Among 51 reviewed articles only five articles [22,30,65,74,95] reported or evaluated fall prediction solutions and two articles [9,54] dealt with both fall detection and prediction. All previously reported solutions attempted to prevent falls by early prediction and alerting the user for imminent falls. Previous reports have only described fall prediction systems, but a working SP-based prevention system linked to assisted devices has not yet been achieved. Wu and Xue [16] proposed a pocket PC-based fall prevention system. This system can detect falls events at least 70 ms before the impact and activate an inflatable hip pad for preventing fall-related hip fractures. Since SPs can be easily substituted for Pocket PCs, this system demonstrates that SP-based fall prevention systems can be designed with the help of other assistive devices like airbags or inflatable hip pads.

4.2.3. Real-Life Falls Analysis

Falls in individuals occur relatively infrequently in real-life even in individuals with increased susceptibility to falls [111]. Therefore, only two of the SP-based solutions reviewed had evaluated their system in real-life falls [31,94]. The remaining articles only evaluated their system within simulated falls situations. Klenk et al. [112] demonstrated that simulated falls and real-life falls differ in terms of acceleration magnitude and dynamics. Consequently, the performances measured on simulated falls situations are considered inadequate for robust testing of fall detection and prevention systems [113]. Evaluation of SP-based fall detection and prevention systems in real-life conditions should therefore be considered a vital area for future research.

5. Conclusions

In this paper we have comprehensively evaluated the existing literature on SP-based solutions for fall detection and prevention. Built-in inertial sensors, open source operating systems, state-of-the-art wireless connectivity and universal social acceptance make SP a very good alternative to conventional dedicated fall detection and prevention tools. However, the performance and usability of current systems remain limited by the relatively lower quality of in-built sensors such as accelerometers in existing SP devices, as well as the need to wear the SP in a fixed position for SP-only solutions. The addition of component parts or additional systems may resolve these issues, but reduces the attractiveness of SP-based solutions. Future research should be aimed at context-aware fall detection and prevention systems which do not require the device to be worn as well as assessment of fall detection and prevention systems in real-life situations.


This research is a part of the Malaysian Elders Longitudinal Research (MELoR) project which is fully funded by University of Malaya (UM) High Impact Research (HIR) Grant from the Department of Higher Education, Ministry of Education, Malaysia (UM.C/625/1/HIR/MOHE/ASH/02) and is also supported by the eScienceFund of the Ministry of Science, Technology and Innovation, Malaysia (SF-017-2013).

Author Contributions

This survey is an outcome of the collective intellectual efforts of all the members of the team. In particular, Mohammad Ashfak Habib prepared the manuscript based on the suggestions of other authors. Mas S. Mohktar, Kheng Seang Lim and Fatimah Ibrahim designed the outline of this paper and also supplied contents for ensuring the comprehensiveness of this paper. Shahrul Bahyah Kamaruzzaman and Tan Maw Pin did the critical revision.

Conflicts of Interest

The authors declare no conflict of interest.


  1. Kalache, A.; Fu, D.; Yoshida, S. WHO Global Report on Falls Prevention in Older Age; World Health Organization Press: Geneva, Switzerland, 2008. [Google Scholar]
  2. Yavuz, G.; Kocak, M.; Ergun, G.; Alemdar, H.; Yalcin, H.; Incel, O.D.; Ersoy, C. A smartphone based fall detector with online location support. Proceedings of the International Workshop on Sensing for App Phones, Zurich, Switzerland, 2 November 2010; pp. 31–35.
  3. Tinetti, M.E.; Speechley, M.; Ginter, S.F. Risk factors for falls among elderly persons living in the community. N. Engl. J. Med. 1988, 319, 1701–1707. [Google Scholar]
  4. Elliott, S.; Painter, J.; Hudson, S. Living alone and fall risk factors in community-dwelling middle age and older adults. J. Community Health 2009, 34, 301–310. [Google Scholar]
  5. Sadigh, S.; Reimers, A.; Andersson, R.; Laflamme, L. Falls and fall-related injuries among the elderly: A survey of residential-care facilities in a Swedish municipality. J. Community Health 2004, 29, 129–140. [Google Scholar]
  6. Ryynänen, O.; Kivelä, S.; Honkanen, R.; Laippala, P. Falls and lying helpless in the elderly. Z. Gerontol. 1992, 25, 278–282. [Google Scholar]
  7. Igual, R.; Medrano, C.; Plaza, I. Challenges, issues and trends in fall detection systems. Biomed. Eng. Online 2013. [Google Scholar] [CrossRef]
  8. He, Y.; Li, Y. Physical Activity Recognition Utilizing the Built-In Kinematic Sensors of a Smartphone. Int. J. Distrib. Sens. Netw. 2013. [Google Scholar] [CrossRef]
  9. Tacconi, C.; Mellone, S.; Chiari, L. Smartphone-based applications for investigating falls and mobility. Proceedings of the 5th International Conference on Pervasive Computing Technologies for Healthcare (PervasiveHealth), Dublin, Ireland, 23–26 May 2011; pp. 258–261.
  10. Noury, N.; Quach, K.-A.; Berenguer, M.; Bouzi, M.-J.; Teyssier, H. A feasibility study of using a smartphone to monitor mobility in elderly. Proceedings of the IEEE 14th International Conference on e-Health Networking, Applications and Services (Healthcom), Beijing, China, 10–13 October 2012; pp. 423–426.
  11. Doughty, K.; Lewis, R.; McIntosh, A. The design of a practical and reliable fall detector for community and institutional telecare. J. Telemed. Telecare 2000, 6, 150–154. [Google Scholar]
  12. Lockhart, J.W.; Pulickal, T.; Weiss, G.M. Applications of mobile activity recognition. Proceedings of the 14th ACM International Conference on Ubiquitous Computing, Pittsburgh, PA, USA, 5–8 September 2012; pp. 1054–1058.
