Method and device for monitoring the activity of at least one person in an infrastructure unit and measuring device

Abstract

Method for monitoring the activity of at least one person (2) in an infrastructure unit (1) supplied with utilities comprising at least electricity and water, wherein each utility is assigned a measuring device (6, 7, 8) for measuring a consumption value of the utilities, wherein the measuring devices (6, 7, 8) provide, at least for electricity and water, temporal consumption value profiles (21, 22) for a predetermined period as profile data, which are evaluated by means of at least one activity criterion that compares the profile with a reference profile describing normal activity for the same period, wherein, in the event of a failed comparison, at least one alarm action is carried out, wherein, as an alarm action, at least one further activity sensor (17) is activated and its sensor data are evaluated by an emergency criterion.upon fulfillment of which an emergency measure is initiated, characterized in that the activity sensors (17) used are floor sensors (17) detecting a person (2) lying on a floor of the infrastructure unit (1) and / or infrared sensors for fall detection and / or motion detectors and / or infrared sensors in heat cost allocators and / or breathing air sensors measuring breathing air in the infrastructure unit (1) and / or current sensors assigned to individual consumers.

Description

[0001] The invention relates to a method for monitoring the activity of at least one person in an infrastructure unit supplied with utilities comprising at least electricity and water, wherein each utility is assigned a measuring device for measuring a consumption value of the utility. The invention also relates to a corresponding device and a measuring device.

[0002] It is common practice to monitor the activity of elderly, care-dependent, or safety-conscious individuals to automatically detect changes in their health. This type of activity monitoring is often referred to as "activity tracking." The number of people requiring care and / or living alone is currently growing significantly and is projected to continue to do so in the future. Accidents in the home and deteriorating health conditions therefore represent a considerable risk to their lives.

[0003] Numerous approaches already exist to provide security for people in apartments or houses, and generally in infrastructure units, through activity monitoring. For example, daily personal visits from service providers are known to occur, although these are usually very time-consuming and expensive. Emergency call devices with an emergency button or a confirmation button are also known to be used, and these devices can be worn on the body. It is also known to install such emergency call devices within the infrastructure unit, particularly an apartment or house, with various operating concepts for these devices. For instance, an emergency call device with a button worn on the body can be activated only when an actual emergency occurs.However, it is also possible, particularly with emergency call systems permanently installed in the infrastructure unit, to provide a button or, more generally, a control element that must be pressed regularly, for example twice a day, to signal that everything is okay. Such emergency call systems, however, also require a significant investment and presuppose that the person is able to operate the corresponding controls.

[0004] Regarding the use of sensors, it has already been suggested that, for example, floor sensors be installed, which could be designed as pressure sensors to detect people lying down within the infrastructure unit. Infrared sensors for fall detection have also been proposed. Motion sensors in rooms of the infrastructure unit can also be used for activity monitoring. Finally, it has been suggested that current sensors be installed on individual appliances if, for example, the appliance in question is used regularly at specific times. A toaster is one example of this.

[0005] The problem with all these options is that additional, costly, and / or complex equipment is required to enable activity monitoring. Furthermore, many of the solutions only target specific events, meaning that—when used alone—they cannot always reliably detect emergencies and respond accordingly.

[0006] From US patent 2007 / 0152837A1, a method and device for monitoring a person's activity are known. This method employs several sensors which are activated by the person being monitored during their daily activities. Sensor activation is understood as triggering an already activated sensor. The activation of the sensors can be compared with a predefined, time-dependent activation pattern to determine whether an alarm needs to be sent to the person.

[0007] A system for the automatic triggering and management of alarms is known from US patent 2016 / 0171866A1. The system includes at least one sensor for collecting data. The collected data is transmitted via a first data connection to a central processing unit (CPU), where an analysis module (2) of the CPU compares the received data with data stored in the analysis module. An alarm can then be triggered regardless of this comparison. If no alarm is triggered, the data is transmitted to a central processing unit, which can perform a further comparison with stored data and, if necessary, trigger an alarm.

