ELEVATOR SYSTEM
FIELD OF THE INVENTION
The present invention relates to elevator systems.
More particularly the present invention relates to a
method, a device, a computer program and a system for
detecting passengers stepping into an elevator car and
exiting from it.
BACKGROUND OF THE INVENTION
It is essential from the standpoint of the efficient
use of elevator systems in buildings to know the
passenger flows inside the building and how many
passengers are in the elevator cars in different
operating situations. More particularly information
about passengers arriving in the elevator cars and
exiting from them on each landing gives detailed
information about the passenger flows of the buildings,
from which it is possible to compile, among other
things, statistics for evaluating and for improving
the efficiency of the use of the elevator systems. By
means of statistics it is also possible to estimate
the service need of elevator systems and to prepare
forecasts of the numbers of passengers to be served.
Up-to-date information about the numbers of passengers
in elevator cars can, for its part, be utilized in
different operating situations, such as e.g. in
interruptions to the operation of the elevator. As a
result of the advantages to be achieved, the need for
measuring passenger flows often becomes a issue to
address when modernizing old elevator systems and/or
when installing a condition monitoring system in an
elevator system, in connection with which it is
desired to integrate the monitoring of passenger
traffic.

The movement of elevator passengers into the elevator
car and out of the elevator car is in prior art
determined by using door photoelectric cells for
detecting the movement of people or by measuring the
load of the elevator car by means of a so-called car
load weighing device e.g. during a stop of the
elevator. The separating capability of a photoelectric
cell is however limited in peak-traffic situations,
especially if there is simultaneous traffic in both
directions at the doors. When using load information,
the load of the elevator at the time of stopping, at
the time of starting, and the smallest load during the
time between these, has been measured. From these
results the number of incoming and outgoing passengers
has been calculated utilizing the average weight of a
passenger. In the method it is assumed that all the
exiting passengers leave the car before the incoming
passengers step into the car, which does not correspond
to the real situation. The divergences of the weight of
actual people and the weight of a normalized elevator
passenger also cause an inaccuracy.
One prior-art solution is disclosed in patent
application EP0528188, in which the arrival in the car
and departure of passengers is detected from changes
occurring in the signal of the car load weighing
device. The method improves the method presented above
that is based on the signal of the load weighing device
but is however imprecise owing to the inaccuracy of the
signal of the load weighing device and the limited
frequency response of the car load weighing device.
More particularly the solution is difficult to
implement when modernizing elevators because connecting
to the load-weighing signal can be awkward or the
elevator car totally lacks a car load weighing device.

An acceleration sensor can be used in an elevator
system for many kinds of measurements. For example,
the acceleration of the elevator car can be monitored
with an acceleration sensor. From the measurements
given by from the sensor it is possible to calculate,
in addition to acceleration, e.g. the position of the
elevator in the elevator shaft and the stopping
accuracy of the elevator floor by floor. Overall a
very comprehensive view of the operation of the whole
elevator can be formed from the measurement results of
the acceleration sensor. One possible embodiment of an
acceleration sensor in connection with elevator
systems is to detect the arrival/departure of
passengers into the elevator car/out of the elevator
car by means of an acceleration sensor fixed to the
elevator car.
Acceleration measurement in itself is not usually
adequate as a basis for signal processing, but instead
generally it is necessary to integrate in order to
ascertain more accurate results. In this case so-
called bias problems (deviations) caused by
installation errors of the sensor are inevitably
encountered. Bias problems are caused by, among other
things, the acceleration sensor never being in
practice fully perpendicular with respect to the
direction of movement to be measured. In addition, if
the acceleration sensor is installed on the roof of
the elevator, it inclines dynamically with the car as
the loading of the car changes. Fig. 1 illustrates one
such situation.
In Fig. 1 the elevator is standing at a floor with
passengers exiting and arriving in the car. The upper
curve in Fig. 1 presents a situation in which the
acceleration as such is integrated into speed v(t) and
speed, for its part, into position x(t).