  13. Incel, O.D.; Kose, M.; Ersoy, C. A Review and Taxonomy of Activity Recognition on Mobile Phones. BioNanoScience 2013, 3, 145–171. [Google Scholar]
  14. Charlesworth, A. The ascent of smartphone. Eng. Technol. 2009, 4, 32–33. [Google Scholar]
  15. Jeon, A.-Y.; Ye, S.-Y.; Park, J.-M.; Kim, K.-N.; Kim, J.-H.; Jung, D.-K.; Jeon, G.-R.; Ro, J.-H. Emergency detection system using PDA based on self-response algorithm. Proceedings of the International Conference on Convergence Information Technology, Gyeongju, Korea, 21–23 November 2007; pp. 1207–1212.
  16. Wu, G.; Xue, S. Portable preimpact fall detector with inertial sensors. IEEE Trans. Neural Syst. Rehabil. Eng. 2008, 16, 178–183. [Google Scholar]
  17. Zhang, T.; Wang, J.; Liu, P.; Hou, J. Fall detection by embedding an accelerometer in cellphone and using KFD algorithm. Int. J. Comput. Sci. Netw. Secur. 2006, 6, 277–284. [Google Scholar]
  18. Hormann, W. Kommunikationstechnologie im Dienst am Menschen. In HAUSNOTRUFSYSTEME Kommunikationstechnologie im Dienst am Menschen; Heilscher, W., Ed.; Wirtschaftsverlag NW: Bremerhaven, Germany, 1980. (In German) [Google Scholar]
  19. Hegde, R.; Sudarshan, B.; Kumar, S.P.; Hariprasad, S.; Satyanarayana, B. Technical Advances in Fall Detection System—A Review. Int. J. Comput. Sci. Mob. Comput. 2013, 2, 152–160. [Google Scholar]
  20. Miller, G. The smartphone psychology manifesto. Perspect. Psychol. Sci. 2012, 7, 221–237. [Google Scholar]
  21. Hansen, T.R.; Eklund, J.M.; Sprinkle, J.; Bajcsy, R.; Sastry, S. Using smart sensors and a camera phone to detect and verify the fall of elderly persons. Proceedings of the European Medicine, Biology and Engineering Conference, Prague, Czech Republic, 20–25 November 2005.
  22. Majumder, A.J.A.; Rahman, F.; Zerin, I.; Ebel, W., Jr.; Ahamed, S.I. iPrevention: Towards a novel real-time smartphone-based fall prevention system. Proceedings of the 28th Annual ACM Symposium on Applied Computing, Coimbra, Portugal, 18–22 March 2013; pp. 513–518.
  23. Choi, W.-H.; Kim, S.-I.; Keum, M.-S.; Han, D.K.; Ko, H. Acoustic and visual signal based context awareness system for mobile application. IEEE Trans. Consum. Electron. 2011, 57, 738–746. [Google Scholar]
  24. Kwon, O.-Y.; Shin, S.-H.; Shin, S.-J.; Kim, W.-S. Design of U-health system with the use of smart phone and sensor network. Proceedings of the 5th International Conference on Ubiquitous Information Technologies and Applications (CUTE), Hainan, China, 16–18 December 2010; pp. 1–6.
  25. Erdem, H.; Uner, A. A multi-channel remote controller for home and office appliances. IEEE Trans. Consum. Electron. 2009, 55, 2184–2189. [Google Scholar]
  26. Bai, Y.-W.; Wu, S.-C.; Tsai, C.-L. Design and implementation of a fall monitor system by using a 3-axis accelerometer in a smart phone. IEEE Trans. Consum. Electron. 2012, 58, 1269–1275. [Google Scholar]
  27. He, Y.; Li, Y.; Yin, C. Falling-incident detection and alarm by smartphone with Multimedia Messaging Service (MMS). E-Health Telecommun. Syst. Netw. 2012, 1, 1–5. [Google Scholar]
  28. Dai, J.; Bai, X.; Yang, Z.; Shen, Z.; Xuan, D. Mobile phone-based pervasive fall detection. Pers. Ubiquitous Comput. 2010, 14, 633–643. [Google Scholar]
  29. Castillo, J.C.; Carneiro, D.; Serrano-Cuerda, J.; Novais, P.; Fernández-Caballero, A.; Neves, J. A multi-modal approach for activity classification and fall detection. Int. J. Syst. Sci. 2013, 1–15. [Google Scholar]
  30. Majumder, A.J.A.; Zerin, I.; Uddin, M.; Ahamed, S.I.; Smith, R.O. SmartPrediction: A real-time smartphone-based fall risk prediction and prevention system. Proceedings of the 2013 Research in Adaptive and Convergent Systems (RACS), Montreal, QC, Canada, 1–4 October 2013; pp. 434–439.
  31. Abbate, S.; Avvenuti, M.; Bonatesta, F.; Cola, G.; Corsini, P.; Vecchio, A. A smartphone-based fall detection system. Pervasive Mobile Comput. 2012, 8, 883–899. [Google Scholar]
  32. Chang, S.-Y.; Lai, C.-F.; Chao, H.-C.J.; Park, J.H.; Huang, Y.-M. An Environmental-Adaptive Fall Detection System on Mobile Device. J. Med. Syst. 2011, 35, 1299–1312. [Google Scholar]
  33. Mehner, S.; Klauck, R.; Koenig, H. Location-independent fall detection with smartphone. Proceedings of the 6th International Conference on PErvasive Technologies Related to Assistive Environments, Rhodes, Greece, 28–31 May 2013.
  34. Sposaro, F.; Tyson, G. iFall: An Android application for fall monitoring and response. Proceedings of the Annual International Conference of the IEEE on Engineering in Medicine and Biology Society (EMBC), Minneapolis, MN, USA, 3–6 September 2009.
  35. Busching, F.; Schildt, S.; Wolf, L. DroidCluster: Towards Smartphone Cluster Computing—The Streets are Paved with Potential Computer Clusters.
