ITAN20130056A1 - SYSTEM FOR DETECTION AND PREVENTION OF OLD FALLS, THROUGH A FLOOR CASSETTE - Google Patents

Description

â € œSystem for the detection and prevention of elderly falls, by means of a floor resonance box ",

The present invention refers to a system capable of autonomously detecting and preventing the fall of a person and of sending an alarm signal to designated persons / entities or other systems.

In particular, we want to give a solution to a problem by now known, that is the safety of having help in the event that an elderly person alone in the house falls to the ground and consequently is unable to call someone.

To understand the importance of this problem, it is sufficient to note that at the national level over one in four elderly (28.3%) of the population, aged 65 and over, lives alone (equal to 3,481,334) and according to statistics 35-40% of subjects over the age of 65 fall at least once a year (2011 Health Observatory Report, National Observatory on Health in the Italian Regions). Furthermore, timely assistance in the event of a broken femur can significantly reduce the onset of complications and increase the effectiveness of the rehabilitation phase.

It is not only the health conditions of the elderly that determine the situations that facilitate falls, but also the structure of the housing unit, its level of infrastructure and furniture. According to the VAOR-ADI database developed by the Center for Aging Medicine of the University of the Sacred Heart, the presence of an unsafe living environment is able to increase the risk of falls at home by about 50%, and in this, the degree of illumination of the environments while traveling is particularly important.

It is therefore extremely important to try to act both in the preventive field and in the field of autonomous detection and timely rescue.

State of the art

In order to make it easier to understand, the state of the art has been divided into fall prevention systems and detection systems.

Fall prevention systems

Regarding the prevention of falls for the elderly, a list of behavioral recommendations has been known for some time, such as the end of every physical activity to train strength and balance, not getting out of bed quickly and many others. Based on these simple concepts, systems have been developed over the past two decades to facilitate or automate the implementation of these recommendations.

Among the most important and object of this invention, that of moving with adequate lighting as the darkness prevents you from seeing the obstacles. With regard to the possibility of using automatic systems for switching on lights in the event of a detected presence and low natural light, there are various solutions (for example, the one proposed by Meierl Thomas, WO9727077A1).

However, being designed for different uses such as energy saving or auto motive, they do not solve the problem in a safe and comfortable way.

In fact, in relation to the prevention of falls, with such traditional systems and sensors, it is difficult to be able to cover the paths tightly and to intervene exactly when the person begins to move from an environment, where until a few seconds before he wanted to stay in the dark to rest and maybe make some small movements while sitting in front of the television or lying on the bed.

Fall detection systems

The problem has already been studied with different approaches.

In particular, there are solutions where parameters related to the person are elaborated which include wearable devices and others where environmental parameters are elaborated with fixed devices.

Solutions based on wearable devices

As for wearable devices (as in the solution proposed by Andrew Wolfe and Thomas Martin Conte, US2010 / 0286567A1, or in the solution proposed by Flinsenberg, W02010 / 044013A1), among the most popular solutions we find sensors such as accelerometers and tilt sensors , through which it is possible to determine the position, speed and acceleration of the person. However, these can be easily forgotten and be uncomfortable in certain activities such as showering.

Solutions based on video analysis

Solutions related to the environment are typically linked to the use of cameras (as in the solution proposed by Anders Fredriksson, US2009 / 0121881A1, or in the solution proposed by C.W. Lin, Z.H. ling, â € œAutomatic fall incidete detection in compressed video for intelligent homecare, â € in Proc. ICCCN 2007, Hawaii, Aug. 2007, pp. 1172-1177) with computerized processing for real-time determination of people's movements. However, problems related to privacy and the complexity of the systems to make the areas in which to protect people totally visible, make this solution difficult to apply.

Other solutions include the monitoring of floor vibrations through piezoelectric sensors (as in the solution proposed by M. Alwan, P. Rajendran, S. Kell et al, "A smart and passive floor vibration based fall detector for elderly⠀ in Proceedings ICTTA '06 , Damascus, Syria, Apr. 2006, pp. 23-28) or with floor accelerometers (as in the solution proposed by F. Werner, J. Diermaier, S. Schmid, P. Panek, "Fall Detection with Distributed Floor-mounted Accelerometers, "Innovatìon, 2011) and the comparison with certain predetermined" fall patterns ". While on the one hand the use of fall patterns eliminates the uncertainty in detecting a certain condition, this however tends to become an obstacle in the definition of the same. Furthermore, the use of large databases and the real-time comparison of what is happening with what is known (essential to have a good reliability with traditional sensors), requires high computational resources. complexity and complexity making such solutions inaccessible.

