WO2024062284A1

WO2024062284A1 - Multifunctional assisstance system, in particular for persons living alone

Info

Publication number: WO2024062284A1
Application number: PCT/IB2023/000645
Authority: WO
Inventors: Bettina Horster; Thomas HORSTER-MOELLER
Original assignee: VIVAI Software AG
Priority date: 2022-08-31
Filing date: 2023-08-31
Publication date: 2024-03-28

Abstract

Multifunctional assistance system, in particular for persons living alone, comprising a number of sensors which are distributed in a living area of a person using the assistance system for the purpose of monitoring predefined activities and predefined parameters, a control and communication unit that is provided in the living area for the purpose of communicating with the sensors and the person, a database that stores non-sensor-based information relating to the person and at least two different alarm chains, and an evaluation unit for evaluating sensor data transmitted via the control and communication unit, wherein the evaluation unit can access the database and is designed to identify any anomalies in the behaviour and/or condition of the person from the sensor data, to classify the anomalies in at least two categories and to trigger an alarm according to at least one of the at least two alarm chains on the basis of the categorization.

Description

MULTIFUNCTIONAL ASSISTANCE SYSTEM, ESPECIALLY FOR PEOPLE LIVING ALONE

TECHNICAL FIELD OF THE INVENTION

The invention relates to a multifunctional assistance system, especially for people living alone, as well as a control and communication unit designed for this purpose.

BACKGROUND OF THE INVENTION

Due to changing lifestyles and increasing ages, the topic of self-determined living in old age has become increasingly relevant in recent years from both a technical and an overall economic perspective. Even if they are mentally fit, many people experience at least certain physical impairments as they get older. Nevertheless, many people still want to live in their own apartment or house for as long as possible, especially since both nursing staff and nursing places are scarce, and in many societies it has long been no longer common for three or even four generations to live under one roof life.

In order to make life easier in a familiar environment, especially for people with certain physical limitations, a wide variety of technical products are already offered under the keyword Ambient Assisted Living (AAL, sometimes also called Active & Assisted Living). Cogvis Software & Consulting GmbH, 1040 Vienna (www.cogvis.ai) offers body-contactless fall sensors that can detect a person's movements in the rooms of an apartment using infrared measurements. On this basis, services and systems from this manufacturer offer automatic functions for fall detection and alarm calls when a fall is detected. The body-contactless fall sensors from Vayyar from Israel (www.vayyar.com) offer similar functions. US 2021/0074138 A1, whose disclosure content is referred to in particular with regard to the technical implementation of automated fall detection, as can also be used in the invention described here, discloses a room monitoring system that uses infrared and other sensors, such as a LIDAR sensor, which records movements and evaluates them in a Kl-truncated manner.

From EP 1 566 782 A1 a fall detection system is known in which a home robot drives to a person who has fallen and communicates with them, for example to trigger one or more calls for help. For this purpose, the European standard DIN EN 50134-2 (VDE 0830-4-2) specifies requirements for dispatch rates of corresponding help call systems.

From EP 3 828 854 A1 a concept for personalizing a fall detection algorithm for a specific person is known, in which sensor data together with feedback information are transmitted to an external device that generates update information for the fall detection algorithm. The update information is then used by the fall detector to update and thereby personalize the fall detection algorithm.

In addition, there are countless "smart home products" for an "intelligent" home that can typically be operated wirelessly via the Internet and/or WLAN and exchange information with other devices, including, for example, light switches, heating controls, sensors for monitoring the switching on of power consumers and adapter plugs for switching on power consumers. So-called smart speakers are known for voice control of such products, which make it possible to evaluate voice commands using voice commands and convert them into corresponding control commands for the products.

The following technical aids are known in particular to support the offering of services, functions or applications via the Internet with a portal and/or voice call function:

Keycloak is an open source software product published at www.keycloak.org to provide single sign-on with identity and access management for

Allow applications and services. Apache Kafka is an open-source software product published at kafka.apache.org to process data streams and provide an interface for loading and exporting data streams to third-party systems.

Microsoft Azure is a Microsoft cloud computing platform available via http://azure.microsoft.com that can be used to offer applications and services.

A cloud communication platform is offered at www.twilio.com, in particular with a programming interface for making and receiving voice calls.

Software and platform solutions are offered at www.traefik.io/traefik/ to make services and applications available, particularly from a portal.

DISCLOSURE OF INVENTION

Although the well-known products definitely fulfill their respective purposes, there is still a lack of an integrative solution that allows not only falls but also other anomalies in a person's daily life to be recognized and also to combine various functions of assisted living in such a way that They can also be used easily and safely by non-tech savvy people. Based on this, the invention is based on the technical task of specifying a corresponding assistance system that can not only combine various functions but also allow them to interact synergistically, as well as a control and communication unit designed for such a system.

The task is accomplished by a multifunctional assistance system with the features of claim 1 or a control and communication unit with the features of claim 24. Advantageous refinements and further developments are the subject of the subclaims.

An assistance system according to the invention includes a number of sensors that are used for monitoring in a living area of a person using the assistance system predefined activities and predefined parameters are distributed. It is suitable for use in your own premises as well as for use in nursing homes, hospitals and assisted living facilities. Depending on the individual setup of the system, it can also be used advantageously in hotels, for example. The term “living area” must therefore be interpreted broadly.

