CN116033865A - System and method for identifying disease-specific odors in passenger vehicles and passenger vehicle - Google Patents

Abstract

The invention relates to a system (10) for identifying a disease-specific smell in a passenger vehicle (1), the system (10) comprising: a plurality of first sensors (11) that detect substances that generate odors; a processing unit (ECU) that obtains data obtained by detecting a substance that generates an odor from the first sensor (11), identifies the odor specific to the disease by machine learning, and generates an indication signal depending on the identification result, wherein the processing unit (ECU) implements a machine learning algorithm that obtains the data of the first sensor (11) and learns or trains to identify a biomarker for the disease in the data; and an interface (12) which provides an indication signal to a passenger (P), a driver and/or an external location ("cloud") of the passenger vehicle (1).

Description

System and method for identifying disease-specific odors in passenger vehicles and passenger vehicle

Technical Field

The present invention relates to a system and method for identifying disease-specific odors in passenger vehicles, and to a passenger vehicle.

Background

Odor sensors are generally known from the prior art. An odor detection system is disclosed, for example, in EP 3 187,852 A1. An odor sensor based on an adsorption mode is disclosed in EP 3 379,240 A1. Furthermore, alcohol detectors in vehicles are known from the prior art, see for example the publication https:// www.dadss.org/, from the research program Driver Alcohol Detection System for Safety (for a safe driver alcohol detection system). Furthermore, artificial neural networks known from the prior art have learned to determine the odor associated with an odor molecule from the structure of the odor molecule, see for example "Machine Learning for Scent: learning Generalizable Perceptual Representations of Small Molecules" by B.Sanchez-Lenelling et al (machine learning for odors: learning a generalized perceived characterization of small molecules) ", arXiv:1910.10685v2[ stat.ML ].

In order to avoid the spread of diseases, in particular to manage infectious diseases or pandemics, it is desirable to identify the pathogen as quickly as possible. In near and/or long haul traffic, infectious diseases or pandemics often cause passengers to lose confidence in passenger vehicles due to the person's desire to prevent infection.

Disclosure of Invention

The task of the invention is to be able to quickly identify diseases and to avoid the spread of diseases in passenger vehicles.

The system, method and passenger vehicle of the independent claims and the parallel claims, respectively, solve this task.

According to one aspect, the present invention provides a system for identifying a disease-specific smell in a passenger vehicle. The system comprises:

a plurality of first sensors, the first sensors detecting substances that generate odors;

a processing unit that obtains data obtained by detecting a substance that produces an odor from the first sensor, identifies the odor that is specific to the disease by machine learning, and generates an indication signal depending on the identification result, wherein the processing unit implements a machine learning algorithm that obtains the data of the first sensor and learns or trains to identify biomarkers for the disease in the data, and

an interface providing an indication signal to a passenger, driver and/or external location of the passenger vehicle.

According to a further aspect, the present invention provides a method for identifying a disease-specific smell in a passenger vehicle. The method comprises the following steps:

detecting odor-causing substances in ambient air including the respiratory air and/or body odor of the passenger;

identifying a disease-specific odor;

informing passengers, drivers and/or operators and/or authorities of the passenger means of the information about one or more identified disease-specific odours; and is also provided with

Depending on the identified disease-specific odour, measures for avoiding the transmission of specific diseases including quarantine measures are initiated.

The system according to the invention is used to perform the method.

According to a further aspect, the invention provides a passenger vehicle comprising a system according to the invention, wherein,

the first sensor of the system is arranged in a seating area, standing area, ventilation and/or air conditioning of the passenger vehicle and

the interface of the system is an interface to a display unit in the passenger vehicle and/or a radio interface to the cloud, the operator of the passenger vehicle and/or to the administrative authorities.

Advantageous embodiments of the invention result from the definition, the dependent claims, the drawing and the description of the preferred embodiments.