[0008] US Patent 2010 / 0145164A1 describes a remote monitoring system for tracking a person's condition. This system utilizes various sensors, which may include sensors for detecting energy or water consumption. Based on previously collected data, an activity profile of the individual can first be created as a reference profile. Subsequently, measured data can be compared with this stored reference profile. If the measured data deviates from the stored reference profile, an alarm is triggered, and a third party is notified.

[0009] US 2015 / 0061859A1 describes a system and a method for implementing security functions in a sensor-equipped smart home.

[0010] US Patent 2011 / 0153103A1 discloses a system and method for predictively displaying the energy consumption of a building. Based on predicted and measured energy consumption values, the system determines whether a threshold for the difference between predicted and measured values ​​is exceeded, triggering an alarm if so. The alarm is transmitted to a user via an electronic and / or physical signal.

[0011] EP 1 861 839 B1 discloses an auxiliary device for controlling and monitoring the operation of household appliances and entertainment devices. The auxiliary device measures the consumption of, for example, water or electrical energy and detects the presence of a person in an infrastructure unit. To detect the presence of a person, as described in paragraph 12, for example, a switch indicating whether a person is at home or not, sensors measuring bed occupancy, or a portable smart bracelet can be used.

[0012] German patent DE 10 2008 044 909 A1 proposes a method for detecting emergencies using remotely read meters. By measuring and analyzing the consumption of electricity, water, and / or gas in a household, emergencies can be detected and alarms forwarded. The analyzed consumption patterns are compared with standard values, whereby consumption profiles for average users can be used, or individually configured benchmark values ​​can be employed. An alarm is triggered if there is a significant decrease in consumption. This alarm message is then transmitted to a control center via established data channels.

[0013] The invention is therefore based on the objective of providing a means for reliable, low-effort activity monitoring of persons in infrastructure units, in particular apartments and / or houses.

[0014] To solve this problem, a method of the type mentioned at the outset provides, according to the invention, that the measuring devices provide, at least for electricity and water, temporal consumption value profiles for a predetermined period as profile data, which are evaluated by means of at least one activity criterion that performs a comparison with a comparison profile describing normal activity for the period, wherein at least one alarm measure is carried out in the event of a failed comparison by the activity criterion.

[0015] According to the invention, it is therefore proposed to analyze measurement data in the form of consumption values ​​over time and to determine, based on a current consumption profile, whether a person's activity level is normal or not. The measuring devices can be, in particular, an electricity meter and a water meter, as are generally known in the prior art. These are configured to output consumption values ​​with a sufficiently high temporal resolution, for example, at least a minute resolution, from which corresponding consumption trends over time can be derived. These trends represent consumption profiles, at least for water and electricity, as profile data. A measured consumption value preferably describes the current consumption at the time of measurement, in particular per unit of time.It should be noted that consumption trends over time can, of course, also be compared with corresponding reference trends in the case of accumulated metering or similar methods, whereby a corresponding offset may need to be taken into account. In such a case, each of the measuring devices therefore regularly, and especially cyclically, provides a currently accumulated consumption value.

[0016] A reference profile, described by reference data, is stored in a storage device of the device performing the method. This reference profile covers all time durations usable in the method. Therefore, a comparison profile for the duration, within the scope of the present invention, preferably means that the reference profile covers the entire time, and that a proportion of the reference data covering all conceivable durations, corresponding to the time duration underlying the comparison, is used for comparison with the profile data. In other words, the reference data underlying the comparison, corresponding to the comparison profile for the duration, are selected from the reference data describing the reference profile for all usable time durations, thus enabling a meaningful comparison.

[0017] The reference profile, and therefore also the comparison profile, describes the normal activity of at least one person in the infrastructure unit. By comparing the current profile data with a corresponding comparison profile of the reference data, it can be determined whether, and especially to what extent, deviations from normal activity exist. If no normal activity is present, i.e., the comparison fails, it can be assumed that a critical situation may exist, which is why at least one alarm measure is taken, for example, a request for the person to take action themselves to confirm the existence of an emergency, or the sending of information to other persons or institutions.

[0018] The problem underlying the invention is thus solved by utilizing existing infrastructure, so that no additional sensors and / or data channels are required. Only the time-measured consumption values ​​of water and electricity are analyzed, and normal behavior is distinguished from a critical condition. This allows for a reliable and cost-effective monitoring of a person's activity level.