  36. Büsching, F.; Post, H.; Gietzelt, M.; Wolf, L. Fall Detection on the Road. Proceedings of the IEEE 15th International Conference on e-Health Networking, Applications and Services (IEEE Healthcom 2013), Lisbon, Portugal, 9–12 October 2013.
  37. Zhao, Z.; Chen, Y.; Liu, J. Fall Detecting and Alarming Based on Mobile Phone. Proceedings of the 7th International Conference on Ubiquitous Intelligence & Computing and 7th International Conference on Autonomic & Trusted Computing (UIC/ATC), Shaanxi, China, 26–29 October 2010.
  38. Quinlan, J.R. C4.5: Programs for Machine Learning; Morgan Kaufmann: San Mateo, CA, USA, 1993; pp. 1–80. [Google Scholar]
  39. John, G.H.; Langley, P. Estimating continuous distributions in Bayesian classifiers. Proceedings of the Eleventh Conference on Uncertainty in Artificial Intelligence, Montreal, QC, Canada, 18–20 August 1995; pp. 338–345.
  40. Chang, C.-C.; Lin, C.-J. LIBSVM: A library for support vector machines. ACM Trans. Intell. Syst. Technol. (TIST) 2011, 2, 27. [Google Scholar]
  41. Theodoridis, S.; Pikrakis, A.; Koutroumbas, K.; Cavouras, D. Introduction to Pattern Recognition: A Matlab Approach: A Matlab Approach; Academic Press: Burlington, VT, USA, 2010. [Google Scholar]
  42. Hjorth, B. EEG analysis based on time domain properties. Electroencephalogr. Clin. Neurophysiol. 1970, 29, 306–310. [Google Scholar]
  43. Terroso, M.; Freitas, R.; Gabriel, J.; Marques, A.T.; Simoes, R. Active assistance for senior healthcare: A wearable system for fall detection. Proceedings of the 8th Iberian Conference on Information Systems and Technologies (CISTI), Lisboa, Portugal, 19–22 June 2013.
  44. Lee, R.Y.; Carlisle, A.J. Detection of falls using accelerometers and mobile phone technology. Age Ageing 2011, 40, 690–696. [Google Scholar]
  45. Chiu, P.P.; Lee, T.K.; Cheng, J.M.; Yeung, S.T. Health guard system with emergency call based on smartphone. Proceedings of the IET International Communication Conference on Wireless Mobile and Computing (CCWMC), Shanghai, China, 14–16 November 2011.
  46. Tamura, T.; Yoshimura, T.; Sekine, M.; Uchida, M.; Tanaka, O. A wearable airbag to prevent fall injuries. IEEE Trans. Inf. Technol. Biomed. 2009, 13, 910–914. [Google Scholar]
  47. Tamura, T.; Yoshimura, T.; Sekine, M. A preliminary study to demonstrate the use of an air bag device to prevent fall-related injuries. Proceedings of the 29th Annual International Conference of the IEEE on Engineering in Medicine and Biology Society (EMBS), Lyon, France, 23–26 August 2007; pp. 3833–3835.
  48. Shi, G.; Chan, C.-S.; Luo, Y.; Zhang, G.; Li, W.J.; Leong, P.H.W.; Leung, K.-S. Development of a Human Airbag System for Fall Protection Using MEMS Motion Sensing Technology. Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems, Beijing, China, 9–15 October 2006; pp. 4405–4410.
  49. Hirata, Y.; Komatsuda, S.; Kosuge, K. Fall prevention control of passive intelligent walker based on human model. Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Nice, France, 22–26 September 2008; pp. 1222–1228.
  50. Hirata, Y.; Muraki, A.; Kosuge, K. Motion control of intelligent passive-type walker for fall-prevention function based on estimation of user state. Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), Orlando, FL, USA, 15–19 May 2006; pp. 3498–3503.
  51. Di, P.; Huang, J.; Sekiyama, K.; Fukuda, T. A novel fall prevention scheme for intelligent cane robot by using a motor driven universal joint. Proceedings of the International Symposium on Micro-NanoMechatronics and Human Science (MHS), Nagoya, Japan, 6–9 November 2011; pp. 391–396.
  52. Di, P.; Huang, J.; Nakagawa, S.; Sekiyama, K.; Fukuda, T. Real-time fall and overturn prevention control for human-cane robotic system. Proceedings of the 44th International Symposium on Robotics (ISR), Seoul, Korea, 24–26 October 2013; pp. 1–6.
  53. Onodera, H.; Yamaguchi, T.; Yamanouchi, H.; Nagamori, K.; Yano, M.; Hirata, Y.; Hokkirigawa, K. Analysis of the slip-related falls and fall prevention with an intelligent shoe system. Proceedings of the 3rd IEEE RAS and EMBS International Conference on Biomedical Robotics and Biomechatronics (BioRob), Tokyo, Japan, 26–29 September 2010; pp. 616–620.
  54. Mellone, S.; Tacconi, C.; Schwickert, L.; Klenk, J.; Becker, C.; Chiari, L. Smartphone-based solutions for fall detection and prevention: The FARSEEING approach. Z. Gerontol. Geriatr. 2012, 45, 722–727. [Google Scholar]
  55. Lee, J.-V.; Chuah, Y.-D.; Chieng, K.T.H. Smart Elderly Home Monitoring System with an Android Phone. Int. J. Smart Home 2013, 7, 17–32. [Google Scholar]
  56. Dai, J.; Bai, X.; Yang, Z.; Shen, Z.; Xuan, D. PerFallD: A pervasive fall detection system using mobile phones. Proceedings of the 8th IEEE International Conference on Pervasive Computing and Communications Workshops (PERCOM Workshops), Mannheim, Germany, 29 March–2 April 2010.
  57. Lopes, I.C.; Vaidya, B.; Rodrigues, J.J. Towards an autonomous fall detection and alerting system on a mobile and pervasive environment. Telecommun. Syst. 2011, 52, 2299–2310. [Google Scholar]
  58. Bai, Y.-W.; Wu, S.-C.; Yu, C.H. Recognition of direction of fall by smartphone. Proceedings of the 26th Annual IEEE Canadian Conference on Electrical and Computer Engineering (CCECE), Regina, SK, Canada, 5–8 May 2013.