The odor-causing substances include odor molecules and odor molecule groups. For example, the odor molecule (R) - (+) -limonene is the predominant odor substance of lemon. Decanal is an odor-generating substance having the characteristic odor of orange peel. Nonanal is an odoriferous substance with citrus odor. Ethyl butyrate is an odor-generating substance having the characteristic odor of pineapple. The odor-causing substances are contained in the ambient air of the passenger vehicle. Ambient air includes respiratory odors and additional odors from other body passages of the living being, i.e., body odors, such as percutaneous body emissions. Diseases in the passenger vehicle can thus be quickly identified in the ambient air. Thus, taking a breath sample by means of a balloon or a small blow tube and performing a complex, time-consuming and expensive evaluation of the breath sample in a gas chromatograph-mass spectrometer to determine the molecules in the breath sample is advantageously replaced by the detection of the sensing of the odor-causing substance according to the invention and the evaluation by means of a machine learning method.

For example, odor-causing substances are detected by means of the identification of odor molecules. These identifications include absorption lines in the absorption spectrum, scattered light specific to the respective molecule, such as raman or rayleigh scattering, adsorption characteristics, and characteristic vibration patterns based on molecular mass or characteristic curves in impedance spectroscopy. The vibration pattern is detected, for example, using an acceleration sensor, such as a microelectromechanical system.

Diseases include viral infections and bacterial diseases. For example, one viral disease is COVID-19 caused by the coronavirus SARS-Cov-2. Biomarkers can be objectively measured and can be indicative of normal biological processes or characteristic biological features of a disease process within the body. For example, biomarkers are contained in the breathing air and/or body taste.

The advantage of identifying disease-specific odors in ambient air, such as in respiratory air and/or body odor, by means of a machine learning algorithm is that no time consuming and uncomfortable chemical methods, blow tubes or surface contact methods need to be used and thus diseases can be identified quickly. The breathing air and other body flavors are constantly emitted and evaluated by the processing unit.

The processing unit includes a computing unit and exchanges data with the memory. For example, the processing unit includes an internal memory. The machine learning algorithm is stored or loaded in memory. Furthermore, the processing unit writes the identified disease-specific odors into a memory from which they can be read. According to one aspect of the invention, data of the sensor is stored in the memory.

Machine learning is a technique that teaches computers and other data processing devices to perform tasks by learning from data, rather than programming for tasks. Machine learning algorithms learn odors from data by classification and/or regression methods. During learning or training, data is processed. The trained machine learning algorithm then identifies odors in the input data without having to perform a comparison with odors stored in memory or a database. The machine learning algorithm includes, for example, an artificial neural network, such as a convolutional network, or a decision tree algorithm, a random forest classifier, or a support vector machine classifier.

According to one aspect of the invention, the processing unit comprises at least one graphics processing unit for implementing processes in parallel, in particular for accelerating the implementation of machine learning algorithms. According to one aspect of the invention, the processing unit is a system on a chip. According to a further aspect of the invention, in particular in combination with the system on chip, the processing unit comprises a neural or neuromorphic circuit which mimics the mammalian olfactory system in hardware and implements a machine learning algorithm thereon.

According to a further aspect of the invention, the processing unit is integrated or can be integrated into the on-board electrical system of the passenger vehicle. According to one aspect of the invention, the sensor is integrated into the on-board electrical system. Thus, communication between the sensor and the processing unit is facilitated. The on-board electrical system is, for example, a CAN bus. According to one aspect of the invention, at least a part of the communication between the sensor, the processing unit and the interface is performed wirelessly, e.g. by means of bluetooth low energy or WLAN.

The indication signal includes a visual signal, an audible signal, and a signal that can be perceived tactilely. According to one aspect of the invention, the indication signal is visually shown on the display unit.

Passengers include humans and animals, for example, pets such as dogs.

The function of the system according to the invention works independently of the integration into the passenger vehicle. Existing passenger vehicles can be retrofitted with the system.

In one embodiment of the invention, the first sensor detects odor-causing substances in ambient air, including the respiratory air and/or body odor of the passenger, by surface adsorption, light interaction and/or frequency and/or amplitude of vibration.

