Method for supervising a thermal preparation process of food and cooking appliance using such method

Abstract

The present invention relates to a method for supervising a thermal preparation process of food (5), performed by a cooking appliance (1), the method comprising: - receiving (131) sensor data (135), the sensor data (135) captured by a sensor arrangement (12) in connection with the thermal preparation process of the food (5); - providing (132) the received sensor data (135) as input data (136) to a trained model (137), the model (137) trained based on a plurality of training data sets; - applying (133) the trained model (137) to the input data (136); and - determining (134), one or more cooking parameters of the food (5) undergoing the thermal preparation process; and - generating (135) or updating (135), based on the determined one or more cooking parameters, one or more of notification data (140) and one or more operating parameters (141) of the appliance (1).

Description

[0001] Description

[0002] METHOD FOR SUPERVISING A THERMAL PREPARATION PROCESS OF FOOD AND COOKING APPLIANCE USING SUCH METHOD

[0003] The present invention relates to a method for supervising a thermal preparation process of food, i . e . a thermal food preparation process , and a cooking appliance , such as a cooking or baking oven, a cooking range , a microwave oven, a steamer, broiler, grill , cooktop, or and combinations thereof . A cooking or baking oven may for example comprise a cavity, such as a muf fle , with one or more heating elements for performing the thermal food preparation process , e . g . cooking, baking, roasting, broiling etc . , and may have a vapor removal system, such as an exhaust system, for example provided for removing, e . g . exhausting, gases or vapors generated and released by the food during the thermal preparation process , such as cooking . In case of a cooking appliance comprising a mu f f 1 e , the vapor removal system may comprise an exhaust system with a ventilator, an exhaust duct and an exhaust opening for sucking air from the muf fle and exhausting the air with the vapors to the environment via the exhaust opening . The underlying invention is applicable for appliances not comprising a cavity or muf fle , such as cooktops , but including a removal system for removing vapors generated during the thermal food preparation process , e . g . cooking .

[0004] According to some known cooking appliances , air drawn from the muf fle may be analyzed and used for predicting a cooking status of a food item cooked in the muf fle .

[0005] US 10 , 244 , 788 B2 describes a cooking appliance comprising a cooking chamber for accommodating a tray for cooking food items . The appliance comprises a controller for monitoring cooking, which receives an indication of a type of the food article within the cooking chamber . The controller is associated with a fluid analysis assembly which comprises a dilution chamber in which cooking vapors drawn from the cooking chamber and ambient air are mixed . Mixing the cooking vapor with ambient air is controlled via a valve associated with the dilution chamber, wherein the valve is configured for adj usting ambient air flow into the dilution chamber dependent on the indication of the type of food . Based on an analysis of the mixed air using a plurality of sensors and the food type , a done cooking status is predicted and alerted to a user when the cooking status is a done cooking status . The known appliance is comparatively complex with regard to predicting the done cooking status and may generate divergent prediction results in di f ferent operating environments .

[0006] EP 1 489 361 A2 describes a method of controlling a cooking process in a cooking oven . The method involves measuring a concentration of atmospheric gases within a cooking chamber of the oven, wherein one or more sensors for measuring the atmospheric gases within the cooking chamber are provided . The known method involves comparative complex signal processing of the sensor signals . WO 2020 / 230056 Al relates to household appliances such as kitchen hoods , steam ovens or electric ovens , comprising a plurality of transducers for detecting humidity, temperature and carbon dioxide concentrations in gases generated when a food item is cooked .

[0007] US 10 , 002 , 965 A describes a cooking apparatus comprising a cooking chamber and a gas sensor configured to detect a gas generated by a food item inside the cooking chamber . A controller is provided to determine a type of the gas and a concentration of the gas detected by the gas sensor based on a change in a resonant frequency of a component of the gas sensor, and to determine a cooking progress state of the food item in the cooking chamber . Information on the determined cooking progress state is transmitted to an external apparatus through a communication device to inform a user of the cooking progress state .

[0008] Albeit the known cooking appliances provide some solutions for predicting a cooking state of a food item, there is still room for improvement . In particular, it is desirable to improve cooking state prediction or determination with particular regard to reliability and accuracy . Further, there is still room for improvement with regard to ease of implementation, particularly with a broad area of application .

[0009] It is an obj ect of the invention to provide an improved method for supervising a thermal preparation process of food, in particular a cooking process for food, and a cooking system configured for carrying out such a method . In particular, a method and system shall be provided that address the above issues .

[0010] This obj ect is solved by the combination of features of the independent claims . Embodiments result from the dependent claims and the exemplary embodiments described below, and in connection with the annexed figures .

[0011] According to an embodiment , a method for supervising a thermal preparation process of food, or, in other words , a thermal food preparation process , is provided . The method is in particular suitable for real-time supervising the thermal preparation process . The thermal preparation process of food may relate to baking, cooking roasting, grilling, steaming, etc . performed by a cooking appliance . The cooking appliance may be a baking oven comprising a muf fle , a range , a cooktop, a microwave oven, a grill , a roaster of any combination thereof .

[0012] The method may comprise the steps described below, preferably carried out continuously or repeatedly during the thermal preparation process of food (herein also : thermal food preparation process or food preparation process ) .

[0013] The method may comprise a step of receiving sensor data, the sensor data captured by a sensor arrangement in connection with the thermal preparation process of the food . The sensor arrangement may comprise one or more sensor units associated with the cooking appliance . For example , the sensor units may be arranged or positioned, in particular near, on, or within the cooking appliance , such that the sensor units can detect or capture data related to the thermal food preparation process . Data related to the thermal food preparation process may comprise food-related data such as : type of food, volume of food, shape of food, color, surface and / or texture of food, composition of food ( i f there are two or more food items or components representing the food to be processed) . Further, the sensor data related to the thermal food preparation process may comprise foodpreparation-related data such as : temperature ( s ) associated with the thermal preparation process , humidity level , moisture content , presence and / or amount of airborne substances ( or : compounds ) released by the food or generated in the course of the thermal cooking process and contained in vapors expelled or sucked away .

[0014] The sensor data comprise sensor data captured by at least one multivariate sensor assembly (which may be form a kind of "electronic nose" ) arranged and configured to detect di f ferent airborne compounds that are present in vapors released from the food or generated during the thermal food preparation process . The sensors are selected such that one or more substances that are present during the thermal preparation and vary in concentration of amount in the course of the thermal preparation can be detected . Respective sensors may for example be adapted to sense substances such as 02( content ) , C02( content ) , volatile components , in particular volatile organic compounds (VOC ) or other volatile substances and compounds released or generated during the preparation process of food, and generally included in cooking vapors . Such substances or compounds may be related, but not limited to , volatile Maillard-Reaction-Products (MRP ) or Stecker-Reaction products , lipids or oxidation products , carameli zation products, thermal degradation products of organic compounds, volatile aroma compounds, etc., in particular melanoidines , pyrazines, aldehydes, ketones, furans acrylamides, terpenes, alcohols, carboxylic acids, esters, aroma components, sulfur compounds, nitrogenous compounds, etc.. Sensors for detecting such substances or compounds may be: optical sensors, metal-oxide sensors (MOS) , in particular MOS for VOCs (such as terpenes, alcohols, aldehydes, organic acids) , PID sensors, NOx sensors. Some of the compounds or substances (such as acrylamides, furans etc.) may detected indirectly, for example based on temperature, temperature profile, degree of browning.

[0015] The method further comprises a step of providing the received sensor data, in particular data corresponding to signals of the sensor units, as input data to a trained (mathematical) model. The model is trained based on a plurality of training data sets, each providing, for respectively one or more (e.g. a mixture or composition of) food items, associations between at least one cooking parameter, such as a cooking degree, e.g. level of doneness, browning etc., related to the one or more food items, and associated compounds, such as one or more volatile organic substances, released by the respective one or more food items or generated during the thermal preparation process of the respective one or more food items. The sensor data or sensor signals may be representative of a concentration, quantity or amount of the compound in a vapor extraction or removal airstream. The model may be a multivariate model suitable for processing multiple inputs (variables or features) as input, in particular sensor data from sensors of different type (volatile compound concentration, volatile compound type, temperature, image data etc-) . The model may be trained to be able to process multivariate input data . Further, the model may be configured to learn relationships between multiple variables and how such variables collectively influence the output . Multivariate variables may relate to one or more of : internal / external food temperature , humidity, cooking time , type of food, image-based color features of the food, concentration / presence of volatile compounds or gases in cooking vapors etc . . The model may learn or be trained how all these inputs , taken together, may correlate with a desired output such as "done" , "underdone" , "overcooked" and / or operating or control parameters for adapting the process .

[0016] The method comprises the further step of applying the trained model to the input data . Accordingly, applying the trained model to the input data will result in output data that are , as a result of the training mentioned beforehand, indicative of or represent one or more cooking parameters related to the food and the underlying thermal preparation process .

[0017] The method comprises the further step of determining, based on output data obtained by applying the trained model to the input data, one or more cooking parameters for the food undergoing the thermal preparation process . For example , the one or more cooking parameters may relate to a level of doneness , a cooking degree , a degree of browning etc .