  59. Rougier, C.; Meunier, J.; St-Arnaud, A.; Rousseau, J. Robust video surveillance for fall detection based on human shape deformation. IEEE Trans. Circuits Syst. Video Technol. 2011, 21, 611–622. [Google Scholar]
  60. Zhao, G.; Mei, Z.; Liang, D.; Ivanov, K.; Guo, Y.; Wang, Y.; Wang, L. Exploration and implementation of a pre-impact fall recognition method based on an inertial body sensor network. Sensors 2012, 12, 15338–15355. [Google Scholar]
  61. Bianchi, F.; Redmond, S.J.; Narayanan, M.R.; Cerutti, S.; Celler, B.G.; Lovell, N.H. Falls event detection using triaxial accelerometry and barometric pressure measurement. Proceedings of the Annual International Conference of the IEEE on Engineering in Medicine and Biology Society (EMBC), Minneapolis, MN, USA, 3–6 September 2009.
  62. Lee, Y.-S.; Chung, W.-Y. Visual sensor based abnormal event detection with moving shadow removal in home healthcare applications. Sensors 2012, 12, 573–584. [Google Scholar]
  63. Li, Y.; Zeng, Z.; Popescu, M.; Ho, K. Acoustic fall detection using a circular microphone array. Proceedings of the Annual International Conference of the IEEE on Engineering in Medicine and Biology Society (EMBC), Buenos Aires, Argentina, 31 August–4 September 2010.
  64. Zhao, Z.; Chen, Y.; Wang, S.; Chen, Z. FallAlarm: Smart Phone Based Fall Detecting and Positioning System. Procedia Comput. Sci. 2012, 10, 617–624. [Google Scholar]
  65. Otis, M.J.-D.; Menelas, B.-A.J. Toward an augmented shoe for preventing falls related to physical conditions of the soil. Proceedings of the IEEE International Conference on Systems, Man, and Cybernetics (SMC), Seoul, Korea, 14–17 October 2012; pp. 3281–3285.
  66. Brezmes, T.; Rersa, M.; Gorricho, J.; Cotrina, J. Surveillance with alert management system using conventional cell phones. Proceedings of the Fifth International Multi-Conference on Computing in the Global Information Technology (ICCGI), Valencia, Spain, 20–25 September 2010; pp. 121–125.
  67. Steinhage, A.; Lauterbach, C. Monitoring Movement Behavior by Means of a Large Area Proximity Sensor Array in the Floor. Proceedings of the 2nd Workshop on Behaviour Monitoring and Interpretation (BMI'07), Kaiserslautern, Germany, 23 September 2008.
  68. Hristova, A.; Bernardos, A.M.; Casar, J.R. Context-aware services for ambient assisted living: A case-study. Proceedings of the First International Symposium on Applied Sciences on Biomedical and Communication Technologies, ISABEL'08, Aalborg, Denmark, 25–28 October 2008.
  69. Tao, S.; Kudo, M.; Nonaka, H. Privacy-preserved behavior analysis and fall detection by an infrared ceiling sensor network. Sensors 2012, 12, 16920–16936. [Google Scholar]
  70. Lopes, I.C.; Vaidya, B.; Rodrigues, J. Sensorfall an accelerometer based mobile application. Proceedings of the 2nd International Conference on Computational Science and Its Applications, Jeju, Korea, 10–12 December 2009.
  71. Suh, M.-K.; Chen, C.-A.; Woodbridge, J.; Tu, M.K.; Kim, J.I.; Nahapetian, A.; Evangelista, L.S.; Sarrafzadeh, M. A remote patient monitoring system for congestive heart failure. J. Med. Syst. 2011, 35, 1165–1179. [Google Scholar]
  72. Viet, V.; Choi, D.-J. Fall Detection with Smart Phone Sensor. Proceedings of the 3rd International Conference on Internet (ICONI), Sepang, Malaysia, 15–19 December 2011.
  73. Kaenampornpan, M.; Anuchad, T.; Supaluck, P. Fall detection prototype for Thai elderly in mobile computing era. Proceedings of the 8th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON), Khon Kaen, Thailand, 17–19 May 2011.
  74. Fontecha, J.; Navarro, F.J.; Hervás, R.; Bravo, J. Elderly frailty detection by using accelerometer-enabled smartphones and clinical information records. Person. Ubiquitous Comput. 2013, 17, 1073–1083. [Google Scholar]
  75. Albert, M.V.; Kording, K.; Herrmann, M.; Jayaraman, A. Fall classification by machine learning using mobile phones. PLoS One 2012, 7, e36556. [Google Scholar]
  76. Hsieh, S.-L.; Su, M.H.; Liu, L.F.; Jiang, W.-W. A Finite State Machine-Based Fall Detection Mechanism on Smartphones. Proceedings of the 9th International Conference on Ubiquitous Intelligence & Computing and 9th International Conference on Autonomic & Trusted Computing (UIC/ATC), Fukuoka, Japan, 4–7 September 2012.
  77. Thammasat, E.; Chaicharn, J. A simply fall-detection algorithm using accelerometers on a smartphone. Proceedings of the Biomedical Engineering International Conference (BMEiCON), Ubon Ratchathani, Thailand, 5–7 December 2012.
  78. Cao, Y.; Yang, Y.; Liu, W. E-FallD: A fall detection system using android-based smartphone. Proceedings of the 9th International Conference on Fuzzy Systems and Knowledge Discovery (FSKD), Sichuan, China, 29–31 May 2012.
  79. He, Y.; Li, Y.; Bao, S.-D. Fall Detection by built-in tri-accelerometer of smartphone. Proceedings of the IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI), Hong Kong, China, 5–7 January 2012.