[0019] The existing consumption data from electricity and water meters is used to monitor activity. Such meters (especially electricity and water meters) are already installed in the relevant infrastructure units, particularly houses and / or apartments, as this type of consumption measurement is mandatory in many countries. These meters generate sufficiently high-resolution consumption data over time, which can be transmitted via a data interface. It should be noted that the basis of the activity monitoring described here is, of course, the data sovereignty of the user, i.e., the individual. This means that data protection can be ensured by agreeing upon the use of the consumption data for personal benefit beforehand.

[0020] Time-based consumption patterns for water and electricity are characterized by regularly or statistically distributed recurring consumption events. For example, typical consumption events regarding the amount and time of water use include the use of a shower, toilet, sink, bathtub, kitchen sink, dishwasher, washing machine, and garden irrigation with a watering can and / or hose. For electricity, examples include the use of a stove, oven, microwave, dishwasher, washing machine, freezer, refrigerator door, television, radio, computer, and lights. Thus, normal daily life exhibits a typical, and especially person-dependent, profile. The recurring patterns that occur are described by the reference data, i.e., the reference profile. The multitude of usage patterns makes it possible to quickly and reliably identify atypical situations.It is particularly advantageous to consider water and electricity consumption together, as they are often correlated. For example, certain actions involving water, such as filling a container with water for boiling, are often directly or indirectly related to actions involving electricity (switching on the stove or a kettle), and vice versa (running a dishwasher and its corresponding water consumption). Furthermore, people often follow specific daily routines, such as showering after their morning ablutions, then preparing breakfast using electrical appliances, then showering again, etc., so that a coherent overall picture emerges when examining consumption trends for electricity and water.

[0021] Specifically, this can involve determining a correlation value describing the agreement with the comparison profile and comparing it to a threshold value. In other words, a correlation between the profile data and the reference data of the comparison profile can be carried out using a generally known method, specifically focusing on the consumption trends. Various influencing factors can be weighted more or less heavily, since, for example, the number of toilet visits and / or the times of use of certain electrical appliances can vary, even if only slightly, but should typically occur within a specific time frame. The result of the correlation analysis is a correlation value that describes the extent to which consumption structures and patterns expected based on the reference data are reflected in the profile data.The threshold value against which the correlation value is compared ultimately describes how "sharp" the activity criterion is; in particular, in embodiments of the present invention, it can also be user-adjustable. If the correlation value is above the threshold value, normal activity is assumed; if it is below the threshold value, abnormal activity is present, so that at least one alarm measure is triggered accordingly.

[0022] In a particularly preferred embodiment of the present invention, person-specific reference data are determined, especially during a learning phase, by evaluating the person's profile data as training data using an artificial intelligence learning algorithm and / or statistical data analysis. People often differ in the nature, sequence, and timing of their regularly recurring activities, particularly those involving electricity and / or water consumption. The reference data are therefore preferably determined for the at least one person to be monitored and describe their normal, person-specific activity.Preferably, the reference profile, from which the comparison profiles for the usable time periods can be derived, is determined in a self-learning phase. This involves generating activity knowledge from the training data, which is determined as profile data from the learning phase, using "smart data analytics." Artificial intelligence algorithms are preferably used to implement so-called machine learning. The term "smart data analytics" comprises "smart data" for structured training knowledge and "analytics" for machine learning techniques, particularly those using artificial intelligence, such as a neural network. However, statistical data analysis methods can also be used additionally or alternatively to enable machine learning based on training data.The result could be, for example, average consumption value trends for at least one person as a reference profile, to which deviation information about commonly occurring deviations and their range is assigned.

[0023] In this context, it is useful if, when an activity criterion is met, the reference data is updated using the profile data on which that fulfillment is based. In other words, the reference profile, and thus the reference data, can be continuously updated and, so to speak, "maintained" in order to adapt it to changes in lifestyle habits, for example.