[0018] The method comprises the further step of generating ( or : determining) or updating, based on the determined one or more cooking parameters , one or more of : notification data for providing a notification representative of the one or more cooking parameters to a user of the cooking appliance, e.g. a visual or acoustical notification on a user interface of the cooking appliance or an external mobile device, such as a smartphone, and one or more operating parameters (or: control parameters) for controlling one or more subsystems of the cooking appliance, wherein the one or more subsystems are at least temporarily involved in the thermal preparation process the food. The one or more subsystems may relate heaters, fans, convection units, vaporizers etc..

[0019] Based on the trained model, the one or more cooking parameters, such a level of doneness etc. can be determined or predicted comparatively accurately. In particular, the method enables a comparatively accurate determination of prediction of an endpoint of the thermal food preparation process, e.g. a cooking or baking process in a cooking chamber of a cooking oven. Further, the method is comparatively flexible and can be used for a plurality of appliances. Due to the possibility of executing the method at least in part on a remote computing device communicatively coupled with the cooking device, it is not absolutely necessary that the cooking device itself is provided with control units or computing devices having a high processing power. Determined sensor data may be received from the cooking appliance in realtime or substantially real-time at the (remote) computing device to generate the operating parameters for transmission to the cooking device under real-time or substantially real-time conditions. According to embodiments, the sensor arrangement comprising the one or more sensor units may further comprise at least one further sensor unit. The at least one further sensor unit may comprise at least one of: one or more optical sensors configured for capturing image and / or video data of the food at least one of prior to and during the thermal preparation process; one or more temperature sensors (such as contact-based temperature sensors, e.g. food-sensors, or non-contact temperature sensors, such as IR or optical temperature sensors) for detecting temperatures associated with the thermal preparation process of the food, one or more humidity sensors for detecting humidity levels associated with the thermal preparation process of the food, one or more acoustic sensors for detecting cooking sounds generated by the food during the thermal cooking process, i.e. sounds emitted as the food undergoes the thermal cooking process, one or more ultrasonic sensors for detecting one or more physical parameters of the food (e.g. shape, volume, position etc.) , detecting means for detecting an actual power level, heating level, and / or power consumption of the cooking appliance, e.g. of a heater etc., for performing the thermal cooking process. The method may comprise providing other data or parameters as input data to the trained model such cooking time, elapsed time since start of the thermal cooking process etc..

[0020] In this connection, the method may further comprise the steps of: receiving further sensor data captured by at least one of the least one further sensor unit, providing the received further sensor data as additional input data to the trained model , applying the trained model also to at least a part or a subset to the additional input data .

[0021] In connection with the further sensor data, the model may further be trained with regard to sensor data associated with the at least one further sensor unit , based on a plurality of further training data sets , each providing, for respectively one or more food items , associations between the at least one cooking parameter, such as a cooking degree , level of doneness , food type , temperature , etc . , related to the one or more food items , and sensor data speci fic to a respective further sensor unit .

[0022] In general , the training data sets and further training data sets may be obtained or originate from multiple thermal food preparation processes conducted under defined conditions , for example , and / or based on respective food items themselves , through images or similar descriptive data .

[0023] In the embodiment mentioned beforehand, the further steps of determining one or more cooking parameters and generating or updating at least one of noti fication data and one or more operating parameters may be carried out analogously . The further sensor data provide additional input and information for improving accuracy of the cooking parameters , thereby enabling, i f applicable , improved supervision of the thermal food preparation process .

[0024] In embodiments , the thermal food preparation process may be conducted automatically, or at least in parts automatically, for example subsequent to the initiation or start of the thermal food preparation process.

[0025] It is to be noted that the present disclosure explicitly considers any combinations and number of different sensors and sensor data mentioned above. Providing a plurality of different sensor data and corresponding sensors is advantageous with regard to covering a plurality of different thermal food preparation processes for a large variety of different foods and food combinations. It is to be noted that even if the method may be implemented for using or processing a plurality of different sensor data from different sensors, the further sensor data may be selected as required or as applicable, for example depending on the type of food, the type of thermal processing (e.g. baking, roasting etc.) . In embodiments, and based on a model that is trained based on training data including the plurality of sensor data for the food items, the method may be conducted without the need to select and / or exclude particular sensor data. By this, the level of automating the thermal food processing may be improved .

[0026] In a particular advantageous embodiment, the method and the trained model involves at least sensor data obtained from one or more multivariate sensor assembly (e.g. an "electronic nose") and from one or more optical sensor units, such as cameras or image capturing devices. Combining such sensor data has shown to provide reliable results, in particular as airborne substances in process vapors and images of the food undergoing the thermal processing or preparation may be comparatively specific such the combined use of respective sensor data may lead to improved determinations of the cooking parameter (s) . Further, using optical image or video sensors has may involve determining the type, volume, shape etc. of the food (e.g. without requiring user input regarding the type of food and / or type of preparation process) , which may contribute to further automation. In addition, the combined use of different sensors or sensor types (e.g. for detecting airborne substances and capturing images of the food) has the advantage that potential deficiencies or limitations of one sensor type may be compensated for by the other sensor type. As an example, if the field of view of an image sensor is partly limited or obscured (by objects or other food items, vapor etc.) the "electronic nose" may compensate for by detecting relevant airborne substances. Further, shortcomings of an image sensor in connection with low contrast or similar food appearance may be compensated for by detecting airborne substances in, e.g. cooking vapors etc.

[0027] According to embodiments, the trained model may comprise (i) a unified model architecture that jointly processes the input parameters to generate the output data, or (ii) a collection of separate models (or: sub-models) , each configured to process a subset of the input data and respectively tailored to process a particular type or a selection or combination of particular tapes of sensor data associated with a type of sensor unit or types of sensor units of the one or more sensor units of the sensor arrangement .

[0028] In embodiments, the method may involve shared input-to- output processing (e.g. considering all available sensor data) or modular input-to-output processing (e.g. considering particular sensor data or groups of sensor data by respective models of sub-models. Further, the method may involve a stacked model architecture. For example, image data of the food may be processed to identify the food or food items, and the execution of the further thermal preparation process and / or the capturing or the (pre- ) processing of sensor data may be based on the identified food or food items. Further, respective data may be provided as information and input data to the model for consideration and determining the cooking parameters. Further, airborne substances may be specific enough for identifying the food or food items, and the execution of the further thermal preparation process and / or the capturing or the (pre- ) processing of sensor data, e.g. of image sensor data, may be based on the identified food or food items. For example, if it is determined that the food relates to dough (e.g. in connection with baking cake or bread) , the image (pre- ) processing may be tailored to dough (cake or bread) etc. Further, available information related to cooking parameters, such as food type etc., level of doneness etc. may be provided as (fixed) input data in connection with applying the trained model to the sensor data, which may be advantageous if the method is carried out continuously or iteratively during the thermal food preparation process.

[0029] According to embodiments, the trained model may be implemented according to (ii) above, and the method may further comprise: aggregating, fusing, concatenating, or merging, at least in part, outputs of the separate models (or sub-models) to determine or generate at least one of: the one or more cooking parameters, display data, and operating parameters. According to embodiments, the method may further comprise controlling the thermal preparation process of the food by: applying at least one of the one or more operating parameters to at least one of one or more subsystems of the cooking appliance (e.g. heaters, fans, convection units, evaporators etc.) , instructing a user interface associated with the cooking appliance and an associated user to issue a (e.g. visual, haptic and / or acoustic) notification representative of at least one of the one more cooking parameters. As an example, the method may involve adapting a heating level of a heating element, an operating speed of a fan etc.. Further, the method may involve stopping the thermal food preparation process and / or generating a notification on a user interface subsequent to determining that an endpoint of the thermal food preparation process is reached.

[0030] According to embodiments, the one or more operating parameters may comprise at least one of: one or more power levels or heating levels of one or more heating subunits of the cooking appliance, an estimated remaining cooking time, a predicted doneness endpoint, an operating level of a cooling fan or of a convection fan, an operating level of a vaporizer (or steam generator) , an operating level of a cooking vapor extraction system or steam exhaust system etc. The one or more heating subunits may be selected from: microwave heating units, resistance heating units, induction heating units, in case of a cooking oven with a cavity: an upper o lower heating unit, a top heater, a griller, an infrared heating unit, etc..

[0031] According to embodiments, the method may be carried out, at least in part , by at least one of : a control unit of the cooking appliance and a remote computing device communicatively coupled with the cooking appliance . The remote computing device , e . g . a computer, in particular server, may be in data communication with the cooking appliance and configured to : process sensor data ( received from the cooking appliance and obtained from sensors of the cooking appliance ) , and provide control signals to the cooking appliance for controlling the thermal food preparation process ( or : thermal food processing) . In embodiments , the sensor units may be part of the cooking appliance . In embodiments , one or more of the sensor units may be external to the cooking appliance and communicatively coupled to the cooking appliance , the control unit and / or the remote computing device for transmitting sensor data, speci fically sensor data associated with the thermal food preparation process , to the appliance , the control unit and / or the remote computing device .