The system further comprises a control and communication unit provided in the living area for communication with the sensors and the person, a database in which non-sensor-based information about the person and at least two different alarm chains are stored, and an evaluation unit for evaluating sensor data transmitted via the control and communication unit, whereby the evaluation unit can access the database and is designed to identify any anomalies in the behavior and/or state of the person from the sensor data, to classify the anomalies into at least two categories and, based on the categorization, to trigger an alarm in accordance with at least one of the at least two alarm chains. An "alarm chain" is understood to mean a computer-readable address and/or instruction list stored in the database, which can be used by the evaluation unit to trigger an alarm in such a way that certain people and/or facilities are informed in a predetermined order. The term "alarm" is understood to mean any form of contact to transmit the information that an anomaly exists. An alarm can be, for example, an automated call to a predetermined person and/or sending a text message to that person. If the person does not respond immediately or, depending on the anomaly detected, within a certain time interval, the alarm is sent to the next person or facility.

In the event of certain behavioral and/or condition anomalies, different people and/or facilities can also be informed simultaneously and/or at least regardless of whether the previous "link" in the alarm chain responded or not. For example, if the assistance system detects a potentially life-threatening situation, automated messages can be sent simultaneously and independently to the emergency doctor, a caretaker and a relative of the person concerned. At least one of the alerting chains can be a dynamically updated alerting chain, ie it can be predeterminable criteria, for example people and/or facilities that are named in an update chain, store and change information about their accessibility and/or it can be provided to locate at least certain people included in an alert chain and to update the chain in such a way that The person closest to the living area monitored by the assistance system is informed first.

The term “database” can be understood to mean a memory of any design that can be accessed by the evaluation unit and can be physically distributed among different units. The term “non-sensor-based information” is understood to mean all types of data relevant to the respective application that do not result from the sensors arranged in the living area, such as the name and age of the person using the assistance system, their medical history, certain current ones Medications, if necessary information about the person's hobbies and life history, etc. The term "abnormalities in the person's behavior and/or condition" is understood to mean all types of deviations from the person's respective "normal" behavior or condition. An anomaly in this sense can be, for example, a fall, but also an unusually high blood pressure, a change in toilet behavior, sudden night activity, etc. An anomaly in this sense can also be a cry for help from the person who uses voice control to trigger the assistance system to trigger an alarm activated. The term “divide into at least two categories” is understood here to mean that the detected anomalies are categorized according to predeterminable criteria, for example according to whether the respective anomaly indicates a possibly life-threatening condition of the person using the assistance system, so that immediate measures such as, for example the alerting of an emergency doctor must be initiated, or whether the anomaly indicates a gradual deterioration in the person's health, for example the onset of dementia. For such anomalies that are not immediately life-threatening, "silent alarms" are triggered, depending on the default settings For example, inform a doctor or nursing staff caring for the person using the assistance system that changes in the person's behavior indicate a deterioration in their health so that appropriate measures can be taken if necessary. In this sense, the assistance system can advantageously be understood as a diagnostic tool that enables appropriate specialist personnel to make diagnoses.

The term “evaluation unit” is not to be understood as a fixed physical unit, for example in the form of a dedicated computer, although it can also be implemented in this way. Typically, the “evaluation unit” is a software component, preferably a dynamically changing, self-learning artificial intelligence that uses specific hardware to carry out the desired evaluation and execute the steps provided in the alerting chain. In this sense, it can be a cloud-based Internet platform that is set up as an evaluation unit according to the invention by appropriate configuration. The hardware can be a cluster or network of hardware units that may take on specific tasks, such as speech recognition or speech generation, or through which corresponding existing services and modules are accessed. The control and communication unit can be coupled to the evaluation unit in such a way that the evaluation unit also checks the status of the control and communication unit, for example the up-to-dateness of the software installed there, and, if necessary, initiates automatic updates. However, it can also be provided that the control and communication unit can carry out such tests itself. The invention advantageously offers an enormous degree of flexibility, so that the assistance system can be easily and cost-effectively adapted to the respective application. For example, if a person moves into an apartment that is already equipped with the assistance system and in which a person previously lived who spoke a different language than the person moving in, the assistance system can simply access a different language module. For example, if new sensors are installed in the monitored living area, the control and communication unit can automatically search for software interfaces/communication protocols required for communication with the sensors. The evaluation unit collects and stores the data transmitted to it via the control and communication unit, preferably in a protected area, preferably cloud-based and in any case not on site in the monitored living area, so that a high level of data security can easily be achieved even if the control system fails. and communication unity can be ensured. The communication between the evaluation unit and the control and communication unit preferably takes place wirelessly via at least one telecommunications network, whereby it can advantageously be provided that the control and communication unit is equipped with its own communication device, for example an LTE device, and does not require a router or a local network .

The multifunctional assistance system is designed especially for people living alone, but can also be used by couples, for example, without any restrictions. It is characterized by its excellent retrofitting capability and generally requires no complex conversion or installation work. It can use standard sensors that communicate with the control and communication unit using standardized communication protocols.