where vo=0, xo=0, To= the moment when the door is open
and passengers are able to move and T is the moment in
time when the closing stage of the door starts, At is
the discrete interval (sampling interval) and N is the
number of samples.
As is seen from Fig. 1, the error caused by the
inclination of the car accumulates in the integration,
in which case speed and position "escape"
uncontrollably. It is possible to attempt to improve
the situation with the fact that the speed of the car
at the start and at the end of the loading cycle is
zero:

The above integrates at first the measured
acceleration for calculating the speed v(t), the
average error ab for acceleration is calculated from
the final error of speed (the end condition of
integration is that the speed of the car must be
zero) . With this term the speed v(t) and finally the

position x(t) are integrated again from the corrected
acceleration. From the lower curve of Fig. 1 it is
seen that the situation improves slightly, but not
however sufficiently. The deviations in the position
of the car produced by the passengers are generally of
the order of magnitude of hundreds of micrometers and
at their maximum of some millimeters. In the lower
curve of Fig. 1 the calculated deviation of the car is
approx. 50mm. On the basis of Fig. 1 it is seen that
the deviations in the position of the car produced by
the passengers are lost in the errors caused by
inclination of the car and reliable detection of the
passengers from a signal thus corrected is very
problematic.
SUMMARY OF THE INVENTION
The purpose of the present invention is to disclose a
new method, device, computer program and system for
detecting passengers stepping into an elevator car and
exiting from it. The term "detect" means in this
context that in the solution according to the
invention the arrival in the elevator car/departure
from the elevator car of an elevator passenger is
detected (observed).
The method, the computer program, the device and the
system according to the invention are characterized by
what is disclosed in the characterization part of
claims 1, 7, 9 and 14. Other embodiments of the
invention are characterized by what is disclosed in
the other claims. Some inventive embodiments are also
presented in the drawings in the descriptive section
of the present application. The inventive content of
the application can also be defined differently than
in the claims presented below. The inventive content
may also consist of several separate inventions,

especially if the invention is considered in the light
of expressions or implicit sub-tasks or from the point
of view of advantages or categories of advantages
achieved. In this case, some of the attributes
contained in the claims below may be superfluous from
the point of view of separate inventive concepts. The
features of the various embodiments can be applied
within the scope of the basic inventive concept in
conjunction with other embodiments.
In accordance with the first aspect of the invention,
a method for detecting elevator passengers is
presented. In the method the vertical acceleration
values of the elevator car are received from the
acceleration sensor and the passengers arriving in the
elevator car and/or leaving the elevator car are
detected on the basis of the vertical acceleration
measurements of the acceleration sensor.
In one embodiment of the invention the calculated
speed of the elevator car is calculated from the
vertical acceleration measurements of the acceleration
sensor, the calculated speed is preprocessed by
rendering the speed of the elevator car as zero
elsewhere except in a situation of loading passengers
or in a situation of unloading passengers, and the
passengers arriving in the elevator car and/or leaving
the elevator car are detected from the preprocessed
speed on the basis of the calculated position of the
elevator car. The term "rendering" means in this
context that the speed is set at zero elsewhere except
in a loading situation or an unloading situation,
after which elimination of the offset (bias) is
performed for each loading situation and unloading
situation. In one embodiment a movement indicator is
used in the preprocessing, which detects a situation
of loading passengers or a situation of unloading

passengers from the fluctuation of movement of the
car.
In one embodiment of the invention the passengers
arriving in the elevator car and/or leaving the
elevator car are detected from the preprocessed speed
on the basis of the calculated position of the
elevator car with the correlation method.
In one embodiment of the invention the passenger last
arriving in the elevator car or leaving from it is
detected for measuring the photoelectric cell delay of
the elevator.
In accordance with the second aspect of the invention,
a computer program is presented. The computer program
is arranged to perform the phases of the method
presented in the method claims 1-5. In one
embodiment of the invention the computer program is
stored on a data processing appliance on a readable
storage medium.
In accordance with the third aspect of the invention,
a device for detecting elevator passengers is
presented. The device is arranged to receive the
vertical acceleration values of the elevator car from
the acceleration sensor and to detect the passengers
arriving in the elevator car and/or leaving the
elevator car on the basis of the vertical acceleration
measurements of the acceleration sensor.
In one embodiment of the invention the device is
arranged to calculate the calculated speed of the
elevator car from the vertical acceleration
measurements of the acceleration sensor, to preprocess
the calculated speed by rendering the speed of the
elevator car as zero elsewhere except in a situation