[0024] In this context, it is also advantageous if the reference data is determined based on the number of people in the infrastructure unit and / or the calendar day, with reference data related to the current number of people and / or the current calendar day being used for comparison in the activity criterion. In other words, this means that the reference data is determined broken down according to specific influencing factors, such as the number of people actually present in the infrastructure unit and / or the calendar day, so that the appropriate reference data can be used for a given number of people and / or calendar day, and thus this data is significantly more accurate and better broken down, further increasing the reliability of activity monitoring. Regarding the calendar day, distinctions can be made, for example, between weekdays, weekends, holidays, different seasons, and the like.

[0025] Furthermore, a possible and expedient configuration allows for the selection of currently usable reference data based on user input describing at least one of the at least one influencing factor, provided reference data is available that is broken down according to at least one influencing factor. The user input can, for example, be provided by the at least one person themselves at an input device, perhaps coded according to specific, defined sets of influencing factors. Thus, a person can, for example, transmit information such as "I am alone," "I am not at home," "I have guests at home," "I am in bed today," and the like to the device performing the process, which can then select the actually relevant reference data for comparison from the reference data broken down according to at least one influencing factor.It should be noted that for at least some of the influencing factors used, such as the calendar day or even the persons present, automatic determination may be provided for and is preferred.

[0026] The comparison can be performed for multiple time periods and / or at least one moving time window. This means that, firstly, it is possible to conduct the comparison on different "time scales" to identify both long-term and short-term trends. Alternatively or additionally, it is useful to use a moving time window, which, for example, could contain a specific number of hours and be continuously tracked, in order to detect a critical situation as quickly as possible in the event of a significant change in consumption patterns.

[0027] As already mentioned, it can also be useful if the activity criterion is user-adjustable with regard to the success threshold for the comparison. This is particularly useful if it should be possible to adjust the stringency of the analysis based on the activity criterion. In this way, reliability can ultimately be weighed against the risk of false triggers.

[0028] The activity criterion can expediently also evaluate consumption profiles for at least one additional consumption medium, in particular heat and / or gas, and / or sensor data from at least one additional activity sensor. It is therefore advantageous, at least when further consumption profiles are available—for example, if a heat meter and / or a gas meter are present as additional measuring devices—to use this information for the evaluation within the activity criterion, since typical consumption patterns, especially those correlated with electricity and water consumption, can emerge here as well. These patterns can be learned and used for comparison with current profile data. In addition to the consumption profiles for water and electricity, the profile data can then also include consumption profiles for heat and / or gas (and / or other consumption media not mentioned here).This further increases reliability.

[0029] It should be noted at this point that the performance of the activity monitoring described here scales with the number of consumption sources considered. If only one measurement were used, for example, only water as a consumption source, monitoring would be possible with very simple means, but rather inaccurate. Monitoring more consumption trends increases the effort, but also the performance (less time delay before inactivity is detected, fewer false alarms). Therefore, it is conceivable to further develop the invention by providing modular expansion packages (software and / or hardware) for a basic package that enables monitoring of consumption trends for electricity and water, for example, at least for gas and / or heat. This would then allow for retrofittable extensions to increase performance.

[0030] Another way to increase reliability is to use additional sensors, but this is less preferred because, due to the activity monitoring described here, these can generally be avoided, at least in continuous operation.

[0031] Nevertheless, it should be noted that deviations from normal consumption patterns typically only become sufficiently clear after a certain period of time. This means that the detection of unusual conditions based solely on electricity and water consumption (and possibly other consumption) takes place within a few hours. If the latency is to be improved, it is certainly possible to use additional sensors.

[0032] According to the invention, at least one additional activity sensor is activated as an alarm measure, and its sensor data is evaluated by an emergency criterion. Upon fulfillment of this criterion, an emergency measure, in particular contacting an emergency service and / or a contact person, is initiated. Therefore, to save effort, additional sensors can certainly be provided, but they do not need to be permanently operational; rather, they can be activated only if the activity criterion fails. If a clear emergency can be detected by these additional sensors, an emergency measure can be initiated, for example, notifying an emergency service, in particular a medical emergency service, and / or contacting another contact person.