With the multifunctional assistance system according to the invention, it becomes possible to "break open" previous "data silos" (of the individual sensors used, such as fall sensors, and the individually used devices, such as smart speakers), and to view the data cumulatively, so that new insights can be gained For example, the assistance system according to the invention can be designed in such a way that it determines that after a change in the medication of the person using the system, the person no longer eats regularly and/or stays in bed for a very long time. For this purpose, for example, a refrigerator sensor, a drawer sensor and a A motion sensor that monitors the bed can be evaluated together with a medication plan stored in the database. The assistance system allows the so-called "ADLs" (Activities of Daily Living - the standard and a measuring system according to which the medical service checks, such as a person's condition and whether he or she receives a level of care).

Further details and advantages of the invention result from the following purely exemplary and non-limiting description of an exemplary embodiment in conjunction with the drawing comprising three drawing figures.

BRIEF DESCRIPTION OF THE DRAWING Fig. 1 shows schematically an apartment equipped with the assistance system according to the invention.

Fig. 2 shows schematically a combination of an evaluation unit with a control and communication unit.

Fig. 3 shows schematically a method according to the invention.

DESCRIPTION OF PREFERRED EMBODIMENTS

1 shows an apartment in which a number of sensors for monitoring predefined activities and predefined parameters are distributed and in which a control and communication unit 10 is provided for communication with the sensors and the person. The database belonging to the assistance system and the evaluation unit are not shown.

The apartment shown has a hallway 12, a kitchen 14, a living room 16, a bedroom 18, and a bathroom 20. The apartment door 22 has a contact sensor 24. Corresponding contact sensors are also on certain cabinets and a refrigerator in the kitchen 14 and a closet in bedroom 18. Using these sensors 24, the closing and opening of the apartment and cupboard doors can be monitored. There is a blood pressure monitor 26 in the kitchen 14. The blood pressure monitor 26 can be a mobile device that can be located in any room.

Motion detectors 28 are provided in all rooms. A fall sensor 30 is provided in at least one room, here in the living room 16.

In the kitchen 14, two current sensors 32 are provided which monitor the switching on of certain electrical appliances, for example a stove and an oven, or certain other kitchen appliances.

A night light 34 is provided in bedroom 18. A bathroom scale 36 is provided in the bathroom 20. In this exemplary embodiment, devices and sensors 22 - 36 can be wirelessly coupled to the control and communication unit 10, so that the information recorded by them can be read out or, in the case of the night light 34, the same can be activated by the control and communication unit 10.

The sensors shown, their positioning and their distribution to the individual rooms are purely exemplary in order to make it easier to understand the invention. Additional and/or different sensors can be provided and these can be distributed differently across the rooms. With appropriate sensors, an assistance system can be implemented that is designed in accordance with claim 1 and can also optionally have the following features.

The control and communication unit 10 can be designed to set up a local wireless communication network via which it can communicate with at least some of the sensors. This has the advantage that the sensors can be easily distributed in the living area to be monitored and placed in such a way that they optimally fulfill their respective task. In addition, defective sensors can be easily replaced and new sensors can be easily integrated into the system without having to lay cables.

The evaluation unit is preferably designed as a non-local computing unit with which the control and communication unit 10 communicates wirelessly or in particular wired. This has various advantages. In this way, the hardware effort in the living area to be monitored can be minimized, the maintenance and updating of the evaluation unit can be simplified (typically the unit is formed by components that are located in a data center) and the data transmitted to the evaluation unit via the control and communication unit can be particularly secured, especially in a so-called “loT” (internet-of-things) system. The algorithms used by the evaluation unit are typically processed in the cloud. If the control and communication unit in the monitored living area is defective or is accidentally disconnected from the power supply, no data can still be lost. In addition, the control and communication unit is preferably designed so that when connected to a The power grid starts automatically and no complex startup procedures have to be carried out. Older people who are not technically savvy sometimes disconnect the control and communication unit in order to connect another device, such as a vacuum cleaner, to the corresponding socket, and were overwhelmed with special start-up procedures.

In a preferred embodiment it is provided that the control and communication unit is designed to check whether it is connected to a local power supply network and, if not, to trigger an alarm according to a predetermined alarm chain. Since the control and communication unit, as a central element of the on-site system, should always be online, this design ensures that interruptions in the activity of the control and communication unit are noticed. For this purpose, the control and communication unit can have its own communication module and an energy storage device in order to be able to communicate with the evaluation unit if the local power network is disconnected or the power network fails.

Alternatively or additionally, it can be provided that the evaluation unit is designed to permanently or intermittently check the functionality of the control and communication unit and, if a functional anomaly is detected, to trigger an alarm in accordance with a predetermined alarm chain.

The control and communication unit is preferably designed to check the functionality of the sensors permanently or intermittently and, if a functional anomaly is detected, to trigger an alarm according to a predetermined alarm chain. The testing of the sensors can be carried out by the evaluation unit as part of remote maintenance.

Preferably, the control and communication unit is prepared to communicate with the sensors via various communication protocols, in particular via at least some of the following protocols: Wi-Fi, Bluetooth®, Zigbee®, Z-Wave, 6L0WPAN, RFID, Cellular loT protocols, NB-IOT, 5G, NFC, LoRaWAN, LTE, Infrared Data Association (IrDA) protocols, and IR remote control protocols from consumer electronics. These can Protocols must be stored in the control and communication unit or be available for retrieval in the evaluation unit. Ultimately, the only important thing is that commercially available sensors can be connected to the control and communication unit without any major installation effort so that the data they record can be forwarded to the evaluation unit.