of loading passengers or a situation of Unloading
passengers, and to detect the passengers arriving in
the elevator car and/or leaving the elevator car from
the preprocessed speed on the basis of the calculated
position of the elevator car. In one embodiment the
device is arranged to use a movement indicator in the
preprocessing, which detects a passenger loading
situation or a passenger unloading situation from the
fluctuation of movement of the car. In one embodiment
of the invention the passengers arriving in the
elevator car and/or leaving the elevator car are
detected from the preprocessed speed on the basis of
the calculated position of the elevator car with the
correlation method.
In one embodiment of the invention the device is
arranged to detect the passenger last arriving in the
elevator car or leaving from it and to measure the
photoelectric cell delay of the elevator on the basis
of the detected last passenger.
In one embodiment of the invention the device
comprises an interface for connecting the device to
separate systems, to which the device is arranged to
convey information about the passengers.
In accordance with the fourth aspect of the invention,
a system for detecting elevator passengers is
presented. The system comprises an elevator car and an
acceleration sensor that measures its acceleration.
The system further comprises analyzer means for
receiving the vertical acceleration values of the
elevator car from the acceleration sensor and for
detecting the passengers arriving in the elevator car
and/or leaving the elevator car on the basis of the
vertical acceleration measurements of the acceleration
sensor.

In one embodiment of the invention the analyzer means
are further arranged to calculate the calculated speed
of the elevator car from the vertical acceleration
measurements of the acceleration sensor, to preprocess
the calculated speed by rendering the speed of the
elevator car as zero elsewhere except in a passenger
loading situation or a passenger unloading situation,
and to detect the passengers arriving in the elevator
car and/or leaving the elevator car from the
preprocessed speed on the basis of the calculated
position of the elevator car. In one ernbodiment the
analyzer means are further arranged to use a movement
indicator in the preprocessing, which detects a
passenger loading situation or a passenger unloading
situation from the fluctuation of movement of the
elevator car. In one embodiment of the invention the
analyzer means are arranged to detect the passengers
arriving in the elevator car and/or leaving the
elevator car are detected from the preprocessed speed
on the basis of the calculated position of the
elevator car with the correlation method.
In one embodiment of the invention the analyzer means
are arranged to detect the passenger last arriving in
the elevator car or leaving from it, and the system
further comprises determination means for measuring
the photoelectric cell delay of the elevator.
As a result of the present invention by monitoring
vertical acceleration it is possible to detect a
passenger leaving the elevator car and arriving in the
elevator car. Additionally, the information produced
by the invention can be used in a condition monitoring
system of the elevator and in monitoring as well as in
forecasting the passenger traffic of the elevator. The
solution according to the invention can be easily

installed in both new elevators and in elevators
already in use.
LIST OF FIGURES
In the following, the invention will be described in
detail by the aid of a few examples of its
embodiments, wherein:
Fig. 1 presents a method for compensating the errors
of the acceleration signal of the elevator car;
Figs. 2a and 2b present the preprocessing of the
acceleration signal of the elevator car with segmented
bias compensation;
Fig. 3 presents the speed and position of the elevator
car calculated with segmented bias compensation;
Figs. 4, 5a, 5b, 6 present a method according to the
invention for detecting passengers with the
correlation method; and
Figs, 7a and 7b present a system according to the
invention as a block diagram.
DETAILED DESCRIPTION OF THE INVENTION
Figs. 2a and 2b present an embodiment of the invention
in which the vertical acceleration signal given by the
acceleration sensor is processed and the passengers
are detected from the processed acceleration signal.
In the embodiment of the invention presented in Figs.
2a and 2b the acceleration signal is preprocessed with
segmented bias compensation.

In the embodiment of Fig. 2a and 2b the speed is set
by default to be zero in the inspection period when
the elevator is standing at a floor. This is done
everywhere else except in those periods of time when a
passenger leaves, arrives or moves in the car. The
inspection period is e.g. the time from the moment
when the door of the elevator car has fully opened to
the moment when the closing phase of the door starts,
but it can also be defined as another period of time
suited to the purpose. The basic assumption is that
the car is in practice stationary other than when
passengers are moving in the car and at the moment
when a passenger arrives or leaves from the car. In
other words

The test function T (•) occurring in the formula above
(the so-called movement indicator), the purpose of
which is to examine movement of the car, can be
implemented e.g. with a sliding variance and with a
time window Wt of a suitable length.