According to one aspect of the invention, the first sensor comprises a membrane that adsorbs odor-causing substances contained in the breathing air and/or body taste of the passenger. For example, the membrane comprises a conductive polymer. For example, the first sensor measures an electrical property of the membrane, such as electrical conductivity, after adsorption of the odor-causing substance, wherein adsorption of the unique odor-causing substance characteristically alters the electrical property of the membrane. According to a further aspect of the invention, the first sensor comprises a plurality of membranes, which each adsorb only a specific substance. Thus, different odor molecules can be identified.

Light interactions include light scattering, back scattering, reflection, transmission, diffraction, and refraction. According to one aspect of the invention, the identification for the respective molecule is characteristic of scattered light, e.g. scattered light from back scattering. In addition to exciting the electronic transitions, molecular vibrations of the odor-causing substance are also excited, including stretching vibrations, rocking vibrations, shearing vibrations, torsional vibrations and/or seesaw vibrations. The stretching vibration is vibration along the bond axes of two atoms in a molecule. Rocking, shearing, torsional and teetering vibrations are deformation vibrations that deform the bonding angle. Molecular vibrational energy is evidenced by the absorption and emission of infrared radiation. Molecules can be identified via their molecular vibrations.

In one embodiment of the system, the at least one first sensor is a tunable lidar sensor. Lidar sensors are implemented to detect odor molecules in dependence upon light backscattered from the vehicle interior space. For example, the lidar sensor emits laser pulses having a wavelength in the infrared range that is harmless to the human body. Tunable means that the lidar sensor is implemented to send out pulses of light of a plurality of different wavelengths in order to identify different odour molecules from the backscattered light. According to one aspect of the invention, the lidar sensor includes drive electronics to transmit different wavelengths. For example, a first wavelength and a second wavelength are transmitted for different wavelengths, respectively, in order to calculate the concentration profile of the odor-causing substance by means of differential absorption. According to a further aspect of the invention, the lidar sensor comprises integrated evaluation electronics for spectroscopy, such as raman spectroscopy. In a further embodiment of the system, the lidar sensor comprises a quality switch. The light pulses are made shorter by means of a quality switch, also called Q-switch. Thus, a high peak power is achieved even at relatively low energies. Due to the low energy it is achieved that: lidar sensors are not harmful to passengers and drivers. Due to the high peak power, the odor molecules become soluble. For example, the quality switch is an electro-optic modulator.

In a further embodiment of the invention, the first sensor comprises a quartz vibration sensor, which determines the frequency and/or amplitude of the vibration of the odor-causing substance.

According to a further aspect of the invention, the first sensor comprises a combination of the diaphragm, lidar sensor and quartz vibration sensor described above.

According to a further aspect of the invention, the system comprises a second sensor measuring the temperature and/or a third sensor measuring the air humidity, wherein the processing unit fuses the data of the first sensor and the data of the second and/or third sensor. By measuring the temperature and/or the air humidity, a disease-specific smell is identified according to the situation. For example, in a first combination of temperature values and/or air humidity values, the concentration of the odor-causing substance is of concern, but in a second combination is not. For example, the temperature and/or the air humidity in the passenger vehicle changes during seasonal operation of the passenger vehicle or as a result of the tempering of the passenger vehicle interior.

According to a further aspect of the invention, the system comprises a fourth sensor measuring the pressure. Advantageously, the processing unit combines the measurements of the first, second, third and/or fourth sensor and/or validates the reliability of these measurements, for example in order to minimize interference effects.

According to a further aspect of the invention, the processing unit performs error correction, in particular based on data obtained by a plurality of the first, second, third and/or fourth sensors, preferably based on data of all sensors. Thus improving the accuracy of the measurement. Furthermore, the identification of disease-specific odors can thus be achieved and controlled within tolerance ranges, for example within concentration ranges.