The control and communication unit typically includes a microphone, a loudspeaker and a display, preferably a touch display, and thus allows auditory and visual communication with the person using the assistance system.

The sensors used preferably include sensors from at least two of the following types of sensors: motion detector, presence sensor, fall sensor, temperature sensor, door opening sensor, window opening sensor, current flow sensor to certain electrical devices, weight sensor, blood pressure sensor, infrared sensor, ultrasonic sensor, LIDAR sensor, blood sugar sensor, medication box sensor , refrigerator sensor, toothbrush sensor, smoke detector, CO2 sensor. It should be noted that in an assistance system according to the invention, “dumb” sensors become “smart” because they suddenly have logic. For example, an electricity meter on the television can show that the person using the system is no longer keeping to their usual rhythm (e.g. news broadcast at 8 p.m.). The ammeter thus receives logic indirectly, namely via the evaluation unit, and in this sense becomes smart.

The evaluation unit is preferably designed to include non-sensor-based data from the database when determining the anomaly and, when an anomaly is detected, to output any changes in the non-sensor-based data as the context of the anomaly. For example, information about a current medication can be stored in the database as non-sensor-based data. If an anomaly such as a change in sleep behavior is detected, the system can inform the person or unit to be alerted according to the respective alert chain that the anomaly occurred at the same time or shortly after a change in medication. The evaluation unit is preferably designed to use self-learning algorithms for evaluation. The system develops automatically and adapts to the new circumstances of the person using the system

In a further preferred embodiment, the control and communication unit is designed to be coupled with smart home control devices. This advantageously allows the system to take on activities far beyond anomaly detection and the person using the system does not have to be familiar with different control units.

In a further preferred embodiment, the control and the system comprises at least one mobile communication device that can be connected to the evaluation unit and / or the control and communication unit and via which a user of the mobile communication device is released information about the person and / or the living area can be displayed. The mobile communication device can be part of at least one of the at least two alerting chains. The communication device can be a standard smartphone and the user can be, for example, a close relative of the person using the system. An adult child can check at any time whether, for example, his mother who needs care is doing well. The mobile communication device can advantageously also be designed to control the smart home control devices after appropriate approval by the person using the assistance system. This son could, for example, control the heating in the living area when he brings his mother back from a visit so that the living area has the desired temperature when he arrives.

The control and communication device is preferably designed to communicate with the person according to predefined or variable criteria. The variable criteria can be criteria learned from the person's previous behavior and/or condition. If necessary and approved, communication can be established with the person without the person having to actively request this. For example, certain emotional anomalies can be prevented or remedied. If, for example, it is determined that the person using the system feels lonely, Kl-supported conversations can be initiated automatically, the keeping of a gratitude diary (spoken and, if necessary, automatically converted into writing) can be initiated and the person be motivated to remember the shameful things in life. Filtering in the form of a so-called sentiment analysis ensures that only positive thoughts are included. Activating morning and evening conversations can be held with the support of a class and the morning and evening can be made positive with harmful sayings. It is also possible to encourage the person to play memory-promoting games that train the brain.

The control and communication unit is preferably also designed to be used as a telephone and/or video telephone.

The assistance system is advantageously designed in such a way that the person can communicate with the system in natural language. This can be done using appropriate computer-based so-called chatbots, whereby the assistance system advantageously allows available chatbots to be used without any special reprogramming.

If the assistance system communicates with the person automatically, the communication can in particular concern one topic from the following list of topics: queries about detected anomalies, reminders of certain appointments and actions, rehabilitation measures, preventive measures, positive psychological intervention, ordering errands. However, this list of topics is far from exhaustive. For example, religious or food-related topics can also be discussed in a classroom-based manner. Reminders of appointments or actions, e.g. eating, drinking, taking medication, can also be issued verbally and appear on the screen of the control and communication unit at the same time. By confirming, you can mark the reminder as completed.

Preferably, the evaluation unit is designed to draw conclusions from current and historical sensor data about at least some of the following states and abilities of the person using the assistance system: state of health, emotional state, cognitive abilities, ability to live independently.

The assistance system according to the invention allows information about the condition and behavior of the person using the system to form an overall picture to evaluate. The information can be shared with other stakeholders such as relatives, doctors or nursing services.

Since conversations are often difficult for people who can no longer hear well, the control and communication unit can have extra powerful speakers and display all dialogues on the screen in clearly legible text.

Thanks to the language assistant and the internet connection, the resident can easily regain access to so-called “digital participation”, even if the resident can no longer manage the transition with IT. Through generative AI, the control and communication unit can conduct natural language dialogues and thus provide information on topics such as nutrition or other advice. A biographical work based on an anamnesis form is also possible. The system gives advice on how to behave. But memories can also be processed. There are several modes of communicating with the control and communication unit. The system makes a significant contribution to being able to manage everyday life independently and safely despite certain health restrictions, but at the same time to be able to quickly contact relatives, carers, etc. in emergency situations. This means that fun and interaction do not have to be neglected. Cognitive disorders can be identified from long-term recording of the behavior of the person using the system. If a person gets up more and more often at night and walks around the apartment, they are probably having problems structuring their day, which can indicate a possible dementia or Parkinson's disease. An integrative approach is a special feature of the system. It is a holistic solution that combines information about activity and behavior profiles to raise the alarm at the right moment or signal the need for support. For example, there are also requests or instructions to take medication at the right time, turn off the stove or have something to drink. The idea of prevention is in the foreground here. The system not only allows acute danger prevention, but also observation of the course of a person's behavior and/or condition. Behavioral changes are carried out in long-term observation in order to document the status of the person. Dialogues recorded with consent can be seen in long-term observation how the language changes. Based on the status analysis, an expert can then assess whether there is a change that needs to be checked by the specialist. The system takes a holistic approach and supports you in managing everyday life.