The threshold value can be set automatically based
on the data material measured during the inspection
period by arranging the values of the test function
τ:(k,a(k)) into their order of magnitude and by

selecting e.g. a sampleas a threshold
value. In this way the selection of the threshold
value is made immune to a fluctuation in individual
values.
Fig. 2a presents the values of a test function
calculated from acceleration with the formula (4) when
the time window Wt is 0.5s. The arrows with the
reference 100 describe the events of the acceleration
curve, in which the test function detects movement in
the car. The dashed line in the lower part of Fig. 2a
presents the threshold value ξ of the test function.
In the other areas the speed is rendered to zero
according to the formula (3). By comparing Figs. 1, 2a
and 2b it is observed that the speed already remains
under better control, but the cumulative deviation is
still troublesome. For example at the moment 133.8
seconds (arrow 102 in Fig. 2b) the car has the
calculated speed -0.003 m/s, although the movement
indicator also says the car is stationary in the next
moment.
The application of formula (2) can be extended to
apply to each area presented in Fig. 2a with the arrow
100. In other words, since the movement indicator T(.)
says when movement of the car starts and ends, the
formula (2) can be applied in segments to each such
area separately. Although the car also inclines to
different sides during loading and unloading, the bias
terms caused by the inclination can be compensated out
in segments by means of the formula (2) such that To
is the moment when the movement indicator x(.) reports
that the movement has started and T is the moment when
correspondingly the movement of the car has ceased.

Fig. 3 presents with this "segmented bias
compensation" principle the calculated speed and
position of the car during a loading situation. Now
the bias errors (deviations) are under control and the
step-like deviations caused by passengers is clearly-
seen in the position of the car. The car has moved
less than 1 mm upwards and downwards from the position
of the starting situation; approx. 900 μm upwards when
a passenger exits and approx. 700 μm downwards when a
passenger steps into the car. An ordinary MEMS (Micro-
Electro-Mechanical Systems) acceleration sensor, or
any other sensor whatsoever with which acceleration
can be measured, can be used as an acceleration
sensor.
From the standpoint of detecting passengers the speed
of the car is better as a signal than the measured
acceleration. Likewise the position of the car is
better as a signal than the speed. The reason for this
is that speed is a magnitude integrated once from
acceleration and position is a magnitude integrated
twice from acceleration. In terms of filter technology
the position integrated from the acceleration
corresponds to a second-order low pass filter. The
vibrations and noise appearing in the original
acceleration signal smooth out effectively and the
actual transition produced by the excitation
"collects" first in the speed and then in the
position. The effect of integration is clearly seen
when the top curve (acceleration) of Fig. 2a and the
speed and position of Fig. 3 are compared. When the
bias errors (deviations) are first compensated from
the position of the car, the exits and arrivals of
passengers can easily be seen. Thus the detection of
passengers is preferably based on the signal
describing the position of the car.

Figs. 4, 5a, 5b and 6 present a method according to
the present invention for detecting passengers.
Passengers are detected with the correlation method.
In the first phase step-like changes in the position
of the elevator car are sought. This is done e.g. with
a sliding variance according to formula (2), in which
case

where x(k) is the position of the car at the sampling
moment k The 0.5s. time window presented earlier can
be used as the width of the window here also. The
sliding variance forms a rounded peak at the point of
the step-like changes of the car according to Fig. 4.
It is possible to endeavor to detect the peaks from
their amplitude. A more reliable result is achieved
however when the detection is performed e.g. by means
of correlation. In this example the peak-shaped
function is taken as the test function and it is slid
from point to along the curve τx(k) (sliding
correlation) , in which case a new curve is obtained to
describe the correlation of the test function to the
tested curve in the environment of each point.