[0032] According to embodiments , one or more of the sensor units of the sensor arrangement are ( i ) built-in sensor units , integral with the cooking appliance ( e . g . physically and functionally incorporated into the appliance as a sensing subunits ) and / or ( ii ) sensor units external to the cooking appliance and communicatively coupled to the cooking appliance and / or a processing device configured for carrying out the method . The external sensor units may be configured for transmitting respective sensor data over a data communication link to the cooking appliance and / or the processing device for use in connection with the method . The expression "integral with" shall, in particular, mean that the sensor unit is not merely externally attached or loosely coupled, but rather: permanently built into the appliance's housing, a cooking chamber, or a control system, functionally connected to its control logic or electronics, and designed as part of the original system. The expression "built-in" shall, in particular, cover sensor units that are externally attached or loosely coupled to the cooking appliance, e.g. using releasable mounts, quick-release mounts, snap-fit connections, magnetic mounts, bayonet-mounts, slide-in mounts, cold shoe mounts, hot-shoe mounts etc.. In particular optical sensors, such as an imaging unit or a camera, may be releasably attached to the cooking appliance, e.g. to a door handle or grip, and attached to the appliance if required. Data exchange between the cooking appliance and such sensors may be based on a wired communication link (e.g. if a hot-shoe connection of a cable is used) or a wireless communication link.

[0033] According to embodiments, the cooking parameter may be one or more of, but not limited to: a type, category or composition of food, a level of doneness, a finishing endpoint of the thermal food preparation process, a browning level, an inner food temperature, outer or ambient temperature, a surface temperature of the food, a humidity level, a food weight, a food volume, a food geometry, a food distribution, a food position, a rack position, a crust level, a surface browning, a surface texture, a Maillard level, a cooking stage, a change of rate of sensor data, a change velocity of senor data, etc.

[0034] According to embodiments, the method may comprise: initiating a thermal food preparation process, e.g. a baking or cooking process; obtaining first sensor data captured by one or more sensor units of the sensor arrangement in a timespan covering a pre-phase and an initial phase of the thermal preparation process; processing the first sensor data, preferably by feeding the sensor data as input data to the trained model (for example, but not limited to a sub-model) , and determining at least one of: food-related parameters or properties (e.g. a food property, food characteristic, such as type of food, food size, food volume, food composition) , and sensor base level signals for the sensor units, e.g. base levels of one or more of the compounds detected by the multivariate sensor and present in the pre-phase or initial phase, such as VOCs etc., humidity, C02, VOC, oxygen level, carbon dioxide level, level of compounds released etc.; based on the food-related properties: automatically selecting or suggesting for selection by a user at least one of a food type, a food category, and a food composition, and / or correcting a program for thermal food preparation; and based on the sensor base level signals, setting respective sensor base levels (which may imply calibration, initialization, offset correction or sensor adjustment) ; performing the thermal food preparation process and continuously or repeatedly performing the steps according to claim 1, (wherein the baseline values may be used for correcting or adapting actual sensor data, and the corrected sensor data may be provided as the input data to the trained model) .

[0035] In connection with this disclosure, capturing sensor data may involve any of the processes of any combinations of: acquiring, collecting, reading or recording data.

[0036] According to embodiments, the method may further comprise, predicting, by applying the trained model to the sensor data, an endpoint for the thermal preparation process of the food, and, if the endpoint is reached, generating a notification to a user and / or stopping or initiating a final phase of the thermal preparation process.

[0037] According to embodiments, a control unit comprising at least one processing unit is provided. The control unit is programmed or configured such that, when operated, the processing unit carries out the steps of a method of supervising a thermal preparation process of food, e.g. a cooking or baking process, described herein. For example, the control unit may be communicatively coupled with a memory storing instructions, which when executed by the control unit cause the control unit to carry out the steps of a method of supervising a thermal preparation process of food, e.g. a cooking or baking process, described herein. As noted herein elsewhere, the control unit may, at least in part, be part of the cooking appliance (e.g. internal to the appliance) or of a remote computing device (e.g. a cloud-based device, network-based device, or mobile device, such as a smartphone, tablet etc. - external to the appliance) , communicatively coupled with the cooking appliance or components thereof , e . g . via a wired or wireless communication link, such that the appliance and control unit , when in operation, are suitable for carrying out a method according to any embodiment described herein .

[0038] According to embodiments , a cooking appliance for thermal food preparation of food, or a cooking system comprising the cooking appliance for thermal food preparation of food, may be provided . The appliance or system may comprise a sensor arrangement as described herein elsewhere , in particular including at least one multivariate sensor assembly configured to detect airborne compounds present in vapors released during the thermal food preparation process from the food, and a control unit according to any embodiment described herein . In particular, the control unit is configured for carrying out a method according to any embodiment described herein .

[0039] In embodiments , a cooking system for thermal food preparation of food may be provided . The cooking system comprises a cooking appliance , a sensor arrangement including at least one multivariate sensor assembly configured to detect airborne compounds present in vapors released during the thermal food preparation process from the food, and a control unit or computing unit comprising one or more data processing units , which may be internal , external and / or remote ( e . g . servers , cloud devices , mobile devices ) to the cooking appliance , configured for carrying out , when in operation, a method according to any embodiment described herein . According to embodiments , which can be claimed separate or in combination with any other embodiment described herein, a cooking appliance , or cooking oven, is provided . The cooking appliance , herein also referred to as appliance , comprises a cooking cavity unit comprising a cooking cavity or cooking chamber ( or oven cavity) , such as a muf fle . The cavity shall be understood as a spaced defined by surrounding walls , such as a bottom wall , a top wall and three side walls , with a front opening to be closed with a front door of the appliance , wherein the cavity is adapted to accommodate food items for being cooking, for example placed on a tray .

[0040] The cooking cavity unit shall , in particular, be understood as a component of the appliance including the cooking cavity and associated with additional components such as thermal insulating elements or layers ( thermal insulation) surrounding the cavity walls and / or housing elements , such as metal sheets , surrounding the cavity and the thermal insulation, for example a kind of chassis .

[0041] The cavity has an exhaust port , for example provided in the top wall . The exhaust port is adapted to draw or exhaust air from within the cavity . During cooking such air typically includes fumes or gases generated by a food item placed within the cavity during cooking . Provided that the exhaust port it is located in a top wall of the cavity, the exhaust port may also be considered as an exhaust chimney, enabling upward air extraction or exhaustion . The cooking appliance may be configured for use as a built-in appliance or oven, wherein stand-alone appliances shall not be excluded .

[0042] The appliance may further include , in accordance with embodiments , a service compartment or service area . The service compartment shall be considered as providing, at least in the installed configuration, e . g . when installed in a kitchen unit , a free space suitable for accommodating or housing one or more components such as electronics , exhaust or cooling fans , exhaust ducts , wiring and the like . The service compartment may for example be provided above the top wall of the cavity, and may be spaced from a corresponding cavity wall by intermediate insulation and housing components or cover elements surrounding or covering the cavity wall ( s ) . The service compartment may at least in part be defined or delimited by walls of a chassis of the appliance , such as for example lateral side walls of the chassis delimiting or enclosing a space located at the top of the appliance , above the cavity . A base area of the service compartment may correspond to an area si ze of a corresponding vertical or hori zontal cross section of the appliance and / or the chassis of the appliance . It is to be noted that the service compartment may be located at an upper or top side of the appliance . However, other locations such as on a lower side , at the back or at lateral sides shall not be excluded .

[0043] The service compartment or service area may be arranged, defined or provided adj acent to an outer face of the cooking cavity unit . An outer face is to be understood as a surface or side of the cavity unit that is oriented outwardly when viewed from the cavity interior .

[0044] The appliance further comprises a cooling fan assembly or vapor extraction assembly comprising a vapor extraction fan, and a multivariate sensor assembly described in more detail below .

[0045] The cooling fan assembly or vapor extraction assembly may be arranged in the service compartment , which may be considered a service compartment area or space . The cooling fan or vapor extraction assembly comprises a cooling fan or exhaust fan that is air- fluidly connected to the exhaust port for sucking air from the cooking cavity . In other words , the cooling fan or exhaust fan, herein also referred to as fan, is configured to extract or suck air from the cavity via the exhaust port , the exhaust port providing a fluidic connection between the cavity interior and the fan . The exhaust port may for example be associated with an opening in a wall , e . g . the top wall , of the cavity . Preferably, the interconnection between the exhaust port and a corresponding suction port of the fan is fluid-tight ( air-tight ) , in particular, to prevent the fan from sucking external air via the interconnection provided by the exhaust port between the suction port of the fan and the cavity . In such embodiments , the exhaust port is configured such that the fan only draws air from within the cavity through the exhaust port and a corresponding input or suction port of the fan . However, in embodiments , the interconnection between the suction port of the fan and the cavity defined by the exhaust port may be adapted to enable the fan sucking external air in addition to air from the cavity interior . This means that the fan may be enabled so suck, via the exhaust port a mixture of ambient air and air from the cavity interior . In embodiments , the fan may comprise a separate suction port provided for sucking ambient air from the service compartment , meaning that the fan may be configured to suck air from the cavity via the exhaust port and from the service compartment via the separate suction port , resulting in a mixture of air drawn from the cavity and the service compartment . For the reason that the temperature in the service compartment is generally signi ficantly lower than the temperature in the cavity, the temperature of air mixture to be exhausted may be considerably lower than that of the cavity interior . This is advantageous for avoiding hot air to be expelled from the appliance . Further, for the reason that the fan assembly is arranged in the service compartment , ambient air is sucked from the service compartment , which air is , at least to some extent , decoupled from external environmental conditions , such as external humidity, external temperature , or substances present in the external environment , which means , that the air mixture expelled by the fan to a large extent mirrors the air or gas composition prevailing within the cavity during cooking .