[0033] Examples of such additional activity sensors that can be used within the scope of the present invention include floor sensors that detect a person lying on the floor of the infrastructure unit, and / or infrared sensors for fall detection, and / or motion detectors, and / or infrared sensors in heat cost allocators, and / or breathing air sensors that measure the CO2 content of the air in the infrastructure unit, and / or current sensors assigned to individual consumers. The use of such activity sensors is already widely known in the prior art and need not be described in more detail here. However, according to the invention, the additional activity sensors used are those that are typically already provided or should be provided in infrastructure units.For example, it has been proposed that in the future infrastructure units, especially apartments and / or houses, breathing air sensors that measure the CO2 content should be provided anyway.

[0034] Preferably, as an alarm measure, a prompt for user input can be issued to the person via an output device, in particular an output device assigned to the infrastructure unit and / or a person. If the person does not perform an input within a specified time period, an emergency measure, in particular contacting an emergency service and / or a contact person, is initiated. To avoid false alarms, it is conceivable to inform other persons only if, within a defined time period, at least one person has not activated an element indicating that no critical condition exists.Such a prompt can be issued, for example, via an output device of the infrastructure unit, such as a computer, television, or the like, or via an output device assigned to the individual, such as a mobile phone. Corresponding controls can also be implemented on specially designated control panels or on the output device itself. Particularly in the context of an adjustable success threshold for comparing the activity criterion, triggering can be made more stringent, or it is possible to shorten the time durations and thus the reaction time, with the risk of more frequent prompts to press the control panel as an "everything is OK" button.

[0035] Of course, it is also possible, in principle, within the scope of the present invention to provide for an alarm notification to be sent to a contact person and / or an emergency service. For example, a person related to the at least one person being monitored can be notified and / or a professional service provider can be activated.

[0036] In addition to the method, the present invention also relates to a corresponding device for monitoring the activity of at least one person in an infrastructure unit that is supplied with consumption media comprising at least electricity and water, wherein each consumption medium is assigned a measuring device for measuring a consumption value of the consumption medium, which has: - a data interface for receiving profile data containing temporal consumption value trends, at least for electricity and water, for a predetermined period of time, which are provided by the measuring devices, - a storage device for storing reference data describing a reference profile that describes normal activity, and - an evaluation device for evaluating the profile data by means of at least one activity criterion that performs a comparison of the profile data with a comparison profile as part of the reference profile for the duration of the reference data and for triggering at least one alarm measure in the event of a failed comparison by an activity criterion.

[0037] All aspects of the inventive method can be applied accordingly to the inventive device, so that the aforementioned advantages can also be achieved with it. Various data interfaces can be used to obtain the consumption values ​​or consumption value trends from the measuring devices.For example, local wireless interfaces such as Bluetooth or Wireless-M-Bus or similar technologies can be used if the device is at least partially installed locally; of course, wired interfaces are also conceivable, at least partially, particularly if the device is implemented as part of a measuring system and / or a wired connection is to be used at least partially, for example, in the case of a remotely located device, especially one responsible for several infrastructure units, such as a server that receives consumption data at least partially via the internet. Clearly, many different configurations of the device are conceivable, although it is preferable to provide a central device that monitors people in different infrastructure units.

[0038] However, it is also conceivable, particularly with the increasing intelligence in measuring devices (keyword "smart meter"), to implement the functionality according to the invention in a measuring device. A measuring device according to the invention for electricity or water as a consumption medium to be supplied to an infrastructure unit thus comprises a device according to the invention. The explanations regarding the method and the device naturally continue to apply in this respect as well.

[0039] Further advantages and details of the present invention will become apparent from the exemplary embodiments described below and from the drawings. These show: Fig. 1 a system for carrying out the method according to the invention, Fig. 2 a flowchart of an embodiment of the method according to the invention, and Fig. 3 an example consumption value curve.

[0040] Fig. Figure 1 shows a system in which the method according to the invention can be carried out. An infrastructure unit 1 is indicated, which in this case can be an apartment or a house. The infrastructure unit 1 is inhabited by at least one person 2 whose activity is to be monitored in order to detect critical, unusual conditions of the person 2. To supply the infrastructure unit 1 with utilities, various supply lines 3, 4, 5 are provided, in this example a supply line 3 for water, a supply line 4 for electricity, and a supply line 5 for gas. Each of the supply lines 3, 4, 5 is assigned a measuring device 6, 7, 8 for recording the consumption of the respective utilities. Thus, the measuring device 6 is a water meter, the measuring device 7 is an electricity meter, and the measuring device 8 is a gas meter.The measuring devices 6, 7, 8 can determine both an accumulated consumption value and a current consumption value related to a unit of time.