where TF is a vector of length m (m odd) containing
samples from the test function, X is a sub-vector
taken from the vector x such that the sample x(k) is
the middlemost in the sub-vector X of length m.
Figs. 5a and 5b present the sliding correlation R(k)
between the test function TF and the sliding variance
τx of the position of the car calculated with the
formula 5. The test function TF at the time 129.5s is
also drawn in Fig. 5a and the correlation value 0
corresponding to this. The test function TF at the
time 130.45s is drawn in Fig. 5b and the correlation
value 1 corresponding to this.
Since correlation examines the correspondence of two
different functions and does not affect the magnitude
between the functions in it, an arriving and an
exiting passenger can be reliably detected with the
correlation function R{k). The peaks of the function
R{k) represent the time of the events. The nature of
the event can be ascertained reliably by examining
from the position of the car in which direction the
car has moved in the environment of the detected peak.
If the car has risen upwards, a passenger has exited
the car. Likewise, if the car has settled downwards, a
passenger has arrived in the car. In Fig. 6 the final
results obtained when a passenger exits from and steps
into the car two different times during the same stop
are collated; at the moment 125.3 s (out), 127.6 s
(in), 130.5 s (out) and 132.9 s (in). The event
subsequent to the time 132.9s (in other words,
settling of the car downwards) is caused by walking
inside the car, which "shakes" the car downwards.
All in all, the detection of passengers can be
performed in many different ways. For example, patent

US 5,518,086 (Tyni) describes a method that uses
neural networks for detecting passengers from the car
load weighing signal. The load weighing signal, either
a floor load weighing device or an upper beam load-
weighing device, corresponds in its nature to the
position signal of the car presented here. This being
the case, the method disclosed in the aforementioned
patent can be used directly in the position
information of the car now presented.
The information about passengers obtained can be used
together with other data of the condition monitoring
system and to form floor-specific traffic statistics
for the relevant elevator. By transferring elevator-
specific statistics to a servicing center, it is
possible to combine information about elevators
serving in the same group and to from the traffic
Statistics of the group. The information can also be
conveyed to the control system of the elevator group,
in which case the control system of the elevator group
can be adapted to the prevailing and/or to the
forecast traffic situation in order to enhance the
efficiency of service of the elevator group.
The solution disclosed in the present invention can be
used in new as well as in existing elevators and also
in elevators manufactured by any manufacturer
whatsoever. Traffic information at different floors
and traffic charts can be offered e.g. as an added-
value service to important customers. Monitoring and
guiding the passenger flows of buildings e.g. in
shopping centers obtains useful information about
passenger numbers.
The solution disclosed in the present invention is
used in one embodiment in condition monitoring, namely

in measuring the photoelectric cell delay. Numerous
intervals that belong to the operating cycle of an
elevator are measured and monitored in an elevator
system, e.g. run time, starting delay, run cycle time,
door-open time, door-closed time, etc. The
photoelectric cell delay is defined as the time from
the moment after the last passenger detected with the
door photoelectric cell to the moment when the doors
of the elevator start to close. Based on the solution
according to the invention for detecting passengers,
the condition monitoring system can now monitor and
supervise the behavior of the photoelectric cell delay
(information about the opening/closing of the door is
obtained e.g. from the condition monitoring system or
directly from the door operator of the elevator car) .
The photoelectric cell delay is one of the aspects
affecting the safety of passengers, ride comfort and
the performance capability of the elevator. The
inoperability of the photoelectric cell can be
detected quickly and reliably e.g. as follows: if the
door does not re-open although the passenger detected
with the acceleration sensor has walked between the
closing door, it can be interpreted as a possible
defect in the photoelectric cell. In addition the
operation of the control of the elevator can be
monitored, in other words whether the control changes
the photoelectric cell delay e.g. in peak-traffic
situations and on entry floors. Generally the solution
according to the invention can be used in connection
with condition monitoring systems for measuring
numerous indicative parameters of the operation and
the utilization rate of the elevator e.g. in assessing
the modernization need of an elevator already in use.
Furthermore, in one embodiment of the invention an
automatic emergency phone call to the service center
is made if the elevator stops between floors and there