  80. Shi, Y.; Shi, Y.; Wang, X. Fall Detection on Mobile Phones Using Features from a Five-Phase Model. Proceedings of the 9th International Conference on Ubiquitous Intelligence & Computing and 9th International Conference on Autonomic & Trusted Computing (UIC/ATC), Fukuoka, Japan, 4–7 September 2012.
  81. Fahmi, P.; Viet, V.; Deok-Jai, C. Semi-supervised fall detection algorithm using fall indicators in smartphone. Proceedings of the 6th International Conference on Ubiquitous Information Management and Communication, Kuala Lumpur, Malaysia, 20–22 February 2012.
  82. Pavlakis, P.; Alepis, E.; Virvou, M. Intelligent mobile multimedia application for the support of the elderly. Proceedings of the Eighth International Conference on Intelligent Information Hiding and Multimedia Signal Processing (IIH-MSP), Piraeus, Greece, 18–20 July 2012.
  83. Fang, S.-H.; Liang, Y.-C.; Chiu, K.-M. Developing a mobile phone-based fall detection system on Android platform. Proceedings of the Computing, Communications and Applications Conference (ComComAp), Hong Kong, China, 11–13 January 2012.
  84. Viet, V.Q.; Lee, G.; Choi, D. Fall Detection Based on Movement and Smart Phone Technology. Proceedings of the IEEE RIVF International Conference on Computing and Communication Technologies, Research, Innovation, and Vision for the Future (RIVF), Ho Chi Minh City, Vietnam, 27 February–1 March 2012.
  85. Zhuang, Y.; Baldwin, J.; Antunna, L.; Yazir, Y.O.; Ganti, S. Tradeoffs in cross platform solutions for mobile assistive technology. Proceedings of the IEEE Pacific Rim Conference on Communications, Computers and Signal Processing (PACRIM), Victoria, BC, Canada, 27–29 August 2013.
  86. Kim, Y.; Kim, S.; Kang, D.; Park, H.; Kim, N.; Yang, S.H.; Kim, Y. A Simple Falling Recognition Scheme for a Human Body by Using Mobile Devices. Proceedings of the 1st International Conference on Advanced Information and Comuter Technology (AICT), Warsawa, Poland, 14–17 August 2013.
  87. Kansiz, A.O.; Guvensan, M.A.; Turkmen, H.I. Selection of Time-Domain Features for Fall Detection Based on Supervised Learning. Proceedings of the World Congress on Engineering and Computer Science, San Francisco, CA, USA, 23–25 October 2013.
  88. Cleary, J.G.; Trigg, L.E. K*: An Instance-based Learner Using an Entropic Distance Measure. Proceedings of the 12th International Conference on Machine Learning (ICML), Tahoe City, CA, USA, 9–12 July 1995; pp. 108–114.
  89. Tiwari, R.; Singh, A.K.; Khan, S.N. Using Android Platform to detect Free Fall. Proceedings of the International Conference on Information Systems and Computer Networks (ISCON), Mathura, India, 9–10 March 2013.
  90. Koshmak, G.A.; Linden, M.; Loutfi, A. Evaluation of the android-based fall detection system with physiological data monitoring. Proceedings of the 35th Annual International Conference of the IEEE on Engineering in Medicine and Biology Society (EMBC), Osaka, Japan, 3–7 July 2013.
  91. Megalingam, R.K.; Unnikrishnan, D.K.M.; Radhakrishnan, V.; Jacob, D.C. HOPE: An electronic gadget for home-bound patients and elders. Proceedings of the Annual IEEE India Conference (INDICON), Kochi, India, 7–9 December 2012.
  92. Hou, Y.; Li, N.; Huang, Z. Triaxial accelerometer-based real time fall event detection. Proceedings of the International Conference on Information Society (i-Society), London, UK, 25–28 June 2012.
  93. Boehner, A. A Smartphone Application for a Portable Fall Detection System. Proceedings of the 27th National Conference on Undergraduate Research (NCUR), La Crosse, WI, USA, 11–13 April 2013.
  94. Cabestany, J.; Moreno, J.M.; Perez, C.; Sama, A.; Catala, A.; Rodriguez-Molinero, A.; Arnal, M. FATE: One step towards an automatic aging people fall detection service. Proceedings of the 20th International Conference on Mixed Design of Integrated Circuits and Systems (MIXDES), Gdynia, Poland, 20–22 June 2013.
  95. Allen, B.; Derveloy, R.; Fell, N.; Gasior, W.; Yu, G.; Sartipi, M. Telemedicine assessment of fall risk using wireless sensors. Proceedings of the 10th Annual IEEE Communications Society Conference on Sensor, Mesh and Ad Hoc Communications and Networks (SECON), New Orleans, LA, USA, 24–27 June 2013.
  96. Yuan, B.; Herbert, J. Non-Intrusive Movement Detection in CARA Pervasive Healthcare Application. Proceedings of the International Conference on Wireless Networks (WORLDCOMP), Las Vegas, NV, USA, 18–21 July 2011.
  97. Liu, S.-H.; Cheng, W.-C. Fall Detection with the Support Vector Machine during Scripted and Continuous Unscripted Activities. Sensors 2012, 12, 12301–12316. [Google Scholar]
  98. Haines, T.P.; Hill, K.; Walsh, W.; Osborne, R. Design-related bias in hospital fall risk screening tool predictive accuracy evaluations: Systematic review and meta-analysis. J. Gerontol. Ser. A Biol. Sci. Med. Sci. 2007, 62, 664–672. [Google Scholar]
  99. A beginner's guide to accelerometers. Available online: (accessed on 14 January 2014).
  100. Falaki, H.; Mahajan, R.; Kandula, S.; Lymberopoulos, D.; Govindan, R.; Estrin, D. Diversity in smartphone usage. Proceedings of the 8th international conference on Mobile systems, applications, and services, San Francisco, CA, USA, 15–18 June 2010.