According to a further aspect of the invention, a machine learning algorithm is learned or trained to identify volatile organic compounds, also known as volatile organic compound (VOC for short). Among the VOCs are, for example: acetone, ethanol, isoprene, nonanal, decanal, alpha-pinene, ethyl butyrate and butyraldehyde, acetaldehyde, propionaldehyde and n-propyl acetate. VOCs are formed in organisms in disease-induced protein changes, cellular changes, or metabolic changes. For example, characteristic protein changes and metabolic changes have been observed in SARS-Cov-2 infected individuals. VOCs excreted via respiration and/or body odor may be a characterization of the disease and thus act as biomarkers.

For example, haoxuan Chen et al, "Breath-borne VOC Biomarkers for COVID-19 (respiratory VOC biomarker for COVID-19)", https:// doi.org/10.1101/2020.06.21.20136523, indicate that ethyl butyrate may be a biomarker for COVID-19. It also shows that butyraldehyde may be a biomarker negative for SARS-Cov-2. According to one aspect of the invention, the machine learning algorithm is trained or trained to identify ethyl butyrate and butyraldehyde in the data of the first sensor, for example via adsorption and/or absorption signals specific to ethyl butyrate and butyraldehyde, respectively.

Selina Traxler et al, "Volatile scents of influenza Aand S.pyogenes (co-) infected cells (volatile odors of influenza A and suppurative bacteria (co) infected cells)", https:// www.nature.com/arotics/s 41598-019-55334-0, for example, indicate that n-propyl acetate is a VOC biomarker for influenza A virus, while the increased emissions of acetaldehyde and propionaldehyde are VOC biomarkers for suppurative bacteria. According to one aspect of the invention, the machine learning algorithm is trained or trained to identify n-propyl acetate, acetaldehyde and propionaldehyde in the data of the first sensor, particularly in combination with identifying ethyl butyrate and butyraldehyde, for example by means of adsorption signals and/or absorption signals specific to n-propyl acetate, acetaldehyde and propionaldehyde, respectively. Thus, according to the invention, diseases such as influenza and/or viral bacterial co-infections can additionally be identified.

According to a further aspect of the invention, a system is used for diagnosing a disease, comprising: including influenza and covd-19. This enables a rapid identification of diseases, such as covd-19 in passenger vehicles, and a rapid initiation of appropriate measures to suppress spread of e.g. covd-19, e.g. by identifying ethyl butyrate and butyraldehyde according to the present invention. Thus, the system may increase passenger confidence in the use of passenger vehicles, especially during periods of pandemic such as covd-19.

In one embodiment of the invention, the passenger and/or the driver is informed about the identified disease-specific smell via a display unit in the passenger vehicle. The display unit comprises a separate display or a display integrated into the head unit of the passenger vehicle. The interface of the system is for example an interface to a television screen in the interior space of the passenger vehicle, to a cockpit display or to a head-up display of the driver. In a further embodiment of the invention, the interface of the system is a radio interface to a fleet operator, to a regulatory agency or to a hospital in order to inform the respective agency about the information of the identified disease-specific odors. According to one aspect of the invention, the interface of the system is an interface to a cloud infrastructure coupled to an operator and administrative authorities. Information, in particular information from a plurality of passenger vehicles, can thus be collected centrally and better distributed.

Initiating measures for annunciating a specific disease and/or for avoiding the spread of a specific disease, comprising:

an indication signal, e.g. 1.5m, informing the passenger of the minimum distance to be kept from the respective nearest passenger;

activating a disinfection system of the passenger vehicle;

alert to the hospital;

alerting authorities in the areas of disease monitoring, disease prevention, infectious diseases and/or non-infectious diseases, including federal health authorities, central authorities, authorities such as the robusty institute, wherein according to a further aspect of the invention, the indication of identified disease-specific odours is forwarded to these authorities via an interface, advantageously in the form of a radio interface, preferably via a cloud infrastructure, and provided to further personnel;

networking, for example via a cloud infrastructure with passengers of a passenger vehicle, passengers of other passenger vehicles, traffic participants and other persons, for example by mobile application software, for example a Corona warning application, for informing information about identified diseases, for tracking contactors and/or preventing the spread of diseases, wherein the networking comprises the provision of indication signals and further information of the above-mentioned mechanisms;

reading a memory in data exchange with the processing unit in order to obtain information in advance about the identified disease-specific smell;

initiating quarantine measures;

collecting the contact information of passengers and driving the passenger transport means to a place outside the hospital or the densely populated area;

wherein, according to a further aspect of the invention, no further passengers are received during travel and are not allowed to leave the passenger vehicle in order to further suppress the propagation.