[0046] The cooling fan assembly further comprises an exhaust duct with one or more walls enclosing or defining an exhaust channel that extends between the fan and an exit opening (or outlet opening) of the exhaust duct . The cooling fan assembly being arranged in the service compartment , implies that the fan and the exhaust duct are arranged in the service compartment .

[0047] The exhaust duct is air- fluidly associated with or connected to the fan such that , in operation, an airstream generated by the fan and including air sucked from the cooking cavity through the exhaust port passes through the exhaust channel and is exhausted via the exit opening .

[0048] One wall of the one or more walls of the exhaust duct may comprise an exhaust duct wall section that merges into the exit opening . The outer side of this exhaust duct wall section faces away from the cooking cavity unit , which shall mean that the exhaust duct wall section is defined on side averted from the cooking cavity unit , in particular averted from the cavity . For example , the exhaust duct may be arranged and extend substantially parallel to a wall of the cavity, preferably substantially parallel to an upper wall of the cavity . I f provided in an upper section of the appliance , and the exhaust duct wall section may then be arranged in an upper part or section of the exhaust duct . In particular, the exhaust duct wall section may, within the boundaries defined by the exhaust duct walls , be spaced at a greatest possible distance from the cavity relative to a fictitious line between the cavity center and the center of area of the exhaust duct wall section . By this , the exhaust duct wall section may be provided in an area of the exhaust duct that is associated with reduced thermal load caused by high temperatures within the cavity during cooking as compared to section closer to the cavity . Further, the exhaust duct wall section, facing away from the cooking cavity unit may be easily accessible , for example from top, which may be beneficial with regard to placing the multivariate sensor described below .

[0049] The expression merging into the exit opening in particular shall be understood in a sense that the exhaust duct wall section is located near and / or substantially immediately adj acent to the exit opening . For example , the exhaust duct wall section may be arranged at a position that is closer to the exit opening than to the fan following, for example , a flowline of an air stream generated by the fan and through the exhaust duct . This position may be considered advantageous because the further the distance from the fan, the lower the temperature of the air sucked from the (hot ) cavity and exhausted via the exhaust duct . This may be beneficial with regard to placing the multivariate sensor at positions related to comparatively low temperature loads . Placement of the multivariate sensor is described in detail below .

[0050] The multivariate sensor assembly may be arranged in a section of a duct wall of the exhaust duct channel , in particular in the exhaust duct wall section, with a sensing area of the sensor directly facing the exhaust channel , in particular the interior of the exhaust channel , such that the airstream passes the sensing area before exiting the exit opening, in particular to enable the multivariate sensor assembly to detect one or more airborne components , described herein elsewhere , present in the airstream . As mentioned above , the placement of the multivariate sensor in the exhaust duct wall defined above is advantageous with regard to lowering the temperature load caused by air sucked from the cavity and with regard to improved accessibility of the multivariate sensor, for example in connection with maintenance and repair . Further, the suggested arrangement of the multivariate sensor is suitable for use with known cooling fan arrangements without requiring substantial redesigns .

[0051] According to embodiments , the multivariate sensor assembly may be configured to detect in the airstream, in particular air mixture , generated by the fan in or through the exhaust channel at least two , three or more volatile substances , such as volatile organic compounds (Volatile Organic Compounds : VOC ) that are exhausted or released by a food item during a thermal food preparation process , in particular a cooking process , into the cavity . For the reason that the fan assembly may arranged in the service compartment , and i f air from the service compartment is mixed with air drawn from the cavity, possible adverse influences from substances present in the outer environment of the appliance , such as in a kitchen environment , may be avoided . Hence , the substances generated or released by the food item may reliably be detected . As mentioned, air from the service compartment may be mixed with the air drawn from the cavity to lower ( cool ) the temperature of the exhaust air stream . On the one hand, this may be to avoid burning hazards for example user' s hands coming close to the exhaust duct . On the other hand, the multivariate sensor may be protected from high temperature loads otherwise generated by the hot (uncooled) air drawn from the cavity . As mentioned cooling may be obtained by mixing the air drawn from the cavity and air from the service compartment . However, active cooling arrangements or elements for cooling the airflow passing the sensing area shall not be excluded . Drawing air from the service compartment may also provide a cooling ef fect within the service compartment , which may be advantageous for cooling components , such as electronic components , arranged in the service compartment and for avoiding accumulation of heat within the service compartment .

[0052] In embodiments , the appliance further comprises a control unit , wherein a or the trained model described herein elsewhere is implemented on the control unit . As indicated the model may be trained based on a plurality of training data sets each providing, for respectively one or more food items , associations between a cooking degree of the one or more food items and related exhaust of one or more volatile substances , e . g . volatile organic substances (VOC ) . The control unit is configured to receive sensor signals of at least the multivariate sensor assembly and use these sensor signals as input data for the trained model . In general , the trained model may be a neuronal network, a machine learning model ( e . g . SVM classi fier ) or similar . By applying the input data to the trained model , a cooking degree for a food item present in the cavity during cooking may be determined or predicted . By using a multivariate sensor assembly, the control unit may predict the cooking degree or status without knowing the food item, type of food and / or food item weight . In particular, the control unit may be adapted to determine the cooking degree or status based on the trained model by using the signals of the multivariate sensor as input data for the trained model and to output data of the trained model as the cooking degree / status or as data representative of the cooking degree / status . In embodiments , not only the cooking status , but also operating parameters for the cooking appliance , and for applying to the cooking appliance for controlling the thermal preparation process , e . g . a cooking or baking process , may be provided by the model .

[0053] In particular due to the advantageous placement of the multivariate sensor and the avoidance of external environmental influences , the cooking degree , such as a level of doneness , may be determined comparatively reliably and accurately, in particular substantially regardless of the installation location of the appliance , or at least under conditions reducing possible environmental impacts as compared to solutions sucking air from an environment outside of a compartment or housing accommodating the appliance . Further, and as mentioned, the control unit may determine the cooking status or degree solely based on the signals from the multivariate sensor, without requiring additional input data such as a cooking time and / or type or kind of food etc . . However, additional sensor data ( from other sensors ) and additional information, e . g . actual operating parameters of the appliance , may be used, which may improve the accuracy and / or ef ficiency of the supervising method .

[0054] The advantageous placement of the sensor assembly in an area of the exhaust duct exposed to comparatively low temperatures enables the use of MOS sensors , which are generally limited to low temperature environments , and which have been found to be ef ficient for detecting VOCs . Using several of such sensors enables detecting the cooking status or degree based on the sensed VOCs only .

[0055] In an embodiment , the cooling fan may comprise a first input port that is air- fluidly connected via the exhaust port to the cavity, a second input port that is air- fluidly connected to or opens into the service compartment , and an output port . The output port may be air- fluidly connected to or opens into the exhaust duct . The fan is configured such that , when in operation, air is sucked from the cavity and service compartment via the first and second input ports , and an airstream comprising a mixture of air sucked from the cavity and the compartment is generated and exhausted through the exhaust duct , past the sensing area, and out of the exit opening to the outer environment of the appliance . As discussed above , the air mixture may be advantageous for reducing the temperature load to the sensor caused by the airstream passing by . Further sucking air from the service compartment may provide a cooking ef fect for the service compartment .

[0056] In embodiments , the exhaust duct wall section may be tapered towards the exit opening, or the exhaust duct wall section may be provided in a tapered section of the exhaust duct . Providing the exhaust duct wall section in a tapered area of the exhaust duct may be advantageous with regard to exposing sensing components , in particular the sensing area, of the senor with the airstream, which may lead to improved sensing accuracy and reliability . Further, such an arrangement may be considered beneficial with regard to flow dynamics related to the airstream generated by the fan . In particular, it was found that such tapered sections may be associated with smooth or laminar flow conditions ( rather than turbulent flow conditions and vortex generation) , which is beneficial for the sensor signal generated from the airstream accurately mirroring the composition and ( relative ) substance quantities of the air within the cavity . In other words , the exhaust duct wall section, and therefore the multivariate sensor, is preferably arranged in an area or section of the exhaust duct , in particular near or close to the exit opening, in which the fan, when in operation, generates a laminar or substantially laminar, airstream through the exhaust duct or within the exhaust channel . Generally speaking, the sensor assembly, preferably the sensing area, may be located in a section of the exhaust duct , in which, during ordinary operation of the cooling fan, the airstream generated by the fan and passing the section represents is substantially uni form or consistent , in particular smooth or laminar . This is ef ficient for accurate VOC detection .

[0057] In preferred embodiments , and as indicated above the multivariate sensor assembly may be mounted near or close to the exit opening . Respective mounting positions may be advantageous with regard to lower temperature load and laminar or smooth flow conditions .

[0058] In embodiments , the multivariate sensor assembly may be mounted such that a first distance between the fan and the sensor assembly, in particular the sensing area, is larger than a second distance measured between the sensor assembly, in particular the sensing area, and the exit opening, the first and second distances respectively measured along a centerline of the exhaust duct or a centerline flowline of the airstream generated by the fan and passing through the exhaust duct . The distances may be measured for example from a center point of the sensor assembly or a side of the sensor assembly or sensing area and a center point of an output port of the fan .

[0059] In embodiments , the first distance may be at least twice , preferably at least three times as large as the second distance .