[0041] The energy consumption is carried out by corresponding appliances 9, 10, and 11 used by Person 2, of which only a few examples are shown here: electrical appliances 9, water appliances 10, and gas appliances 11. Electrical appliances 9 can include, for example, a stove, an oven, a microwave, a dishwasher, a washing machine, a freezer, a refrigerator, a television, a radio, a computer, and / or light sources. Water appliances 10 include, for example, a shower, a toilet, a washbasin, a bathtub, a kitchen sink, a dishwasher, a washing machine, and a garden tap. Gas appliances 11 include, for example, a stove, a gas boiler for heating, and a gas boiler for hot water.

[0042] Person 2 exhibits typical usage patterns within infrastructure unit 1, which can be described by a reference profile based on their recurring patterns. Such a reference profile can be determined by evaluating profile data over an extended period as training data, employing methods such as machine learning, particularly artificial intelligence, and / or statistical analysis. The reference data describing the reference profile is broken down by the number of people in infrastructure unit 1 (if applicable) and by calendar day, representing a full day at a time. This allows for the extraction of comparative profiles from the reference profiles for predefined time periods.

[0043] Current reference data describing the current reference profile of person 2 are stored in a storage device 12 of a device 13 according to the invention, which is shown here external to the infrastructure unit 1, but in other embodiments can also be located within the infrastructure unit 1, in particular even as part of one of the measuring devices 6 or 7. In the embodiment according to Fig. In this document, device 13 is configured as a central server on which activity monitoring for multiple infrastructure units 1 and corresponding personnel 2 can be performed. To enable this, device 13 is connected to measuring devices 6, 7, and 8 via communication links 14, which also include the internet. Relevant consumption values ​​can be received via a data interface 15, thus generating consumption trends over time for at least water and electricity. However, it is expedient and intended here to also consider other consumption trends, for example, regarding gas consumption (measuring device 8), both in the reference data and during activity monitoring.

[0044] The resulting consumption value profiles are fed as profile data to an evaluation unit 16 of the device 13, which evaluates the profile data using an activity criterion. This criterion ultimately compares the current consumption value profiles with a comparison profile extracted from the reference data for a specific period of time—that is, a selected portion of the reference data, which will be discussed in more detail below. If this comparison fails, an exceptional situation is assumed, and alarm measures are initiated. One of these alarm measures utilizes an output device 18, in the form of a mobile phone 19, assigned to person 2. Person 2 is prompted to make an input using control elements 20 to indicate that everything is nevertheless in order.

[0045] It should also be noted that additional activity sensors 17 may be present in infrastructure unit 1, preferably activity sensors 17 that are already installed, for example, a breathing air sensor that measures the CO2 content of the air in various rooms of infrastructure unit 1. Sensor data from such activity sensors 17 can also be included in the aforementioned activity criterion or used subsidiarily as an alarm measure.

[0046] Fig. Figure 2 shows a flowchart of an embodiment of the method according to the invention, as it can be carried out by the device 13. In step S1, consumption profiles are recorded, at least for water and electricity, using the measuring devices 6, 7, and optionally also for gas using the measuring device 8. These consumption profiles are evaluated in step S2 using an activity criterion. The consumption profiles relate to at least one time period. For this time period, for example, an interval of several hours within a day, a portion of the reference data corresponding to the duration, which thus represents a comparison profile, is selected. The comparison profile describes normal activity of person 2 in the infrastructure unit 1.It should be noted that the reference data for the comparison profile do not necessarily have to be consumption value trends, but can also describe their characteristic patterns and relationships, in particular cross-correlations between water consumption and electricity consumption (and, where applicable, gas consumption). At least one of the time periods is defined as a sliding time window.

[0047] As part of the activity criterion in step S2, the profile data for the specified time period are compared with the reference data describing the comparison profile for the same period by determining a correlation value between the profile data and the reference data describing the comparison profile. This correlation value indicates the extent to which the patterns and relationships contained in the profile data correspond to the patterns and relationships in the reference data describing the comparison profile. In other words, the correlation value is a measure of agreement with regard to patterns and relationships.