According to a further aspect of the invention, the method is implemented by a computer.

Further aspects of the invention relate to a computer-readable storage medium/computer-readable data carrier. The computer readable storage medium/computer readable data carrier comprises machine learning algorithms and software instructions and/or hardware instructions which, when implemented by a processing unit, cause the processing unit to implement the method according to the invention.

The first sensor and according to a further aspect of the invention the second and/or third sensor is arranged in a seating area, a standing area, a ventilation installation and/or an air conditioning installation of the passenger vehicle. According to a further aspect, the first and/or second and/or third sensor is arranged on a horizontal and/or vertical holding bar of the passenger vehicle or on a safety belt. The seating and standing areas also include cabins and compartments. This arrangement enables odor detection directly at the source of the odor and also enables odor not directly detected in the interior space, for example due to air flow, but propagating in ventilation and/or air conditioning facilities.

According to a further aspect of the invention, the passenger vehicle can be operated automatically. The passenger vehicle comprises: an environment recognition sensor that recognizes an environment of the passenger vehicle; a controller that obtains data from the environmental recognition sensor and determines adjustment and/or control signals for trajectory planning; and actuators for longitudinal and/or transverse guidance, by means of which the control unit regulates and/or controls the automated operation of the passenger vehicle. The environment recognition sensor comprises a camera, a radar, a laser radar and a microphone. The controller is, for example, an ADAS/AD ECU, i.e., an advanced driving assistance system and automatic driving (advanced driving assistance system und autonomous driving). Automated operation ranges from assisted driving to autonomous driving, i.e. unmanned. In particular for highly automated operation, this includes: measures are initiated to automate the special travel of passenger vehicles to sites outside of hospitals or densely populated areas to avoid diffusion.

According to one aspect of the invention, the passenger vehicle is a bus, mass transit, rail vehicle, vessel, aircraft, or cable car. The present invention thus improves passenger confidence in the use of buses, mass transit vehicles, rail vehicles, ships, aircraft and cable cars, especially when pandemic cases such as covd-19 occur.

Drawings

The invention is illustrated in the following examples. Wherein:

FIG. 1 shows an embodiment of a system according to the present invention;

fig. 2 shows an embodiment of a method according to the invention;

FIG. 3 shows an embodiment of a passenger vehicle according to the present invention;

FIG. 4 shows a further embodiment of a passenger vehicle according to the invention;

FIG. 5 shows a further embodiment of a passenger vehicle according to the invention; and is also provided with

Fig. 6 shows a further embodiment of a passenger vehicle according to the invention.

Detailed Description

The system 10 in fig. 1 comprises a first sensor 11 which detects odor-causing substances via surface adsorption and/or via light interaction. This is method step V1 in fig. 2. The data of the sensor is evaluated by the processing unit ECU, i.e. the electronic controller. The processing unit ECU determines the biomarkers for the disease in the data of the sensor 11 by means of machine learning and thus identifies the disease-specific smell, which corresponds to method step V2. In method step V3, the passenger P and the administrative authorities are informed about the identified disease-specific smell, for example via the interface 12 of the system 10. In method step V4, for example, quarantine measures are initiated.

The passenger vehicle 1 in fig. 3 is a bus B, for example a city bus. In fig. 4, the passenger vehicle 1 is a mass-transit vehicle PM. In this embodiment, the dog is shown as a further passenger P whose smell, including breathing air and body smell, is detected by the sensors 11, 13, 14. In fig. 5, the passenger vehicle 1 is a rail vehicle, for example a long-haul train, a subway or a tram. In fig. 6, the passenger vehicle 1 is a ship S, for example a business ship or a pleasure boat.