is a passenger or passengers in the elevator car. The
condition monitoring system of the elevator knows the
position in meters of the elevator car in the shaft,
and likewise it knows the door zones and when a stop
is made between door zones { = floors) . In addition the
condition monitoring system is able to detect, based
on the acceleration signal of the car, the nature of
the stop; it is able to distinguish an emergency stop
from a normal stop. In this embodiment the condition
monitoring system can if necessary activate an
emergency phone call to the service center if it
appears that the elevator is not able to start moving
by its own efforts. At the same time the system can
supply technical data about the event, such as the
estimated number of passengers in the car, between
which floors the elevator is, the stopping method
(emergency/normal) , etc.
Fig. 7a presents one preferred embodiment of the
system according to the invention. The system of Fig.
7a comprises an elevator car 708, which has stopped at
a floor 704. There are two passengers in the elevator
car from before. A third passenger 710 is stepping
into the elevator car 708 from the floor 704. When the
passenger 710 steps into the elevator car 708, the
acceleration sensor 700 fixed to the elevator car 708
registers the vertical movement of the elevator car
708. The measurements of the acceleration sensor 700
are conveyed to the processing unit 702 along the
connection 706. The connection 706 can be a wireless
or a wired connection. It is also obvious (as an
exception to Fig. 7a) that in connection with the
elevator car 708 can be a device that collects the
measurement results gathered by the acceleration
sensor 700, and the device sends the results to the
processing unit 702. The operation of the processing

unit 702 is described in more detail in conjunction
with Figs. 2-6.
In one embodiment of Fig. 7a the processing unit 702
presents a part of a more extensive monitoring system
and/or condition monitoring system, which is
implemented in the elevator system already in its
construction stage. Fig. 7b presents a second solution
according to the invention to implement the monitoring
of passengers. In the embodiment presented in Fig. 7b
the monitoring is implemented as a separate solution
e.g. only after the construction stage. In this case
one or more interfaces 714 can be arranged to the
processing unit 702, via which information can be
obtained from the processing unit 702 e.g. for a
monitoring system 712, for a remote system of the
servicing center/service center, for the control
system of an elevator and/or an elevator group or any,
other similar separate system whatsoever. Information,
such as e.g. information about the opening/closing of
the doors, can also be conveyed to the processing unit
702 via the interface. The actual analysis of results
obtained from the acceleration sensor can be performed
with a computer program saved in a suitable memory,
which is arranged when run on a data processing
appliance to perform the analysis phases presented in
the invention.
The invention is not limited solely to the embodiments
described above, but instead many variations are
possible within the scope of the inventive concept
defined by the claims below.

CLAIMS
1. Method for detecting the arrival in the
elevator car/departure from the elevator car of
elevator passengers, characterized in that the method
comprises the phases:
the vertical acceleration values of the elevator
car are received from the acceleration sensor; and
the arrival in the elevator car and/or departure
from the elevator car of an elevator passenger are
detected on the basis of the vertical acceleration
measurements of the acceleration sensor.
2. Method according to claim 1 characterized
in that the detection phase further comprises the
phases:
the calculated speed of the elevator car is
calculated from the vertical acceleration measurements
of the acceleration sensor;
the calculated speed is preprocessed by rendering
the speed of the elevator car as zero elsewhere except
in a passenger loading situation or a passenger
unloading situation; and
the arrival in the elevator car and/or departure
from the elevator car of an elevator passenger is
detected from the preprocessed speed on the basis of
the position of the elevator car.
3. Method according to claim 2, characterized
in that in the preprocessing phase:
a movement indicator is used in the preprocessing,
which detects a passenger loading situation or a
passenger unloading situation from the fluctuation of
movement of the car.
4. Method according to any of claims 1-3
above, characterized in that in the detection phase:
the arrival in the elevator car and/or departure
from the elevator car of an elevator passenger is
detected from the preprocessed speed on the basis of