[0060] In embodiments , the sensor assembly, preferably the sensing area, may be located in a section of the exhaust duct , in which, during ordinary operation of the cooling fan, a velocity, in particular an average velocity, of the airstream generated by the fan through the exhaust duct is at least one of in a range from 15% and 85% , and a range from 30% and 60% of a maximum airstream velocity within the exhaust duct . Respective sections can easily be determined for a given exhaust duct and exhaust duct design based on aerodynamic considerations , in particular measurement , calculations or simulations . The mentioned range ( s ) , are advantageous with regard to avoiding sensor placement in regions leading to biased or unfavourable measurements not representative of the substances (VOC ) or substance concentrations in the cavity . In this connection the airstream velocity shall be understood as the velocity of the airstream, or flow velocity, measured along the exhaust duct . Regarding for example regions involving turbulent flow dynamics or vortexes , while the velocity of the air particles may be high, the ( average ) flow velocity of the airstream is low, under- or over-representing the substance or substance-mix concentration in the cavity . Regarding relatively high flow velocities , the measurements by the sensor assembly may also be faulty and not be representative of the substances actually generated by the food in the cavity .

[0061] In embodiments , the cooling fan may be a radial or centri fugal fan, preferably with curved impeller blades . The fan may comprise an impeller having an axis of rotation that is perpendicular to a wall , preferably an upper wall , of the cavity . In other words , a rotational plane of the impeller may be parallel to this wall .

[0062] In embodiments , a width of the exhaust duct , measured in a plane perpendicular to the axis of rotation ( or parallel to the rotational plane ) , may increase in a funnel-shaped, spiral manner from the output port of the fan towards the exit opening . In embodiments , the exit opening may extend substantially over the whole width, in particular the frontal width, of the cooking appliance measured, for example , in a plane perpendicular to the axis of rotation . The width may for example be related to a frontal width of a control interface covering the service compartment at the frontal side of the appliance . The exit opening may extend over the frontal width between a first end and a second end, which may define lateral ends of the exit opening .

[0063] The sensor assembly, preferably the sensing area, may in embodiments be located in an area in which a curvature of flowlines of the airstream within the exhaust duct and near the exit opening is in a range from 0 to Cmecwherein Cmed represents the average or median of airstream curvatures prevailing over the width of the exit opening measured perpendicularly to the axis of rotation . Such regions have been proven suitable for obtaining sensor signals adequately mirroring the substance ( s ) (VOCs ) or their concentrations in the cavity .

[0064] In embodiments , the fan may be a radial or centri fugal fan with an impeller, with an axis of rotation of the impeller being perpendicular to a wall , preferably an upper wall , of the cavity, wherein a width of the exhaust duct , measured in a plane perpendicular to the axis of rotation, may increase in a funnel-shaped, spiral manner from the output port towards the exit opening . The exit opening preferably extends substantially over the whole width of the appliance , and in a plane perpendicular to the axis of rotation between a first end and a second end . The sensor, in particular the sensing area may be located in a region located near the exit opening and confined between a first and a second fictitious line . The first fictitious line is drawn, following a tangential velocity of a tip of the impeller, from a first point of an outer circumference of the impeller, or of a casing surrounding the impeller, to the first end . The second fictitious line is drawn, following a tangential velocity of a tip of the impeller, from a second point of an outer circumference of the impeller or the casing of the impeller, to the second end . The second point is located downstream the first point relative to the direction of rotation of the impeller, which may be downstream the direction of the airflow generated by the impeller . Providing the sensor within the area defined or confined between the two fictitious lines has been found advantageous for obtaining appropriate measurements regarding the substances (VOCs ) generated by food cooked in the cavity . According to particularly preferred embodiments , the sensor assembly comprises at least one metal-oxide- semiconductor (MOS ) sensor, the MOS sensor adapted to sense one or more organic substances generated or released by food being cooked in the cavity, or more general during the thermal preparation process . The sensor assembly may comprise single ( or separate ) sensor elements , such as MOS sensors , and / or an array of sensor elements , such as MOS sensors , where the sensor elements may represent sensing subunits of the sensor assembly . In particular a sensor unit may comprise an array of two or more MOS sensors . It has been found that MOS sensors are particularly suitable for determining respective substances in an accurate and reliable manner . In particular, MOS sensors are suitable for use and placement in the areas identi fied above , which areas may be considered as optimal for MOS-based sensing regarding flow velocity and / or temperature of the airstream and / or relative amount of substances (VOC ) as compared to the cavity interior .

[0065] In embodiments , the sensor assembly may comprise two or more sensor units respectively configured for detecting at least one organic substance , and wherein the control unit is configured to determine , based on the trained model , the cooking degree , control parameters and / or operating parameters for the appliance from on a response pattern originated from a combination of two or more signals of the sensor units . In this connection, the response pattern may be used as input data for the trained model to generate output data representative of the cooking degree or cooking status of the food item, the control parameters and / or operating parameters . Using a sensor assembly that is sensitive for a plurality of , i . e . two or more , di f ferent (volatile cooking) substances , contributes to obtaining precise and accurate results , without necessitating, for example input data indicating cooking time , food type , weight , or si ze etc . However, user input or additional information from a user may be requested from the user in case that

[0066] In particular, it has been found by testing under real cooking conditions that the cooking degree or cooking status determined based on the sensor signals , in particular the response pattern, and the trained model is in good correspondence and appropriately or optimally mirrors the actual cooking degree or cooking status of the food item in the cavity . These optimal results can in particular be obtained by using MOS sensor technology, e . g . one or more MOS sensors , placed in accordance with the above embodiments in speci fied areas of the exhaust duct .

[0067] In embodiments , the sensor assembly may be attached to an outer wall of the exhaust duct , and air-tightly overlaps or closes an opening in the outer wall , the opening defined in or by the exhaust duct wall section . The sensing area of the sensor assembly attached accordingly faces the interior of the exhaust duct or proj ects into the exhaust duct . The sensing area may, as mentioned, proj ect at least to some extent , into the exhaust duct . It is also possible that the sensing area is substantially flush with the exhaust duct in the area of the exhaust duct wall section, or that the sensing area is , at least slightly, set back relative to the opening plane of the opening . Such an attachment enables a comparative easy and technically simple implementation combined with good or optimal sensing results , in particular cooking status or cooking level predictions or determinations .

[0068] In embodiments , the cavity may comprise a bottom wall , a top wall , three sidewalls connecting the bottom and top walls and a front opening . A front door may be provided for closing the front opening during cooking .

[0069] With such a configuration of the cavity : the service compartment and the cooling fan assembly, which includes the cooling fan and exhaust duct , may be arranged in a section above the top wall ; the exhaust port may pass through the top wall and may be connected to the first input port ; the second input port of the fan may be air- fluidly connected to or open into the service compartment , in particular a space defined by the service compartment .

[0070] Further, with such a configuration : the exhaust duct , in particular the exit opening, may extend substantially over the whole width of a front surface of the cooking appliance ; and the exit opening may open into or at a space between an upper side of a front door configured for closing the front opening and a lower side of a front panel , in particular an upper front panel , e . g . a frontal control panel , covering a frontal side of the service compartment .

[0071] Further, with such a configuration : an optical sensor unit , such as a camera for capturing images or videos of the food, may be placed placed in a way, that it provides an angled view (not directly from the top or the side ) to the food placed in the cavity; the optical sensor unit may be placed at any location at the oven where it has an unobstructed view to the food placed in the cavity; the location for placing or mounting the camera may be at least one of : at , on, or in a door handle of a door to the cavity; in or at the door ; at the cavity, in particular in a frontal or backward left or right inner corner of the cavity, e . g . in or at a back corner of the cavity, in particular where the side walls of the cavity are connected to the top wall ; or in or at one of the inner, upper edges or corners of the cavity .

[0072] Such configurations are advantageous with regard to usage of space and exhaust of the air . In particular exhausting the air at the front side described above between the door and the front panel is advantageous with regard to simple duct routing and obtaining appropriate laminar flow areas for placement of the sensor assembly . Further, such configurations are advantageous with regard to optimal sensor placement and detection . In particular, the mentioned locations for the sensor units are advantageous with regard to reliably detecting volatile compounds in cooking vapors and features of the food based in optical detection using a camera, for example .

[0073] In embodiments , the cooking appliance may comprise an outer housing and / or a chassis , that may least partially surround or cover the cooking cavity unit and service compartment . Such embodiments may be related to built-in appliance types and stand-alone appliance types .

[0074] As noted elsewhere , in embodiments , the sensor assembly comprises the multivariate sensor unit for detecting one or more volatile compounds and at least one optical sensor, in particular an image capturing device and / or a video camera, preferably implemented internally or externally with regard to the cavity of the cooking appliance . Using a combination of volatile compounds and image / video data, i . e . a combination of respective sensors and one or more associated trained models , has been found to be suitable for automating cooking, in particular for fully automated cooking . At least one of the at least one optical sensor may be implemented or integrated with the cooking appliance , wherein the at least one optical sensor may be positioned such that image or video data of the food positioned in or inserted into the cavity may be captured for e . g . detection of the type , shape , si ze etc . of the food .