Information about health status can be presented as a dashboard and/or as a traffic light system. The person using the system can decide who sees which data. For example, many people do not want their children to see what they do in detail (e.g. number of trips to the toilet).

The assistance system can support both prevention and rehabilitation, e.g. after strokes, for example with activating games and language exercises.

Fig. 2 shows a system with one of several possible configurations according to the invention. An evaluation unit 10 is connected via a telecommunications network 20 or WAN (Wide Area Network) to smartphones as alarm receiving devices 30 and body-contactless fall sensors 40 and a control and communication unit 50. In Fig.

2 to simplify the illustration, only one of the alarm receiving devices and only one of the fall sensors. It goes without saying that the evaluation unit 10 is designed to interact with a large number of control and communication units in different living areas.

The control and communication unit 50 has an audio input and audio output device 52 with loudspeaker and microphone, a processor 51 in conjunction with a known memory unit 56 for reading in and out of data and commands, and a touchscreen 54 for further inputs and / or outputs via display.

The evaluation unit 10 includes a database 60 in which home addresses of the control and communication unit 50 and at least one of the fall sensors 40 are assigned for a link of each of the fall sensor outputs and the audio outputs and inputs of the control and communication unit with one of the home addresses, a loT hub 7 with APIs for the fall sensor outputs 70, only one of which is shown in Fig. 2, a speech processor unit 80 for converting a stored text sequence into audio output and the audio input into another stored text sequence for voice control, one the APIs for The alarm unit 82 uses the fall sensor outputs 70 and implements the following functions:

Receiving the fall sensor outputs using the fall sensor output APIs 70,

Initiating a voice dialog on the control and communication unit 50 according to the alarm chain and, if necessary, sending a call for help with indication of the residential address according to the link, namely if at least one condition is present, including the condition that a received one of the fall sensor outputs signals an alarm, in particular a fall alarm,

Canceling the alarm when a refusal to help is detected, sending the call for help to the alarm receiving device 3 according to the scheduling.

In the embodiment shown, the fall sensor 40 is part of a fall detection system set up with the network 20 and the fall sensor outputs are evaluations of, for example, infrared recordings from a sensor in the apartment. The API 70 is then designed as a gateway for receiving the evaluations by the evaluation unit.

The evaluation unit 10 also includes a helper database 62, in which a list of helper contacts with a ranking and helper availability times is assigned to each home address. The evaluation unit 10 is configured with a helper database module 94 to allow authorized form-based remote access to the helper database 62 and thus to serve as a portal for authorization-dependent access to the helper database 62 for deleting, changing, reading and / or adding. Helper availability periods can also be managed and, in particular, entered.

The function of sending the call for help is set up to create a selection of the helper contacts of the list, whose assigned helper availability times include, to send the call for help and request a confirmation to one of the alarm receiving devices 30 corresponding to the highest-ranking helper contact of the selection according to the ranking, to check the receipt of the confirmation and the number of helper contacts in the Selection and to update the selection if the number is greater than 1 and in the absence of confirmation for a predetermined period of time, namely by removing an entry corresponding to the highest-ranking helper contact, to repeat the sending of the call for help and the request for confirmation.

With a wireless connection to the control and communication unit 50, further sensors 42, 44 are connected to the processor of the control and communication unit for sending further sensor outputs, as are the audio input and audio output device 52. In the database, the link is also a link of the further sensor outputs to one of the residential addresses. The evaluation unit 10 also includes APIs for the further sensor outputs 72, only one of which is shown in Fig. 2.

The evaluation unit 10 also includes an analysis unit 84 implementing the following functions:

Creating profiles that are specific to addresses among the residential addresses, based on the link and with the other sensor outputs and, if applicable, the fall sensor output, and

Detecting and classifying changes in profiles over time.

The further sensors 42, 44 are sensors for monitoring the streaking of doors and/or the switching on of power consumers such as a stove.

The evaluation unit 10 also includes a classification adaptation module 96 in order to be able to serve as a portal for authorization-dependent access to the classifications for reading and / or to factors of the classifications for reading, changing, deleting or supplementing through authorized form-based remote access. The function of triggering a voice dialogue and providing a call for help is advantageously a function that requires the existence of the condition that the classification signals danger.

The evaluation unit 10 includes a machine learning unit 86 that implements the following functions:

Collecting feedback information by recording for a set of addresses under the home addresses of the deletion or adding factors of the classifications,

Updating the classification factors for the set of addresses, repeating the collection of feedback information and updating the classification factors with iteration of the classification factors to minimize an amount of detected de-alerting and/or deletion or adding to the classification factors.