  101. Igual, R.; Plaza, I.; Martín, L.; Corbalan, M.; Medrano, C. Guidelines to Design Smartphone Applications for People with Intellectual Disability: A Practical Experience. In Ambient Intelligence-Software and Applications; Springer: Berlin, Germany, 2013; pp. 65–69. [Google Scholar]
  102. De Urturi Breton, Z.S.; Hernández, F.J.; Zorrilla, A.M.; Zapirain, B.G. Mobile communication for intellectually challenged people: A proposed set of requirements for interface design on touch screen devices. Commun. Mobile Comput. 2012, 1, 1–4. [Google Scholar]
  103. Fisk, A.D.; Rogers, W.A.; Charness, N.; Czaja, S.J.; Sharit, J. Designing for Older Adults: Principles and Creative Human Factors Approaches, 2nd ed.; CRC Press: Boca Raton, FL, USA, 2009. [Google Scholar]
  104. Brownsell, S.; Bradley, D.A.; Bragg, R.; Catlin, P.; Carlier, J. Do community alarm users want telecare? J. Telemed. Telecare 2000, 6, 199–204. [Google Scholar]
  105. Kangas, M.; Konttila, A.; Winblad, I.; Jamsa, T. Determination of simple thresholds for accelerometry-based parameters for fall detection. Proceedings of the 29th Annual International Conference of the IEEE on Engineering in Medicine and Biology Society (EMBS), Lyon, France, 22–26 August 2007.
  106. Schulze, B.; Floeck, M.; Litz, L. Concept and Design of a Video Monitoring System for Activity Recognition and Fall Detection. In Ambient Assistive Health and Wellness Management in the Heart of the City; Mokhtari, M., Khalil, I., Bauchet, J., Zhang, D., Nugent, C., Eds.; Springer: Berlin/Heidelberg, Germany, 2009; Volume 5597, pp. 182–189. [Google Scholar]
  107. Mastorakis, G.; Makris, D. Fall detection system using Kinect's infrared sensor. J. Real-Time Image Process. 2012. [Google Scholar] [CrossRef]
  108. Zhang, C.; Tian, Y.; Capezuti, E. Privacy Preserving Automatic Fall Detection for Elderly Using RGBD Cameras. In Computers Helping People with Special Needs; Miesenberger, K., Karshmer, A., Penaz, P., Zagler, W., Eds.; Springer: Berlin/Heidelberg, Germany, 2012; Volume 7382, pp. 625–633. [Google Scholar]
  109. Li, Y.; Ho, K.; Popescu, M. A microphone array system for automatic fall detection. IEEE Trans. Biomed. Eng. 2012, 59, 1291–1301. [Google Scholar]
  110. Han, M.; Lee, Y.-K.; Lee, S. Comprehensive Context Recognizer Based on Multimodal Sensors in a Smartphone. Sensors 2012, 12, 12588–12605. [Google Scholar]
  111. Noury, N.; Fleury, A.; Rumeau, P.; Bourke, A.; Laighin, G.; Rialle, V.; Lundy, J. Fall detection-principles and methods. Proceedings of the 29th Annual International Conference of the IEEE on Engineering in Medicine and Biology Society (EMBS), Lyon, France, 22–26 August 2007.
  112. Klenk, J.; Becker, C.; Lieken, F.; Nicolai, S.; Maetzler, W.; Alt, W.; Zijlstra, W.; Hausdorff, J.; van Lummel, R.; Chiari, L. Comparison of acceleration signals of simulated and real-world backward falls. Med. Eng. Phys. 2011, 33, 368–373. [Google Scholar]
  113. Bagalà, F.; Becker, C.; Cappello, A.; Chiari, L.; Aminian, K.; Hausdorff, J.M.; Zijlstra, W.; Klenk, J. Evaluation of accelerometer-based fall detection algorithms on real-world falls. PLoS One 2012, 7, e37062. [Google Scholar]
Figure 1. Common basic architecture of fall detection and fall prevention systems.
Figure 1. Common basic architecture of fall detection and fall prevention systems.
Sensors 14 07181f1 1024
Figure 2. Taxonomy of smartphone-based systems based on sensing mechanism and sensor placement.
Figure 2. Taxonomy of smartphone-based systems based on sensing mechanism and sensor placement.
Sensors 14 07181f2 1024
Figure 3. Taxonomy of smartphone based fall detection and prevention algorithms.
Figure 3. Taxonomy of smartphone based fall detection and prevention algorithms.
Sensors 14 07181f3 1024
Figure 4. Taxonomy of communication patterns in smartphone-based fall detection and prevention systems.
Figure 4. Taxonomy of communication patterns in smartphone-based fall detection and prevention systems.
Sensors 14 07181f4 1024
Figure 5. Estimation of the number of SP based fall detection and prevention studies.
Figure 5. Estimation of the number of SP based fall detection and prevention studies.
Sensors 14 07181f5 1024
Table 1. Smartphone built-in sensors and their uses.
Table 1. Smartphone built-in sensors and their uses.
Built-in Sensors of SPUsual Use in SPUse in Fall Detection & Prevention
AccelerometerSenses the changes in orientation of SP and adjusts the viewing angle accordingly.[60]
GyroscopeDetects angular momentum (roll, pitch and yaw); facilitates game.[60]
MagnetometerSenses the Earth's magnetic field; works as a digital compass.[60]
BarometerMeasures atmospheric pressure; facilitates weather widgets.[61]
Image SensorProvides still picture and video capturing facilities.[62]
MicrophoneSound capture.[63]
Wi-Fi sensorFacilitates wireless communication through Wi-Fi.[64]
Bluetooth SensorFacilitates wireless communication through Bluetooth.[60]
Location sensors (GPS)Targets or navigates by map or picture with the help of GPS satellites.[2]
Temperature SensorMeasures temperature; facilitates weather widgets.[65]
Humidity SensorMeasures humidity; facilitates weather widgets.[65]
Ambient Light SensorAdjusts the display brightness.[66]
Proximity SensorDetects how close our SP's screen is to our body.[67]
Touch SensorHelps to operate the SP through touching.-
NFC SensorEstablishes communication between similar device by touching or bringing them into proximity.[68]
Infrared SensorCan sense temperature.[69]
Back-Illuminated sensorAdjust the light captured while capturing a photograph.-
Table 2. Comparison of smartphone-only fall detection and prevention systems.