the calculated position of the elevator car with the
correlation method.
5. Method according to any of claims 1- 4
above, characterized in that:
the arrival of the passenger last arriving in the
elevator car or the departure of the passenger last
leaving it is detected; and
the photoelectric cell delay of the elevator
aforementioned is measured on the basis of the
aforementioned detection.
6. Computer program, characterized in that it
is arranged to perform the phases of the method
presented in the method claims 1 - 5.
7. Computer program according to claim 6,
characterized in that the computer program is stored
on a data processing appliance on a readable storage
medium.
8. Device for detecting the arrival in the
elevator car/departure from the elevator car of
elevator passengers characterized in that the device
is arranged to receive the vertical acceleration
values of the elevator car (708) from the acceleration
sensor (700) and to detect the arrival in the elevator
car and/or departure from the elevator car of an
elevator passenger on the basis of the vertical
acceleration measurements of the acceleration sensor
(700).
9. Device according to claim 8, characterized
in that the device is arranged:
to calculate the calculated speed of the elevator
car (708) from the vertical acceleration measurements
of the acceleration sensor (700);
to preprocess the calculated speed by rendering
the speed of the elevator car (708) as zero elsewhere
except in a passenger loading situation or a passenger
unloading situation; and

to detect the arrival in the elevator car and/or
the departure from the elevator car of an elevator
passenger from the preprocessed speed on the basis of
the calculated position of the elevator car.
10. Device according to claim 9,
characterized in that the device is arranged to use a
movement indicator in the preprocessing, which detects
a passenger loading situation or a passenger unloading
situation on the basis of the movement of the elevator
car (708).
11. Device according to any of claims 8-10
above, characterized in that the device is arranged to
detect the arrival in the elevator car (708) and/or
the departure from the elevator car (708) of an
elevator passenger from the preprocessed speed on the
basis of the calculated position of the elevator car
(708) with the correlation method.
12. Device according to any of claims 8-11
above, characterized in that the device is arranged to
detect the last passenger arriving in the elevator car
(708) or the last passenger leaving it and to measure
the photoelectric cell delay on the basis of the
aforementioned detection.
13. Device according to any of claims 8-12
above, characterized in that the device comprises at
least one interface (714) for connecting the device to
one or more separate systems; and the device is
arranged to convey information about the passengers to
at least one aforementioned separate system.
14. System for detecting the arrival in the
elevator car/departure from the elevator car of
elevator passengers, which system comprises:
an elevator car (708); and
an acceleration sensor (700) that measures the
acceleration of the elevator car (708) ;
characterized in that the system comprises
analyzer means (702) for receiving the vertical

acceleration values of the elevator car (708) from the
acceleration sensor (700) and for detecting arrival in
the elevator car (708) and/or the departure from the
elevator car (708) of an elevator passenger on the
basis of the vertical acceleration measurements of the
acceleration sensor (700) .
15. System according to claim 14,
characterized in that the analyzer means (702) are
further arranged:
to calculate the calculated speed of the elevator
car (708) from the vertical acceleration measurements
of the acceleration sensor (700);
to preprocess the calculated speed by rendering
the speed of the elevator car (708) as zero elsewhere
except in a passenger loading situation or a passenger
unloading situation; and
to detect the arrival in the elevator car (708)
and/or departure from the elevator car (708) of an
elevator passenger from the preprocessed speed on the
basis of the calculated position of the elevator car
(708) .
16. System according to claim 15,
characterized in that the analyzer means (702) are
further arranged to use in the preprocessing a
movement indicator, which detects a passenger loading
situation or a passenger unloading situation from the
fluctuation of movement of the car (708).
17. System according to any of claims 14 - 16
above, characterized in that the analyzer means (702)
are arranged to detect the arrival in the elevator car
(708) and/or the departure from the elevator car of an
elevator passenger from the preprocessed speed on the
basis of the calculated position of the elevator car
(708) with the correlation method.
18. System according to any of claims 14 - 17
above, characterized in that the analyzer means (702)
are arranged to detect the arrival of last passenger

arriving in the elevator car (708) or the departure of
the last passenger leaving it, and in that the system
further comprises determination means (702) for
measuring the photoelectric cell delay of the doors of
the elevator (708) on the basis of the arrival or the
departure of the detected last passenger.

The present invention discloses a method, a device, a computer program and a system for detecting the arrival in the elevator car/departure from the elevator car of elevator passengers. In the method the vertical acceleration values of the elevator car are received from the acceleration sensor and the passengers arriving in the elevator car and/or leaving the elevator car are detected on the basis of the vertical acceleration measurements of the acceleration sensor.