Solutions sufficed on the analysis of environmental noise

The use of the sound produced by the fall has also been widely discussed, in particular, however, reference is always made to the airborne noise contained within the environment being measured, with varied analyzes from case to case. For example with the use of particular geometric arrangements of microphones (as in the solution proposed by Y. Li, Z. Zeng, M Popescu, S. Member, K.C. Fellow, "Acoustic Fall Detection IJsing a Cìrcular Microphone Array," Health (San Francisco), 2010, pp. 2242-2245) followed by different types of signal analysis, often with recognition of â € œpatternsâ € as in the case of vibrations or accelerations. All this undoubtedly leads to problems related to environmental noises, just think of a simple television on to understand how this can alter the noise produced by the fall. These solutions also present the same criticalities described above, regarding the use of patterns and the consequent high computational resources.

Solutions with pUt sensorì fusion

In order to reduce false alarms, solutions have been proposed in which multiple sensory sources are used such as accelerometers and aerial microphones in wearable devices (as in the solution proposed by Noury N, â € œA smart sensor for the remote follow up of activityâ € , Medicine, 2002, or in the solution proposed by Doukas C, â € œ Advanced Patient or Elder Fall Detection based on Movement and Sound Dataâ €, Systems Engineering, 2008) or with floor accelerometer and aerial microphone (as in the solution proposed by Israel Gannot and Yaniv Zigel, WO 2009/113056 A1). However, this approach is once again difficult in its application as foreseeing the use of patterns and the concomitance of the two factors (sound and acceleration), as well as being even more onerous from the computational point of victory, it tends to suffer from side effects air sound as described above.

Summary of the invention

A first object of the present invention is to provide a system which does not have the drawbacks of the aforementioned systems and which is particularly simple, effective and reliable. The second purpose of the invention is to provide a system that has a low computational burden and consequently that is easy to implement. Of particular importance is also the fact of preventing the fall, always guaranteeing adequate lighting of the environments during a movement. By detecting the movements through the sound waves transmitted by the perishing, it is possible to determine with certainty the movement of the person, so as not to interfere in the desired lighting comfort, when there are no movements but simple movements. Finally, obviously, the possibility of detecting the presence of people in the environments in order to send notifications of various kinds, such as vitality notifications.

These and other purposes are achieved in accordance with the present invention by means of a system comprising a resonance chamber with a microphone inside it (supported by a membrane on the floor and isolated from environmental airborne noise by means of an insulator) and a simple microcontroller to process the signal and send signals to the alarm sending unit Using a resonance box to capture the noise coming from the floor and isolating it from the airborne noise of the environment, it is possible to simply discriminate in frequency the impact of a heavy object such as a person with respect to to a tool or to what an elderly person can handle, obtaining a better reliability and efficiency than the known art.

Brief description of the figures

Below is a brief description of the figures shown:

Fig, 1 conceptual scheme of the survey unit;

Fig.2 vertical section of the resonance chamber with microphone in 1: 1 scale;

Fig. 3 axonometric view of the resonance chamber with microphone;

Fig. 4 conceptual scheme of the processing algorithm.

Detailed description of the invention.

The system is composed of one or more detection units (typically one for each room to be monitored) and an alum sending unit which, received the signal from the detection unit (for example via wireless), will send the alert to entities / persons in charge ó to other systems.

In fig. 1 which schematizes in a conceptual way a simplified view of the detection unit, we can see a sound box with microphone (1), which has the purpose of picking up the sound waves present inside the floor and making them resonate to its own tune. The signal obtained from the sound box with microphone (1) is preamplified by means of a preamplifier (8) and converted into digital through an analog / digital converter (9). Once converted, the signal is processed through a processing unit (10), according to the data set by the user through a user interface (12). In the case of slow movement, the processing unit (10) acquires the ambient lighting level through the brightness sensor (13), and switches on the lighting source such as the lamp (14), if this does not proved adequate.

If a fall is detected, the processing unit (10) sends the fall signal through the communication module (11) to the alarm sending unit.

The resonance chamber with rmcrofbno (1), as schematized in fig. 2 and fig. 3, is composed of a microphone capsule (5), for example electret, contained in a resonance chamber (2) consisting of a container ( 3) and a membrane (4), resting on the floor.

An example of realization of the resonance box (2) is represented in fig. 2, through a container (3) and a membrane (4), made of plastic or metal or wood, with

membrane fA \

However, shape, size and materials can be varied according to the need for efficiency and dimensions, consequently the materials and shapes described are only a conceptual example. The resonance chamber (2) must be examined isolated from the airborne noises of the environment, by means of a layer of acoustic insulation (6) and an additional container (7) in order to prevent these from altering the sound waves picked up inside the floor.