[0048] This is in view of the exemplary consumption value curves 21, 22 of the Fig.3 explained in more detail. It should be noted that consumption curves 21 and 22 are greatly simplified and illustrated and do not necessarily correspond to actual consumption curves. Consumption curve 21 is for water, and consumption curve 22 is for electricity. A situation in the morning is shown. Person 2 typically gets up and uses the toilet, with the water consumption from flushing symbolized by peak 23, and the subsequent handwashing by peak 24. Person 2 then goes to the kitchen of infrastructure unit 1 to fill up a container of water (peak 25) and to use a coffee machine or kettle (peak 26 of electricity consumption curve 22). The remaining electricity consumption curve 22 is characterized by the use of various electrical appliances 9 in the kitchen (area 27).After breakfast, person 2 leaves the kitchen of infrastructure unit 1 and goes back to the bathroom to shower, peak 28 of the water consumption curve 21. If person 2 then blow-dries their hair, this can be seen again as a peak 29 in the electricity consumption curve 22.

[0049] It becomes clear that a multitude of recurring and / or correlated events occur in the normal daily routine of person 2. This correlation also extends to relationships between the consumption patterns 21 and 22 for water and electricity, revealing sequences or temporal coincidences, for example, regarding the use of a coffee machine or kettle, or washing machines and dishwashers. Therefore, it is the combined analysis of consumption patterns 21 and 22, at least for water and electricity, that lends the described approach its reliability and robustness. For example, through "smart data analytics" using machine learning, these patterns and relationships can be identified, extracted, and summarized in the reference data as a reference profile, differentiating between calendar days (seasons, holidays, weekends, weekdays, etc.).) and also the number of people actually present 2.

[0050] Especially when reference data broken down by influencing factors (e.g., comprehensive calendar day, number of people, basic activity information) is available, a step S0 may be conceivable prior to step S1 if not all influencing factors can be determined automatically anyway. In this step, user input describing at least one influencing factor is received, which can then be taken into account in step S2 when selecting the appropriate reference data.

[0051] In step S2, the activity criterion is considered fulfilled if the correlation value exceeds a threshold. This threshold, as well as the length of the at least one time period considered, can be adjusted by the user to determine how easily or heavily the activity criterion is triggered. If the activity criterion is fulfilled, normal activity of person 2 is assumed, and new profile data is collected again in step S1 for the next cyclical monitoring step.

[0052] If the activity criterion, more precisely the comparison, fails, the process proceeds to step S3. There, at least one alarm action is carried out, specifically a particular alarm action in which a prompt for user input is issued to person 2 via output device 18. It should be noted that alternatively or additionally, fixed output devices assigned to infrastructure unit 1 can also be used. Then, in step S4, the system waits for a predetermined period. If the user input occurs within this predetermined period, i.e., if person 2 confirms that everything is OK, the process continues with step S1. However, if person 2 does not confirm that everything is OK, an emergency is assumed, and at least one emergency action is executed in step S5.Such an emergency measure may include, in particular, contacting an emergency service, such as an ambulance service and / or other professional service provider, and / or contacting a contact person, such as a relative.

[0053] It should also be noted that other alarm measures may of course be provided additionally or alternatively, for example, contacting a contact person and / or an emergency service and / or activating activity sensors 17 or evaluating sensor data from activity sensors 17.

[0054] Finally, as a further example, it should be noted that many recurring patterns in a household regarding consumption can also be linked to a refrigerator. A refrigerator represents a central element of domestic life. Therefore, an extension of the invention could also provide for the additional measurement of the energy consumption of an interior lighting device of the refrigerator using a dedicated sensor, from which further patterns suitable for refining activity monitoring could be derived. Reference symbol list 1 infrastructure unit 2 Person 3 Supply line 4 Supply line 5 Supply line 6 Measuring device 7 Measuring device 8 Measuring device 9 consumers 10 consumers 11 consumers 12 Storage device 13 Device 14 Communication link 15 Data interface 16 evaluation units 17 Activity sensor 18 Output device 19 Mobile phone 20 Control element 21 Consumption value trend 22 Consumption value trend 23 Peak 24 Peak 25 Peak 26 Peak 27 area 28 Peak 29 Peak S1 - S5 step