The evaluation unit 10 also includes an appointment reminder unit 88 for appointment reminders via audio output. The server unit 1 also includes a voice input unit 90 with a function, for example to complete the profiles via voice input. The evaluation unit 10 is also set up as a portal for form-based remote settings of queries for voice input in order to supplement the profiles or, via a classification module, to be able to adapt 96 classifications, in particular associated factors or weightings, via a portal in a form-based and authorization-dependent manner. The evaluation unit 10 is also set up with the appointment reminder unit 88 as a portal for setting appointment reminders in an authorization-dependent manner. For example, voice queries for blood pressure readings can be set up or reminders about food and/or fluid intake.

In the control and communication unit 50, the processor 51 is connected to a memory such as a flash memory 56 with the required operating system, drivers and applications. In this exemplary embodiment, the evaluation unit 10 includes a chatbot 92 for interaction with the audio output and input unit 52, in particular for voice games.

The control and communication unit 50 also includes a touchscreen 54 as a display (not shown).

Fig. 3 is a process diagram that is consistent with a process according to the invention. The process diagram is divided by vertical lines into processes in an apartment shown on the left, processes of interaction with an assistant outside the apartment shown in the middle and processes in the evaluation unit 10 shown in Fig. 1 shown on the right.

The method shown in Fig. 3 is a method for operating the network of hardware units shown in Fig. 2, wherein data connections are provided from the fall sensor 40 to the evaluation unit 10, between the control and communication unit 50 and the evaluation unit 10 and from the same to the alarm receiving device 30. The following steps are carried out using the alarm unit 82:

51 Receive the fall sensor outputs using the fall sensor output APIs 70,

52 Triggering a voice dialog on the control and communication unit 50 according to the link and scheduling a sending of a call for help with information about one of the residential addresses according to the link, in the presence of at least one condition, comprising the condition that a received one of the fall sensor outputs signals a fall ,

53 Execution of the voice dialogue on the control and communication unit 50 when accessing a voice processor unit 80,

S13 Cancellation of termination by voice control when the further text sequence is recognized as a rejection of help,

S6 Sending the call for help to the alarm receiving device 3 according to the scheduling. A fall sensor 40 provides a fall sensor output in step S1 signaling a fall event. For fall sensor output, a system can also be used in which a sensor unit in the apartment is connected to a cloud-based fall sensor platform, which ultimately provides the fall sensor output in step S1.

To receive the fall sensor output, an API 70 is provided in the evaluation unit, which allows the evaluation unit, based on the fall sensor output, to continue with the S2 sending of a voice dialog on the control and communication unit 50, which is in the same apartment as the fall sensor 40 or its Sensor unit is located. The database 60 ensures via a link that the voice dialogue is sent out exactly on this control and communication unit 50, which is located in this apartment. This occurs under the condition that a received fall sensor output signals a fall. The voice dialogue is then executed in step S3 on this control and communication unit 50. For this purpose, audio streaming is set up between control and communication unit 50 and evaluation unit 10. The speech dialogue is therefore carried out in step S3 through access by the control and communication unit 50 to this speech processor unit 80. The speech processor unit 80 can also be one that cooperates with a cloud-based speech recognition service. The voice dialogue is carried out in step S3 on this control and communication unit 50 in order to be able to provide feedback from a person in the apartment to the evaluation unit 10. An alarm confirmation is received by the evaluation unit 10 in step S4 both when there is a request for help as a response and when a predetermined period of time has elapsed after step S3 has been triggered. As described above, the evaluation unit 10 is configured with a helper database 62 in which a list of helper contacts with a ranking and helper availability times is assigned to each home address. When the alarm confirmation is received, a selection of the currently available helper contacts from the list is created in step S4.

In step S6, the call for help is then sent and a confirmation is requested to one of the alarm receiving devices corresponding to the highest-ranking helper contact of the selection according to the ranking. For this purpose, an email and an SMS are sent to an assistant's smartphone and received there in step S7. According to the process described above, the sending of a call for help is scheduled by specifying one of the residential addresses according to the link, in the presence of at least one condition, comprising the condition that a received fall sensor output signals a fall. Advantageously, the scheduling can also be brought forward when the further stored text sequence is recognized as a request for help.

Using a browser-based application on an assistant's smartphone, the assistant sends in step S8 a confirmation that sends step S10 in the evaluation unit 10 or a rejection that sends step S9 in the evaluation unit 10.

Alternatively, via the process flow of step S7 that bypasses step 8, a rejection that triggers step S9 in the evaluation unit 1 is automatically carried out after a predetermined period of time has elapsed after step S6, in which step S10 was not triggered.

Alternatively, running the S3 voice dialog with voice control allows the voice control to cancel the scheduling if help rejection is detected. The scheduling of sending a call for help is canceled by voice control in step S13, if necessary even after it has been sent. A false alarm can be marked as a false alarm or uncritical. In addition, it is provided that by entering in step S12 on a smartphone in step S13, the scheduling of sending a call for help can be canceled, if necessary, even after sending. Scheduling means determining that and from what point in time alert messages are to be sent in accordance with step S6.