The user interface (12), named in fig.l and fig. 4, consisting for example of a display and selectors, is housed in the detector unit allowing, as shown in fig. 4, to select the type of floor (19) and the threshold values (20), such as the falling threshold and the displacement threshold, in order to adapt the system to any need in an easy and immediate way.

The processing unit (10), named in fig.1, can simply consist of a microcontroller which will have to process the signal in real time according to the processing algorithm shown in fig.4. This algorithm first of all provides for a spectral processing (15) such as a fast Fourief transform, in order to obtain the signal amplitude on n limited frequency bands (for example 20 + 40, 40 + 60, 60 + 80, ..., 480Ã 500Hz).

Then the value reached (16) is calculated using the formula:

bi X Cf

where bi is the amplitude of the i-th brada obtained from the spectral processing (15) and ctà is the i-th weight coefficient, associated with the said brada bt, through a predetermined configuration depending on the type of floor (19 ) set via the user interface (12).

Depending on the type of floor (19), therefore, different pre-established configurations of the weight coefficients are set and vary from 0 to 1 (for example in the case of parquet = 1, c2 = 1. ... .c4 = 0.6 ,. ..., Ce = 0,4, ...., CB = 0, ... cn = 0), thus defining a kind of spectral imprint proper for that type of floor. These easily distinguishable footprints are however attributable to five large families of floors (ceramic, stone, linoleum, parquet, wood), n circumvented value (16) avenue then compared with the threshold values (20) in order to detect the fall or a simple shift and act accordingly as previously described. Furthermore, the type of floor (19) and the threshold (20), fall and displacement values can also be set through an automatic setting mode (23). This is done through the fall of a specific object at a fixed distance (for example at a distance of 2 meters from the detection unit), and the comparison of the amplitude values obtained on the various frequency bands (&!, ... ., bn) with predetermined values (in order to bring it back into a type of floor and set the thresholds in proportion, on the basis of the value achieved calculated for that known impact of weight and distance).

Claims (8)

CLAIMS: 1. System for the detection and prevention of falls of elderly people consisting of an alarm sending unit and one or more detection units consisting of: - a resonance box (2), in order to make the sound waves coming from the floor resonate; - a microphone (5), positioned inside the said resonance chamber (2), in order to pick up the sound waves present inside the floor; - a brightness sensor (13), in order to check the adequate degree of illumination during a detected movement; - an illumination source, typically a lamp (14), to generate the right degree of illumination during a detected movement; - a module for communication (11) with the other units; - a user interface (12) to select the various operating modes, the type of floor, the drop and shift threshold values; - a processing unit (IO) to process the signal, recognize the Fall, send signals to the alarm sending unit, recognize movements, detect brightness, switch on the light source. 2. System for the detection and prevention of falls of the elderly according to claim 1, characterized by the fact that the sound box (2) must be formed by a container (3) and a membrane (4) and contain inside a microphone (3). 3. System for the detection and prevention of falls of the elderly according to claims 1 and 2, characterized in that the membrane (4) of the sound box (2) must be in contact with the floor, in order to make the sound waves coming from the floor inside the box, of resonance (2). 4. System for the detection and prevention of falls of elderly people according to one or more preceding claims, characterized by the membrane (4) and the container (3) of the sound box (2), are made of plastic or metal material or woody. 5. System for the detection and prevention of falls of elderly people according to one or more preceding claims, characterized by the false membrane (4) having a surface between 0.5 and 100 cm <2>. 6. Sistrana for the detection and prevention of falls of elderly people according to one or more preceding claims, characterized by the fact that the resonance box (2) must be isolated from airborne noises, to prevent them from altering the sound waves picked up at the interior of the floor. 7. Sistrana for the detection and prevention of falls of elderly people according to one or more previous claims, characterized by the fact that the processing unit (10) is configured for: â– carry out a spectral processing (15), in order to obtain the amplitude of the signal on n limited frequency bands; calculate the value reached (16), using the formula n / b {x Cj where is the width of the i-th band obtained from the spectral processing (15) and et is the i-th weight coefficient, fixed to the said bit band through a predetermined configuration depending on the type of floor (19) set it via the user interface (12); compare (17) said value reached (16), with the threshold values (20) fall and displacement, set them through the user interface (12), in order to send the alarm signal (18) in case of detected fall o switch on the light (22) in the event of a movement detected in an environment that is not adequately lit. 8. System for the detection and prevention of falls of persons according to one or more preceding claims, characterized in that the parameters of the type of floor (19) and the threshold values (20) for falling and moving, can also be set through a automatic mode (23) for setting, through the fall of a specific object at a fixed distance, and the comparison of the amplitude values obtained on the various frequency bands, with predetermined values.