[0075] Non-limiting embodiments of the present invention will now be described, by way of example , with reference to the accompanying drawings . Same or functionally corresponding elements are referenced with same reference signs . The drawings need not be to scale . In the drawings :

[0076] - FIG . 1 is a cross-sectional view of a cooing appliance , such as a baking oven;

[0077] - FIG . 2 is a 3D sectional view of the cooking appliance ; - FIG. 3 is a detail of the cross-sectional view of FIG. 1;

[0078] - FIG. 4 relates to a first top view of the cooking appliance ;

[0079] - FIG. 5 shows an illustration indicating a flow velocity of an airstream;

[0080] - FIG. 6 relates to a second top view of the cooking appliance ;

[0081] - FIG. 7 shows an example for a sensor of a sensor assembly;

[0082] - FIG. 8 shows a sensitivity diagram of a sensor;

[0083] - FIG. 9A to 9C show further sensitivity diagrams;

[0084] - FIG. 10 and 11 relate to an exemplary embodiment of an impeller for a cooling fan;

[0085] - FIG. 12A to 12C show perspective views of oven components or an oven and indicate possible sites for positioning a camera; and

[0086] - FIG. 13 shows a process diagram of an example of a method for supervising a thermal food preparation process .

[0087] FIG. 1 is a cross-sectional view of a cooking appliance 1, such as a baking oven. The cooking appliance 1 may for example be implemented as a built-in oven. The cooking appliance 1 (referred to in short as appliance 1 below) comprises an oven chassis 2 (frame, scaffold or supporting structure) for supporting components of the appliance and / or for covering parts of the components.

[0088] The appliance 1 comprises a cavity 3 defined by a bottom, top, and three side walls. The cavity 3 is open at one side called the front side. A corresponding opening is closed by a door 4, which is provided for enabling the insertion of food 5 (e.g. a piece of dough) or food items into the cavity 3. The door 4 may be hinged in order to rotate around one of the edges of the front side, preferably a lower horizontal edge. The door 4 covers the front area of the cavity 3.

[0089] The cavity 3 has means for supporting food 5, such as oven shelves for supporting one or more trays for placing food 5 within the cavity 3.

[0090] Heating means 6 are associated with the cavity 3, and are adapted to heat the cavity interior 7. The heating means 6 may include: one or more heating elements (e.g. resistance heating elements) exposed to the cavity interior 7 or located outside of the cavity 3, microwave (MW) heating means, gas burner heating elements, steam cooking elements, but are limited to such heating elements.

[0091] In the example of FIG. 1, the heating means 6 comprises a ring-shaped sheathed heating element.

[0092] Above the door 4 and cavity 3, there is a service compartment 8, which, in the given example, is still contained in the oven chassis 2 . According to the embodiment of FIG . 1 , this service compartment 8 hosts a control unit 9 , which may comprise an electronic board and / or one or more processors or data processing units , able to receive instructions from a user interface provided at a frontal control panel 10 .

[0093] The control unit 9 is adapted and configured for selectively activating the electric loads of the appliance , such as heating means 6 , a fan 11 , a cavity illumination and the like .

[0094] The electric loads are activated at least in part based on algorithms implemented on the control unit 9 , that may utili ze data acquired from signals generated by a ( i . e . at least one ) sensor assembly 12 , which may comprise a sensing area with one or more sensors for sensing temperature , humidity and / or one or more sensors for detecting one or more organic substances as described in further detail above and below . In particular, in one operational mode , the appliance 1 , more speci fically the control unit 9 , may be operable to detect or determine a cooking status or cooking degree of a food 5 placed in the cavity 3 based on sensor signals or a response pattern determined from such signals detected by one or more sensors adapted to sense one or more organic substances generated by the food 5 while being cooked, which is described on more detail below .

[0095] The sensing area or the one or more sensors may include an MOS sensor, which will be described in further detail below, and may further comprise a cavity thermal sensor (not shown) . The sensor assembly and sensing area referred to below in particular relate to sensor ( s ) suitable for detecting one or more organic substances (VOC ) .

[0096] On the control unit 9 , a trained mathematical model may be implemented, as described further above and below, the model for determining of predicting a cooking state or cooking level of the food based on the sensor signals as input parameters for the model . Further, the control unit 9 may, at least in part , automatically control a cooking process based on the trained model , where sensor data of sensors to the appliance 1 are fed to the model , and the model is trained to output data related to cooking state , level of doneness , operating parameters for the appliance , etc . as described herein elsewhere .

[0097] As can be seen from FIG . 1 and FIG . 2 , a front surface of the service compartment 8 is covered with the control panel 10 , which is substantially flush with the front surface of the door 4 .

[0098] The control panel 10 may host , as components of the user interface , a visual feedback means and input means for controlling the oven .

[0099] The cavity walls of the cavity 3 are surrounded by an insulation layer 13 , that has a certain thickness and is configured for thermal insulating the cavity . The cavity 3 , the cavity walls and the thermal insulation 13 may be considered as part of a cooking cavity unit installed in the chassis 2 . The cavity 3 comprises an exhaust port 14, preferably and without limitation located in the cavity' s upper side or top wall, which may be identified as the "roof" of the cavity 3.

[0100] Inside the service compartment 8, there is another space or service space, e.g. a delimited space, which in the present example represents an exhaust duct 15, and which may also be called a "cooling channel". The exhaust duct 15 is defined by a boundary wall made of metal or plastic, wherein the boundary wall encloses an exhaust duct channel represented by the inner volume of the exhaust duct 15.

[0101] In the given example, the boundary wall delimits the exhaust duct channel, wherein the boundary wall, formed e.g. from sheet metal in one piece, may be attached to an outer wall of the chassis 2, wherein this outer wall may define a bottom wall of the exhaust duct 15.

[0102] In the given examples, the boundary wall and the bottom wall also define a fan housing 16 for accommodating an impeller 17 of the fan 11. This housing has the form of a volute. The impeller 17 is connected to a drive motor 18 for driving the impeller 17. In the given example, the fan is 11 implemented as a radial or centrifugal fan 11, in which the axis of rotation A of the impeller 17 is perpendicular to an airstream 19 generated by the fan 11 (at least with regard to the immediate proximity of the impeller 17) . Further, the axis of rotation A is perpendicular to the top wall.

[0103] The exhaust port 14 is in fluid communication with the exhaust port 14 , i . e . it is air- fluidly connected to the exhaust port 14 . The exhaust port 14 is configured for enabling the fan 11 to suck air from the cavity 3 .

[0104] The exhaust duct 15 extends between the fan 11 and an exit opening 20 thereof . The exit opening 20 is arranged in a space between the control panel 10 and the upper side of the door 4 when closed . Following the airstream from the fan 11 to the outlet opening 20 , a height or diameter of the exhaust duct 15 or exhaust channel , relative to vertical cross sections , gradually decreases , in particular near and towards the exit opening 20 . Preferably, the width of the exit opening 20 extends substantially over the whole width of the appliance 1 , such as over the width of the control panel 10 and / or the width of the door 4 .

[0105] The exhaust duct 15 is in air- fluid communication with the fan 11 such that , when operating the fan, the airstream 19 generated by the fan 11 , including air sucked from the cavity 3 through the exhaust port 14 , passes through the exhaust duct 15 and is exhausted at the exit opening 20 .

[0106] In particular, the exhaust port 14 is connected to a first input port 21 associated with the fan 11 ( FIG . 3 ) .

[0107] In the given example , the fan 11 , or in more detail , the fan housing 16 , includes one or more ( fixed) openings providing an air- fluid connection between the impeller 17 and the service compartment 8 . The one or more openings represent a second input port 22 , wherein, in operation, the fan 11 also sucks air from the service compartment 8 via these one or more openings .

[0108] This means , in operation, the fan 11 rotates around the vertical axis A, sucks air in axial direction from the cavity 3 and the service compartment 8 , and expels the mixed air in radial direction through the exhaust duct 15 and finally through the exit opening 20 to the environment . This means , that , in operation, the fan 11 may suck hot air including substances (VOC ) generated by food cooked in the cavity, from below and relative cool air from the service compartment 8 from above , and expel the gas mixture via the exit opening 20 .

[0109] As can be seen in more detail in FIG . 3 , the fan 11 has two fluid inlets , represented in the give example by the first and second input ports 21 , 22 , which are arranged above and below the impeller 17 , respectively . The fluid inlets are located approximately in the area of the axis of rotation A.

[0110] With continued reference to FIG . 3 , an upper side of the exhaust duct , in more detail an upper wall of the exhaust duct 15 that is oriented and located in an area averted from the cavity 3 , includes or defines an exhaust duct wall section 23 . In the exhaust duct wall section 23 , the sensor assembly 12 is provided, the sensor assembly 12 comprising, for example a multivariate sensor, comprising one or more MOS sensors , for detecting one or more organic substances generated by food 5 being cooked or heated in the cavity 3 .

[0111] The exhaust duct 14 , fan 11 , and the exhaust duct wall section 23 that includes the sensor assembly 12 are designed and implemented such that the air mixture M and a corresponding airstream 19 generated by the impeller 17 of the fan 11 by mixing i ) air Al sucked from the cavity 3 and ii ) air A2 sucked from the service compartment 8 and expelled by the fan 11 passes a sensing area 24 of the sensor assembly 12 , which faces the exhaust duct 15 .

[0112] During a cooking process , the food 5 will release a gas mixture , including for example one or more organic substances (VOC ) , that will take the path as described above : it will be sucked up by the fan 11 through the exhaust port 14 and subsequently blown radially by the fan 11 towards the exit opening 20 . Before reaching the exit opening 20 , the air mixture M or airstream 19 will impinge on the sensing area 24 . Accordingly, the sensor assembly 12 , which is configured for detecting one or more corresponding substances , is able to detect respective substances .