In addition, it is provided that by entering S14 on the touchscreen 54 provided on the control and communication unit 50 or, if necessary, alternatively a button provided on the control and communication unit, or by voice control in step S15, the sending of a call for help can also be scheduled, if necessary is canceled after sending. In step S5 it is checked whether the termination has been canceled. If the termination has not been canceled, step S5 checks whether the number of helper contacts in the selection is greater than 1. In step S5, the selection is updated if the number of growers is greater than 1 and if there is no confirmation signaled in step S9 for a predetermined period of time. The update is carried out by removing an entry corresponding to the highest-ranking helper contact, which update is then followed by step S6 in order to alert a next helper, whereas otherwise if the number is 1, step S6 is bypassed and the already alerted helper is alerted again.

Functions or steps are therefore implemented in the evaluation unit for checking receipt of the confirmation and the number of helper contacts in the selection and for updating the selection if the number is greater than 1 and if there is no confirmation over a predetermined period of time, namely by removing an entry corresponding to the highest-ranking helper contact, repeating the sending of the call for help and requesting confirmation.

A confirmation in step S8 also triggers the display S11 of a message about the confirmation from the assistant on the touchscreen 54 of the control and communication unit 50. In step S10, the scheduling of sending a call for help is canceled and the helper is further requested to contact us via a message on the smartphone. Due to the cancellation of the scheduling of sending a call for help, it is ensured that no further alert sending can take place with step S6 after step S5. A function or a procedural part can also be provided in order to ensure that after step S5 no further alerting can be sent with step S6 due to the cancellation of the scheduling of sending a call for help, a message is sent to all helpers for whom step S6 was carried out and from whom no confirmation via step S10 was received. These assistants can then be informed that an alarm has been dealt with. The consultation takes place via a voice call between the smartphone and the control and communication unit 50, which is also provided with a voice call and telephone function set up for this purpose. The assistant then decides and either triggers step S13 with their smartphone or comes into the apartment. After assistance is provided, input S14 is then made on a touchscreen 54 provided on the control and communication unit Or, if necessary, alternatively a button provided on the terminal device Or through the voice control, through which input in step S15 the sending of a call for help is scheduled (possibly also after the Sending a call for help) is canceled.

The server device, in particular the portal mentioned and/or the relevant form-based entries via the units or modules 88, 94, can be set up according to one or more of the following features or functions:

Edit user profile incl. two-factor authentication, set appointments incl. voice reminder, set medication incl. voice reminder, view medication plan, set drinking reminder incl. voice reminder or drinking noise, data acquisition and evaluation in particular of the sensor outputs, manage sensors, in particular via classification adjustment module 96, Observe vital signs based on sensor outputs, manage authorizations for relatives and nursing services, event analysis based on sensor outputs, alarm processing, user management, alarm chain management, device management, remote access to control and communication units, especially via teleport,

Remote software update on control and communication units, data migration. For this purpose, times recorded for each visible sensor output, in particular fall sensor output, can also be recorded by the evaluation unit 10, stored and made available in the portal.

The following APIs can be set up in the evaluation unit 10 APIs for voice interfaces, in particular for economy control, such as

API Twilio SMS, WhatsApp and phone calls,

API Vayyar fall sensor,

API Withings Or Nokia.

The following techniques, features and products can be used to implement the evaluation unit 10:

Domain driven design, distributed streaming solution with Kafka including data replication, data aggregation, partitioning, load balancing and resiliency,

Identity management with Keycloak including 2FA, OpenlD Connect, OAuth2, authentication and authorization,

Microservice architecture with Kubernetes and Docker, secured connection between end device and cloud via Azure loT hub, automatic data synchronization between portal and end device, device simulator, data analysis I AEDLs, Quartz system scheduler.

Step S11 described above preferably shows transparently how an alarm chain is processed and who will take care of this case.

For the sensor outputs, interfaces can be implemented via Bluetooth to a scale, a blood pressure monitor and a blood glucose monitor. Vital data can be transmitted completely independently of the device via the audio input and audio output device and with the help of the speech processor unit 80. This promotes the participation effect, personal responsibility and offers the possibility of continuing to use familiar devices. Example texts for the alarms according to step S6 described above are:

Fall: "Fall alarm at Mrs. Maria Mustermann, Exampleweg 14, 12345 Somewhere, Time: 19:20 12.08.2020, Tel. +49 1234 567891011 Or +49 9876 543210. Are you concerned? 1=Yes, 9=No"

Dangerous situation: "Conspicuously long inactivity for Ms. Maria Mustermann, Exemplarweg 14, 12345 Somewhere, time: 7:20 p.m. August 12th, 2020, Tel. +49 1234 567891011 Or +49 9876 543210. Are you worried? 1=Yes, 9=No"

Vital signs: "Unusually high blood pressure 167/102 in Ms. Maria Mustermann, Exemplarweg 14, 12345 Somewhere, time: 7:20 p.m. August 12, 2020, Tel. +49 1234 567891011 Or +49 9876 543210. Are you worried? 1=Yes, 9 =No"

Advantageously, step S6 can be used to alert another application via API. The German emergency service Maltester uses its own system to manage incoming emergency calls. This can then be addressed via the configurable API, and an API for reporting alarms can then also be provided.

In addition to the touchscreen 54, the evaluation unit 10 can also advantageously control outputs on smart home devices such as lighting devices via the control and communication unit 50, also in response to some of the sensor outputs and / or to voice inputs into the audio input and output unit 52 and, if necessary. also in accordance with settings that are offered with the evaluation unit 10 via the portal in a form-based and authorization-dependent manner.