Table 2. Comparison of smartphone-only fall detection and prevention systems.
YearArticleObjectiveSP PositionSensor(s)Algorithm(s)Alerting Feature(s)
2009[34]DetectionAnyAccelerometerTBA (Adaptive: depends on user provided parameters)SMS (time, GPS coordinates, password for activating bidirectional voice call).
[70]DetectionTrouser PocketAccelerometerTBA (Fixed)SMS, voice call, vibration, sound.
2010[28]DetectionChest, Waist, ThighAccelerometer & gyroscopeTBA (Fixed)Sound alarm, voice call.
[2]DetectionTrouser PocketAccelerometerDiscrete Wavelet Transform (DWT)SMS (GPS coordinates), email (Google Map), twitter messages.
[56]DetectionChest, Waist, ThighAccelerometerTBA (Fixed)Audible alarm, voice call.
[37]DetectionWaistAccelerometerC4.5 DT, NB and SVMSMS
2011[9]DetectionWaistAccelerometerTBA (Fixed)E-mail and/or SMS.
[44]DetectionWaistAccelerometerTBA (Fixed)SMS (date, time, location)
[71]DetectionPocketAccelerometerTBA (Fixed)SMS (name, time, GPS coordinates, street address)
[72]DetectionHand, Shirt or Trouser PocketAccelerometer & gyroscopeTBA (Fixed), One-Class SVMNot found
[45]DetectionNot foundAccelerometerTBA (Fixed)Audible alarm, SMS (GPS coordinates), voice call (manual), remote server draws help path
[73]DetectionShirt PocketAccelerometerTBA (Fixed)SMS
2012[26]DetectionWaistAccelerometerTBA (Fixed)SMS (time, GPS data), draw help path
[27]DetectionWaistAccelerometerTBA (Fixed), Median filter attenuate noiseMMS (time, map of suspected fall location, and GPS coordinate)
[31]DetectionWaistAccelerometerTBA (Fixed), ANN 1 based pattern classifierNotification contains GPS coordinates.
[54]uFall for Detection, uTUG for PreventionWaistAccelerometer, GyroscopeTBA (Fixed)E-mail or SMS, recorded signals are sent to remote server, audio cue (for uTUG)
2012[74]Prevention (GUG)WaistAccelerometerSegmentation, filtering, dispersion measures calculationNot found
[75]DetectionWaist (Back)AccelerometerSVM, SMLR 2 in SP, NB, DT, KNN 3 in PCNot found
[76]DetectionShirt or Trouser PocketAccelerometerTBA (Considers axis wise data separately)Not found
[77]DetectionShirt PocketAccelerometerTBA (Adaptive)Not found
[78]DetectionShirt PocketAccelerometerTBA (Adaptive)Text message
[79]DetectionWaistAccelerometerTBA (Fixed), Median Filter,MMS (time, GPS coordinate, Google map)
[80]DetectionTrouser PocketAccelerometerSVM classifierVibration, sound alarm, SMS (time, location, & health information)
[64]DetectionWaistAccelerometer, Wi-Fi moduleDT Classifier, location estimation using RSSI 4SMS (name, time, location)
[81]DetectionHand, Pocket, waistAccelerometer, GyroscopeSemi-supervised learningNot found
[82]DetectionNot foundAccelerometer, GyroscopeNot foundSMS (location),
[83]DetectionChest, Waist, ThighAccelerometerTBA (Adjusted based on user's profile)SMS
[84]DetectionHand, PocketAccelerometer, GyroscopeTBA (Fixed)Not found
2013[57]DetectionTrouser PocketAccelerometerTBA (Fixed)SMS (date, time, GPS data), voice call, vibration, sound.
[8]DetectionChestAccelerometer, Gyroscope, & MagnetometerFisher's discriminant ratio and J3 criterionMMS (time, map of suspected fall location, GPS coordinate)
[22]PreventionTrouser PocketAccelerometer & GyroscopeC4.5 DT classifier, Hjorth mobility and complexity [42]Alert the user about imminent fall by using message & vibration.
[33]DetectionWaistAccelerometerTBA (Fixed)SMS, voice call, others: twitter, email, Facebook.
[55]DetectionNot foundAccelerometerTBA (Fixed)SP trigger PC via Wi-Fi, PC send alert via SMS, emails or/and voice calls
[58]DetectionWaistAccelerometerTBA (Fixed)SMS (time, GPS data), draw help path
[85]DetectionNot foundAccelerometerTBA (Fixed)Not found
[86]Detection(User's height 164 cm)AccelerometerTBA (Fixed)Server displays current states and triggers an alarm
[87]DetectionTrouser PocketAccelerometerOneRAttributeEval, ReliefFAttributeEval SVMAttributeEval, K* [88], C4.5, NBSMS (GPS coordinate)
[89]Detection (Free Fall)Not foundAccelerometerDisplacement based algorithmsSMS (GPS coordinate)
[90]DetectionWaistAccelerometerTBA (Fixed)SMS

1.Artificial Neural Network;2.Sparse Multinomial Logistic Regression (SMLR);3.k-Nearest Neighbours (KNN);4.Received Signal Strength Indication.

Table 3. External components, used in SP-based fall detection and prevention solutions.
Table 3. External components, used in SP-based fall detection and prevention solutions.