In the bus B, the first sensor 11, the second sensor 13 and the third sensor 14 are arranged in a standing area "standing" of the bus B, for example on horizontal and vertical holding bars, in ventilation facilities AV and air conditioning facilities AC, and in a mass transit means PM also partly in a seating area "sitting". Thus, the ambient air including the breathing air and the body taste of the passenger P is detected together with the temperature and the air humidity. The data of the sensors 11, 13, 14 are fused in the processing unit ECU and the odor specific to the disease is evaluated. The evaluation result is displayed on a display unit "display" in the bus B via the interface 12. The interface 12 comprises a radio interface WLAN via which the evaluation results are provided to an external location "cloud".

List of reference numerals

1. Passenger transport means

10. System and method for controlling a system

11. First sensor

12. Interface

13. Second sensor

14. Third sensor

ECU processing unit

P-passenger

"cloud" external location

"sitting" seating area

Standing area

AV ventilation facility

AC air conditioning facility

"display unit

WLAN radio interface

B bus

PM mass transport means

Rail vehicle

S ship

V1-V4 method steps

Claims (7)

1. A system (10) for identifying a disease-specific smell in a passenger vehicle (1), the system (10) comprising:

-a plurality of first sensors (11) detecting substances that generate odors;

a processing unit (ECU) that obtains data obtained by detecting a substance that generates an odor from the first sensor (11), identifies the odor that is specific to the disease by machine learning, and generates an indication signal depending on the identification result, wherein the processing unit (ECU) implements a machine learning algorithm that obtains the data of the first sensor (11) and learns or trains to identify biomarkers for the disease in the data, and

-an interface (12) providing an indication signal to a passenger (P), a driver and/or an external location ("cloud") of the passenger vehicle (1).

2. The system (10) according to claim 1, wherein the first sensor (11) detects odor-causing substances in the breathing air comprising the passenger (P) and/or in the ambient air of the body taste by surface adsorption, light interaction and/or frequency and/or amplitude of vibration.

3. System (10) according to claim 1 or 2, comprising a second sensor (13) measuring temperature and/or a third sensor (14) measuring air humidity, wherein the processing unit (ECU) fuses the data of the first sensor (11) and the data of the second sensor (13) and/or the third sensor (14).

4. A system (10) according to any one of claims 1 to 3, which is used for diagnosing a disease, comprising: including influenza and covd-19 viral diseases and variants of viral diseases.

5. A method for identifying a disease-specific smell in a passenger vehicle (1), the method comprising the steps of:

detecting odor-generating substances (V1) in the breathing air and/or in the ambient air including the smell of the passenger (P);

identifying a disease-specific smell (V2);

-informing passengers (P), drivers and/or operators and/or authorities of said passenger means (P) about one or more identified disease-specific odours (V3); and is also provided with

Depending on the identified disease-specific odour, initiating measures (V4) for annunciating specific diseases and/or for avoiding the spread of specific diseases including quarantine measures,

wherein the system according to any of claims 1 to 4 is used for performing the method.

6. Passenger vehicle (1) comprising a system (10) according to any one of claims 1 to 4, wherein,

the first sensor (11) of the system (10) is arranged in a seating area ("sitting"), standing area ("standing"), ventilation (AV) and/or Air Conditioning (AC) of the passenger vehicle (1), and

the interface (12) of the system (10) is an interface to a display unit ("display") in the passenger vehicle (1) and/or a radio interface (WLAN) to the cloud, the operator and/or the administrative authorities of the passenger vehicle (1).

7. Passenger vehicle (1) according to claim 6, comprising: an environment recognition sensor that detects an environment of the passenger vehicle (1); a controller that obtains data of the environment recognition sensor and determines adjustment and/or control signals for trajectory planning; and an actuator for longitudinal and/or transverse guidance, by means of which the controller regulates and/or controls the automated operation of the passenger vehicle (1).

Images