Claims

PATENT CLAIMS

1. Multifunctional assistance system, in particular for people living alone, comprising a number of sensors that are distributed in a living area of a person using the assistance system for monitoring predefined activities and predefined parameters, a control and communication unit provided in the living area for communicating with the sensors and the person, a database with non-sensor-based information about the person and at least two different alerting chains and an evaluation unit for evaluating sensor data transmitted via the control and communication unit, wherein: the evaluation unit can access the database and is designed to do so to detect any anomalies in the behavior and/or condition of the person from the sensor data, to divide the anomalies into at least two categories and, based on the categorization, to trigger an alarm according to at least one of the at least two alarm chains.

2. Assistance system according to claim 1, characterized in that the control and communication unit is designed to set up a local wireless communication network via which it can communicate with at least some of the sensors.

3. Assistance system according to claim 1 or 2, characterized in that the evaluation unit is designed as a non-local computing unit with which the control and communication unit communicates wirelessly or wired.

4. Assistance system according to one of claims 1 - 3, characterized in that the control and communication unit is designed to check whether it is connected to a local power supply network and, if not, to trigger an alarm according to a predetermined alarm chain.

5. Assistance system according to one of claims 1 - 4, characterized in that the control and communication unit is prepared to communicate with the sensors via various communication protocols, in particular via at least some of the following protocols: Wi-Fi, Bluetooth®, Zigbee® , Z-Wave, 6L0WPAN, RFID, Cellular loT protocols, NB-IOT, 5G, NFC, LoRaWAN, LTE, Infrared Data Association (IrDA) protocols, and IR remote control protocols from consumer electronics.

6. Assistance system according to 1 - 5, characterized in that the control and communication unit comprises a microphone, a loudspeaker and a display, preferably a touch display.

7. Assistance system according to one of claims 1 - 6, characterized in that the sensors include sensors from at least two of the following types of sensors: motion detector, presence sensor, fall sensor, temperature sensor, door opening sensor, window opening sensor, current flow sensor to certain electrical devices, weight sensor, infrared Sensor, ultrasonic sensor, LIDAR sensor, blood pressure sensor, blood sugar sensor, medicine box sensor, refrigerator sensor, toothbrush sensor, smoke detector, CO2 sensor.

8. Assistance system according to one of claims 1 - 7, characterized in that the evaluation unit is designed to include non-sensor-based data from the database when determining the anomaly and, when an anomaly is detected, to output any changes in the non-sensor-based data as the context of the anomaly.

9. Assistance system according to one of claims 1-8, characterized in that the evaluation unit is designed to use self-learning algorithms for evaluation.

10. Assistance system according to one of claims 1 - 9, characterized in that the control and communication unit is designed for coupling with smart home control devices.

11. Assistance system according to one of claims 1 - 10, characterized in that it comprises at least one mobile communication device that can be connected to the evaluation unit and / or the control and communication unit and via which a user of the mobile communication device can receive information about the Person and/or the living area can be displayed.

12. Assistance system according to claim 11, characterized in that the mobile communication device is part of at least one of the at least two alarm chains.

13. Assistance system according to claim 10 and one of claims 11 or 12, characterized in that the mobile communication device is designed to control the smart home control devices after appropriate approval by the person using the assistance system.

14. Assistance system according to one of claims 1 - 13, characterized in that the control and communication device is designed to communicate with the person according to predefined or variable criteria.

15. Assistance system according to one of claims 1 - 14, characterized in that the system is designed so that the person can communicate with the system in natural language.

16. Assistance system according to one of claims 14 or 15, characterized in that the communication with the person concerns at least one topic from the following list of topics: queries about identified anomalies, reminders of certain appointments and actions, rehabilitation measures, preventive measures, positive psychological intervention , ordering errands.

17. Assistance system according to one of claims 14 - 16, characterized in that the variable criteria are criteria learned from the person's previous behavior and / or condition.

18. Assistance system according to one of claims 1 - 17, characterized in that the control and communication unit has its own communication module and an energy storage device in order to be able to communicate with the evaluation unit if the local power network is disconnected or the power network fails.

19. Assistance system according to one of claims 1 - 18, characterized in that the evaluation unit is designed to check the functionality of the control and communication unit permanently or intermittently and, if a functional anomaly is detected, to trigger an alarm in accordance with a predetermined alarm chain.

20. Assistance system according to one of claims 1 - 19, characterized in that the control and communication unit is designed to check the functionality of the sensors permanently or intermittently and, if a functional anomaly is detected, to trigger an alarm in accordance with a predetermined alarm chain.

21. Assistance system according to one of claims 1 - 20, characterized in that the evaluation unit is designed to draw conclusions from current and historical sensor data about at least some of the following states and abilities of the person using the assistance system: state of health, emotional state, cognitive abilities , ability to live independently.

22. Assistance system according to one of claims 1 - 21, characterized in that the control and communication unit is designed to be used as a telephone and / or video telephone.

23. Assistance system according to one of claims 1 -22, characterized in that the at least one alarm chain is a continuously updated alarm chain.

24. Control and communication unit for an assistance system according to one of claims 1 - 23.