Component NameFeaturesUsed In
SensorTag (TI)Temperature, Humidity, & Pressure Sensor, Accelerometer, Gyroscope, Magnetometer, Bluetooth, 8051 Microcontroller[43]
Shimmer2 (Shimmer)Accelerometer, 802.15.4 standard Radio, Bluetooth Module, MSP430 Microcontroller[31]
GPSADXL2-axis Accelerometer (Two), GPS Module[21]
BlueGiga WRAPBluetooth RS-232 cable replacer[21]
CameraVideo Camera[29]
X6-2 Mini (Gulf Coast)Accelerometer[75]
BC5 (CSR Inc.)Bluetooth Module[92]
EZ430 Chronos (TI)Accelerometer, Pressure, Temperature & Battery Voltage Sensor, Bluetooth Module, MSP430 Microcontroller[93]
CC1111 (TI)USB RF Access Point[93]
LIS344ALH (STMicro)Accelerometer[94]
BlueGiga WT12Bluetooth Module[94]
XBee RF (Digi)ZigBee Module[94]
XU-Z11 (Digi)USB to ZigBee Adaptor[94]
XR-Z14-CW1P2 (Digi)ZigBee Wall Router[94]
Bed Presence (Ibernex)Detects the absence of user on bed[94]
PIC24F (Microchip)Microcontroller[65,94]
Piezoresistive sensorsCan measure mechanical stress[30]
WiFly ShieldAble to connect to 802.11b/g wireless networks[30]
NODE (Variable Tech)Accelerometer, Gyroscope, Magnetometer, Bluetooth Module[95]
Table 4. Fall detection and prevention systems using smartphone and other external units.
Table 4. Fall detection and prevention systems using smartphone and other external units.
YearArticleObjective *Sensor(s)SP PositionExternal Sensor's PositionSP—External Unit ConnectivityAnalysis UnitAlgorithm(s)
2005[21]DSP camera, External accelerometerAnyWaistBluetoothExternal PCNot found
2010[28]DSP accelerometer, gyroscope & magnetometer, Several external magnets (35 mT)Trouser right (left) PocketJust above left (right) kneeMagnetic FieldSPTBA (Fixed), Hausdorff distance
2011[32]DExternal accelerometer & gyroscopeAnyWaist, left & right ankleZigBeeSPCenter of gravity clustering algorithm
[96]DSP accelerometer & gyroscopeNot foundChest, Finger tipBluetoothExternal PCTBA (Fixed)
2012[31]DExternal accelerometerAnyWaistBluetoothSPANN Based Pattern Classifier
[91]DExternal accelerometerAnyChestBluetoothExternal Arduino BoardTBA (Fixed)
[92]DExternal accelerometerNot foundChest/WaistBluetoothSPTBA & Binary DT
[65]PExternal bend, temperature & humidity sensor, accelerometer, gyroscopeNot foundShoe-SoleBluetoothSPSVM, Fast ANN & TBA
2013[29]DSP accelerometer & GPS receiver, External video cameraChestWall mountedClient/Server networkSP & Network PCBoth TBA & machine learning
[43]DSP GPS Module, External accelerometerAnyTorsoBluetoothExternal UnitNot found
[93]DExternal accelerometerAnyWristBluetoothExternal PCTBA (Fixed)
[94]DExternal accelerometer, gyroscope, bed presence sensorAnyWaistBluetoothExternal UnitNot found
[30]PSP accelerometer & gyroscope, External pressure sensor (4 units),Pocket or HandShoe-SoleWi-FiSPHjorth mobility and complexity, Energy Integral
[95]PExternal accelerometer & gyroscope (two sets)Not foundChest and ArmBluetoothSPTBA (Fixed)

*“D” represents Detection and “P” represents Prevention.

Table 5. Declared performances of the SP based fall detection and prevention solutions.
Table 5. Declared performances of the SP based fall detection and prevention solutions.
ArticleObjectiveDeclared Performance
[8]DetectionThe total classification accuracy is 95.03% (accuracies for static, transitions, dynamic, and falls are 98.75%, 94.625%, 91.8%, and 97.63%, respectively)
[9]DetectionBoth specificity and sensitivity are 100%, except the case when fall dynamics is completely in the vertical direction
[22]Prevention99.8% accuracy in gait abnormality detection
[28]DetectionAverage of false negative values is 2.13% and the false positive value is 7.7%
[30]Prevention97.2% accuracy in gait abnormality detection
[31]DetectionObtained 100% sensitivity, specificity, and accuracy
[33]DetectionSensitivity 83.33% and a specificity 100%
[44]DetectionSpecificity and sensitivity are 81% and 77% respectively
[56]DetectionWaist is the best position to attach the phone and gives average false negative value of 2.67% and false positive value of 8.7%.
[58]DetectionAccuracy 94% (50 samples for the test and 47 of these samples are correct)
[64]DetectionPrecision & Recall (respectively) for DT: 100% & 75.8%; for SVM: 99.81% & 75.43%; for NB: 98.67% & 73.20%
[37]DetectionAccuracy for DT is 98.85%, for SVM is 86.47%, and for NB is 87.78%
[72]DetectionAccuracies are 75% (while typing SMS), 87.5% (while listening), 77.9412% (SP in chest pocket) and 84.2857% (SP in pants pocket)
[75]DetectionIdentify falls with 98% accuracy and classify the type of falls with 99% accuracy
[76]DetectionAverage sensitivity & specificity are 97% & 100% respectively
[78]DetectionSensitivity 92.75% and specificity 86.75% (for adaptive TBA)
[80]DetectionAverage recall is 90% and precision is 95.7%
[81]DetectionSensitivity 85.3% and specificity 90.5%
[83]Detection72.22% sensitivity and 73.78% specificity
[84]DetectionSensitivity 80%, specificity 96.25% and accuracy is 85%
[86]DetectionAccuracy is 86% in lying and 100% in falling
[87]DetectionPrecision & Recall (respectively) for NB: 83.8% & 82.0%; for J48 DT: 88.2% & 88.3% for K-Star: 88.9% & 88.6%
[90]Detection90% specificity, 100% sensitivity and 94% accuracy
[92]DetectionOverall accuracy of 92%
[96]DetectionFalls (active) accuracy 95.2%, Falls (inactive) accuracy 95.7%
Sensors EISSN 1424-8220 Published by MDPI AG, Basel, Switzerland RSS E-Mail Table of Contents Alert
Back to Top