Claims

Method for monitoring the activity of at least one person (2) in an infrastructure unit (1) supplied with utilities comprising at least electricity and water, wherein each utility is assigned a measuring device (6, 7, 8) for measuring a consumption value of the utilities, wherein the measuring devices (6, 7, 8) provide, at least for electricity and water, temporal consumption value profiles (21, 22) for a predetermined period as profile data, which are evaluated by means of at least one activity criterion that compares the profile with a reference profile describing normal activity for the same period, wherein, in the event of a failed comparison, at least one alarm action is carried out, wherein, as an alarm action, at least one further activity sensor (17) is activated and its sensor data are evaluated by an emergency criterion.upon fulfillment of which an emergency measure is initiated, characterized in that the activity sensors (17) used are floor sensors (17) detecting a person (2) lying on a floor of the infrastructure unit (1) and / or infrared sensors for fall detection and / or motion detectors and / or infrared sensors in heat cost allocators and / or breathing air sensors measuring breathing air in the infrastructure unit (1) and / or current sensors assigned to individual consumers. Method according to claim 1, characterized in that, within the framework of the comparison, a correlation value describing the agreement with the comparison profile is determined and compared with a threshold value. Method according to claim 1 or 2, characterized in that person-specific reference data are determined in a learning phase by evaluating profile data of the person (2) as training data in the learning phase using a learning algorithm of artificial intelligence and / or statistical data evaluation. Method according to claim 3, characterized in that, when the activity criterion is met, the reference data are updated using the profile data underlying the fulfillment. Method according to claim 3 or 4, characterized in that the reference data are determined depending on a number of persons in the infrastructure unit (1) and / or on a calendar day, wherein reference data relating to the current number of persons and / or the current calendar day are used for the comparison in the activity criterion. Method according to one of the preceding claims, characterized in that the comparison is carried out for several time periods and / or for at least one sliding time window as a time period and / or the activity criterion is user-adjustable with regard to a success threshold for the comparison. Method according to one of the preceding claims, characterized in that the activity criterion additionally evaluates consumption value profiles (21, 22) for at least one further consumption medium and / or sensor data of at least one further activity sensor (17). Method according to one of the preceding claims, characterized in that, as an alarm measure, a request for an operator input is issued to the person (2) via an output device (18), wherein if the person (2) does not perform an operator input within an operating time period, the emergency measure is initiated. Method according to one of the preceding claims, characterized in that, as an emergency measure, contact is initiated with an emergency service and / or a contact person. Method according to one of the preceding claims, characterized in that, as an alarm measure, an alarm information is sent to a contact person and / or an emergency service. Device (13) for monitoring the activity of at least one person (2) in an infrastructure unit (1) supplied with utilities comprising at least electricity and water, wherein each utility is assigned a measuring device (6, 7, 8) for measuring a consumption value of the utility, comprising: - a data interface (15) for receiving profile data containing temporal consumption value trends (21, 22) at least for electricity and water for a predetermined period of time, which are provided by the measuring devices (6, 7, 8),- a storage device (12) for storing reference data describing a reference profile describing a normal activity and - an evaluation unit (16) for evaluating the profile data by means of at least one activity criterion for comparing the profile data with a comparison profile as part of the reference profile for the duration of the reference data and for triggering at least one alarm measure in the event of a failed comparison by an activity criterion, wherein the alarm measure is the activation of at least one further activity sensor (17) and an evaluation of its sensor data by an emergency criterion, wherein an emergency measure is initiated upon fulfillment of this criterion, characterized in thatthat the activity sensors (17) used are floor sensors (2) detecting a person lying on the floor of the infrastructure unit (1) and / or infrared sensors for fall detection and / or motion detectors and / or infrared sensors in heat cost allocators and / or breathing air sensors measuring breathing air in the infrastructure unit (1) and / or current sensors assigned to individual consumers. Measuring device (6, 7, 8) for electricity or water as a consumption medium to be supplied to an infrastructure unit (1), comprising a device (13) according to claim 11.

Images