[0113] Based on the trained model implemented in the control unit 9, the control unit 9 determines or calculates , based on the sensor signals or a response pattern generated therefrom as input parameters for the trained model , a cooking state or degree of the food 5 as an output parameter . The control unit 9 may, depending on the determined cooking state or degree , operate the loads of the cooking appliance 1 , e . g . the heating means , the fan 11 , a convection fan 26 ( in case that the appliance is a convection oven of includes such operational modes ) , a display of the control panel 10 and other loads . For example , i f the cooking status indicates "unfinished" the control unit 9 may instruct to continue cooking, wherein it is possible for the control unit 9 to adapt cooking parameters ( e . g . heater power etc . , ) depending on the status . Similarly, the control unit 9 may instruct the fan 11 to continue operation, for example with same or di f ferent (higher / lower ) speed . Further, the control unit 9 may indicate the cooking status on a display of the control panel to inform a user accordingly . On the other hand, i f the cooking status indicates " finished" , the control unit 9 may stop cooking or initiate a finishing cooking routine , may instruct the fan 11 to stop or continue for a predetermined residual time ( same seed or di f ferent from previous speed) , and may cause the display to alert the cooking status etc .

[0114] In all , the sensor assembly 12 is arranged in the exhaust duct wall section 23 , with the sensing area 24 of the sensor assembly 12 directly facing the exhaust channel of the exhaust duct 15 such that the airstream 19 and air mixture M pass the sensing area 24 before exiting the exit opening 20 , wherein the air mixture M includes air sucked from the cavity 3 and the service compartment 8 through the inlet ports 21 , 23 .

[0115] The sensor assembly 12 is preferably constituted by an array of least three sensors of MOS type . Each sensor of the sensor assembly 12 or array may be adapted to sense one or more than one (multiple ) of the substances relevant for determining the cooking degree or status . In case that a sensor is able to detect several substances , such a sensor may provide multiple outputs or a single output comprising a combination of single sensor signals , which may be merged into a single output signal .

[0116] According to preferred embodiments , shown for example in connection with FIGs . 1 to 3 , the exhaust duct wall section 23 is tapered towards the exit opening 20 . In particular, the exhaust duct wall section 23 may be arranged in a tapered wall of a funnel-shaped section of the exhaust duct 15 .

[0117] The sensor assembly 12 may be mounted such that a first distance DI between the fan 11 , more precisely an output port 27 ( FIG . 4 ) of the fan 11 and the sensor assembly 12 , in particular the sensing area 24 , is equal to or larger than a second distance D2 measured between the sensor assembly 12 , in particular the sensing area 24 , and the exit opening 20 , the first and second distances DI , D2 respectively measured along a centerline , represented for example by a flowline 28 of the airstream of the exhaust duct 15 . In FIG . 4 a spiral flowline 28 is shown, and distances DI and D2 are sketched accordingly .

[0118] According to embodiments , sensor assembly 12 , preferably the sensing area 24 , is located in a section of the exhaust duct 15 , in which, during ordinary operation of the fan 11 , the airstream 19 generated by the fan 11 and passing the section is substantially uni form or consistent , in particular smooth or laminar . In FIG . 5 flowlines of the airstream 19 are sketched and a corresponding location is indicated by a dashed line defining an area suitable for placement of the sensor assembly 12 . As can be seen from FIG . 4 , flowlines left to the sketched area are not laminar, and may include vortexes or turbulences . Such flows , having a low average flow velocity with regard to the airflow from the fan 11 to the outlet opening 20 , may not mirror or be representative of the substances or relative substance concentrations prevailing within the cavity 3 during cooking .

[0119] According to embodiments , a suitable mounting area for the sensor assembly 12 or sensing area 24 may be such that , during ordinary operation of the cooling fan 11 , a velocity of the airstream generated by the fan 11 through the exhaust duct 15 is at least one of in a range from 15% and 85% , and a range from 30% and 60% of a maximum airstream velocity within the exhaust duct 15 . The area surrounded in FIG . 4 by the dashed line may be representative of respective mounting locations .

[0120] According to embodiments , the mounting area for the sensor assembly 12 may be defined as follows : Given a radial fan 11 as in the above example , with a width of the exhaust duct , measured in a plane perpendicular to the axis of rotation, increasing in a funnel-shaped, spiral manner from the output port 27 of the fan 11 towards the exit opening 20 , and given that the exit opening 20 for example extends substantially over the whole width of the appliance , the sensor assembly 12 or the sensing area 24 , may be located in an area in which a curvature of flowlines of the airstream 19 within the exhaust duct 15 and near the exit opening 20 is in a range from 0 to Cmeci• Cmed represents the average or median of airstream curvatures prevailing over the width of the exit opening 20 measured perpendicularly to the axis of rotation A. Such curvatures for example prevail in the area marked as F any by a dashed line in FIG. 4.

[0121] Expressed in different terms, based on ta top view of the cooking appliance 1, the fan 11, and the exhaust duct 15, when diverting a virtual top viewing plane by two lines which are perpendicular to each other, pass through the axis of rotation A, and are perpendicular to the axis of rotation A into four sections S1...S4, one of the areas, i.e. the area SI in FIG. 4, represents a suitable mounting area for the sensor assembly 12. Two of these sections, i.e. SI and S2 are located in in a frontal section of the virtual plane P, i.e. a section adjacent or near at the front of the cooking appliance 1, wherein section SI defines an area suitable for sensor mounting. The section SI is, considering the length of the flow path or the airstream 19 within the exhaust duct 15 from the output port 27 to the exit opening 20, (in average) closer to the output port 27 as compared to the other frontal section S2. The area of overlap of the upper wall or the exhaust duct 15 and the section SI includes suitable mounting locations for the sensor assembly 12.

[0122] In an embodiment illustrated in connection with FIG. 6, and given that the fan 11 is a radial fan with impeller 17 having an axis of rotation A that is perpendicular to the upper wall the cavity 3, for example, with a width of the exhaust duct 15 , measured in a plane perpendicular to the axis of rotation A increasing in a funnel-shaped, spiral manner from the output port 27 towards the exit opening 20 , the exit opening, preferably extending, substantially over the whole width of the cooking appliance 1 and in a plane perpendicular to the axis of rotation A between a first end El and a second end E2 , a suitable mounting area for the sensor assembly 12 , i . e . a suitable location for the exhaust duct wall section 23 , may be as follows :

[0123] The mounting area is located in a region located near the exit opening 20 and confined between a first and a second fictitious line LI , L2 .

[0124] The first fictitious line LI is drawn, following a tangential velocity Vt ( l ) of a tip of the impeller 17 , from a first point TP1 of an outer circumference of the impeller 17 , or of a casing < f the impeller 17 , to the first end El . The first line LI may be considered as a first tangent to the impelle 17 , or the casing of the impeller 17 .

[0125] The second fictitious line L2 is drawn, following a tangential velocity Vt ( 2 ) of a tip of the impeller 17 , from a second point TP2 of an outer circumference of the impeller 17 , or the casing of the impeller 17 , to the second end E2 . The second line LI may be considered as a second tangent to the impeller 17 , or the casing of the impeller 17 .

[0126] The second point TP2 is located downstream to the first point TP1 relative to the direction of rotation of the impeller 17 or the airstream 19 generated by the impeller 17 . The area confined or defined between the two fictitious lines LI and L2 in a ( fictitious ) plane that is perpendicular to the axis of rotation A is highlighted by a hatching in FIG . 6 .

[0127] A suitable mounting area may be considered as a region of overlap between the hatched area and the exhaust duct 15 , preferably located in a tapered section of the exhaust duct 15 towards the exit opening 20 . This means the to the exhaust duct wall section 23 may be selected to be within this mounting area .

[0128] FIG . 7 shows (not to scale as compared the FIG . 1 to 6 ) an example of a sensor 30 of the sensor assembly 12 . The sensor 30 may be a MOS sensor for detecting one or more substances generated by food during cooking or baking . The sensor 30 may include an outer housing 31 shielding sensing elements accommodated therein, and having one or more " sni f fer" openings 32 penetrating through the housing 31 . The " sni f fer" openings 32 enable gaseous substances included in the airstream 19 passing through for detection by the sensing elements . Mounting locations for the sensor assembly 12 discussed above are particular suitable for such a sensor 30 and enable comparatively exact and reliable detection of respective substances . Further, the mounting locations are advantageous with regard to flow velocity, uni form flow characteristics and temperature load of the sensor ( s ) . Speci fically, the given boundary conditions of the locations for the sensor provide advantageous temperatures regarding the fact that the fan 11 sucks hot air from the cavity 3 , and provide advantageous locations for detecting the substances representative of the substances and relative substance concentrations within the cavity 3 .

[0129] Some MOS (Metal Oxide Semiconductor ) sensors are particularly sensitive to concentration of various types of gases . A variation of resistance of the metal oxide due to adsorption of gases may be measured and used as a signal input by the oven control unit , in particular for input parameters for the trained model ( the trained model described in more detail further above ) . Such sensors may be speci fic to several gas compounds .

[0130] FIG . 8 shows a sensitivity diagram (Rs / RO vs . analyte concentration) of a corresponding sensor, 30 configured and suitable for sensing substances such as air methane , carbon monoxide , lOS-butane , Ethanol , and Hydrogen . Rs / RO is the ratio expressing the relative change in the resistance of the sensor when it is exposed to a target analyte or substance compared to its baseline resistance in the absence of the analyte or substance .

[0131] For obtaining accurate prediction results for the cooking degree or cooking level , several sensors may be used, respectively sensitive to di f ferent substances or analytes .

[0132] FIG . 9A to 9C show sensitivity diagrams (Rs / RO vs . analyte concentration) of three di f ferent sensors , which may be used in combination for the sensor assembly 12 , the sensors sensitive for di f ferent groups of analytes and having different specific sensitivities, e.g. [air, methane, CO, iso-butane, hydrogen, ethanol] , [air, methane, CO, iso-butane, ethanol, hydrogen] , and [air, ethanol, hydrogen, methane, iso-butane, propane] . (The square brackets respectively framing a respective group) .

[0133] Using such different sensors has been proven so result in comparatively exact, reliable and reproducible predictions for food cooked or baked in the cavity 3, in particular without requiring inputting a cooking time and / or a priori knowledge of food type, weight and / or size.

[0134] An exemplary algorithm, based for example on a trained model, may include: acquiring sensor signals of the sensors of a senor arrangement while a cooking process is going on; pre-treatment of the sensor signals, including for example Loess-filtering, derivative processing (such as rate of change, noise reduction, etc.) , and / or adaptive processing (such as adaptive filtering, self-calibration, dynamic range adjustment, adaptive signal processing, etc.) ; extracting features from the pre-processed signals ; pattern recognition (e.g. based on a trained model, such as a trained support vector machine classifier, a supervised machine learning algorithm, a trained neuronal network etc.) ; and predicting a cooking state, cooking degree of the food (e.g. based on a response pattern of the signals) , or determining cooking parameters for adapting operation of the appliance etc..

[0135] In connection with the underlying invention, it has been found that the fan 11, in particular a fan 11 of a cooling system of the cooking appliance 1, is an advantageous, even optimal position for sensors of the types mentioned above. With such sensors (in particular of MOS-type) , the sensor electronics may be damaged by the high temperature found in the oven cavity, if exposed directly to them. The cooling system, preferably comprising a channel with sheet metal walls, or plastic walls, and a cooling fan and fan motor enable sufficient cooling of the hot air sucked from the cavity so as to avoid heat-induced sensor damage.

[0136] It has also been found that radial, in particular centrifugal fans are advantageous for use with the sensor assembly. A corresponding fan may be a double sided centrifugal fan comprising an impeller as shown in FIG. 10 and 11.

[0137] FIG. 10 shows a perspective view of an impeller 17 from a side facing the service compartment 8, and FIG. 11 shows a perspective view of an impeller 18 from a side facing the cavity 3. The impeller 17 comprises a central, substantially circular base 33. A plurality of blades 34 at project from the base 33, wherein the blades 33 are curved in a spiral like manner and extend from a central region of the central base outwards, wherein the central base 33 extends partially into the spaces between respectively adj acent blades 34 , in particular for improving flow dynamics and for improving mechanical stability of the blades 34 .

[0138] Such an impeller 17 is able to suck up a gas mixture both from the upper side and from the lower side , and blowing a corresponding air mixture radially ( to an output port ) relative to a plane perpendicular to the axis of rotation A. The impeller, according to the above examples , is located inside the fan housing 16 . As shown in FIG . 1 to 4 , the fan housing 16 is open near the axis or rotation A enabling the impeller 17 to suck air also from the service compartment 8 , which generates a cooling airflow through the service compartment 8 for cooling electronics or other heat-sensitive components arranged therein, and at the same time is ef ficient for cooling down the temperature of air sucked from the cavity 3 . As indicated above , with a ( cooling) fan assembly as described in connection with the figures , hot air coming from the cavity 3 will mix with cool air service compartment 8 , which is enables an airflow 19 with temperatures below 100 ° C, which is advantageous for the sensor array in terms of reduced temperature loads .

[0139] FIGs . 12A to 12C show perspective views of oven components or an oven, respectively, and indicated possible sites for positioning an image sensor apparatus , in particular a camera, for capturing images of the food 5 , in particular when place inside the cavity 3 . Speci fically, FIG . 12A shows a perspective , sectional frontal door 4 of the cooking appliance 1 . The door 4 comprises an outer door frame 35 , which is , as usual , hingedly connected to a body or frame of the appliance 1 . A glass pane 36 comprising several sheets is attached to the door frame 35 and makes up a transparent frontal window 36 enabling a view to the cavity interior . The door 4 further comprises an outer door handle 37 , which may be attached to the frame 35 and / or to the glass pane 36 . As shown and indicated by arrows in FIG . 12A, the camera may be attached or placed at di f ferent locations associated with the door 4 , for example on or in the handle 37 , on or at the glass pane 36 and / or within the glass pane 36 , for example between two of the sheets of the glass pane 36 . The position of the camera is preferably in an upper section of the door 4 , such that the camera has an angled view into the cavity 3 . As mentioned elsewhere herein, an angled view may be advantageous as compared to a top view with regard to detecting various features of the food, e . g . shape , contour, height , width etc . The location for placing the camera on or in the glass pane instead on or in the door handle , may have the advantage of better protection against outer impacts . However, the camera is preferably placed in a location of the door where a cooling airstream passes by, such that the heat load to the camera caused by cooking processes can be reduced . As the handle 37 and the glass pane 36 are moved away from the cavity 3 in the open position of the door, e . g . when food is entered, cameras attached to the door 4 may not provide images of the food during insertion . In order to be able to capture images also during insertion of the food, further locations for the camera may be considered.

[0140] In this connection, FIG. 12B illustrates a perspective view of the cooking appliance 1, showing a view towards and into the cavity 3. Arrows indicate preferred locations for placing or mounting the camera. For example, the camera may be placed in or at upper, rear corners, and / or in or at upper lateral or backward edges of the inner walls of the cavity 3. In particular, the camera may be located at the intersections of the sidewalls, back wall, and / or top wall of the cavity 3, i.e. at or in locations where respective walls converge. The cameras may be placed such that the cameras can capture an angled view of food placed in the cavity 3.

[0141] FIG. 12C illustrates a perspective view of the cooking appliance 1 from the other lateral side as compared to FIG. 12B, indicating mounting locations for the camera as described previously.

[0142] The cooking appliance 1 may comprise a single camera or several cameras placed at different locations (e.g. handle and back corner (s) ) . Images captured by the cameras may be used as appropriate in connection with supervising a cooking process. Respective images may be used respectively alone and / or may be merged for obtaining image-based information on the food (type, volume etc.) and cooking process (browning, surface texture etc.) .

[0143] FIG. 13 shows a process diagram of an example of a method for supervising a thermal food preparation process. The method comprises : receiving 131 sensor data 135 , the sensor data 135 captured by a sensor arrangement , e . g . sensor 12 , in connection with the thermal preparation process of the food, the sensor arrangement 12 comprises one or more sensor units 3 ) associated with the cooking appliance 1 , and the sensor data 135 comprising sensor data captured by at least one multivariate sensor assembly 12 arranged and configured to detect airborne compounds present in vapors released during the thermal food preparation process ; providing 132 the received sensor data 135 as input data 136 to a trained model 137 , the model 137 trained based on a plurality of training data sets , each providing, for respectively one or more food items , associations between at least one cooking parameter related to the one or more food items , and associated compounds released by the respective one or more food items or generated during the thermal preparation process of the respective one or more food items ; applying 133 the trained model to the input data ; and determining 134 , based on output data 138 obtained by applying the trained model 137 to the input data 136 , one or more cooking parameters 139 of the food 5 undergoing the thermal preparation process ; and generating 135 or updating 135 , based on the determined one or more cooking parameters 139 , one or more of : noti fication data 140 for providing a notification representative of the one or more cooking parameters to a user of the cooking appliance 1, and one or more operating parameters 141 (or: control parameters 141) for controlling one or more subsystems of the cooking appliance 1, wherein the one or more subsystems are at least temporarily involved in the thermal preparation process the food 5.

[0144] The steps may be performed at least in part by the control unit 9 (e.g. data collection etc.) , wherein the trained mode may be implemented on the control unit 9 of the appliance, which control unit may execute the steps related to the model 137 and subsequent steps. In embodiments, the model 137 may be implemented at a remote (e.g. cloud-based) system (not shown in the figures) and execute the steps related to the trained model. Subsequent steps may be performed either by the remote system or by the control unit 9 as appropriate. For example, adapting control parameters or operating parameters of the appliance 1 and generating notifications may be performed by the control unit 9, e.g. at or by the cooking appliance 1.

[0145] As shown in FIG. 3, hot air (Al) from the cavity 3 is sucked by the fan 11 and expelled into the exhaust duct 15. Parallel, cool air A2 coming from the service compartment 8 is also sucked by the fan 11. The resulting air flow 19 is expelled between the door 4 and control panel 10. Since the resulting air temperature will be lower than the temperature in the cavity, but the air mixture flow will still contain the substances or particles released by the food in the cavity while cooking, the areas for placement of the sensor assembly 12 as described above are particularly advantageous for sensor placement .

[0146] Regarding for example MOS sensors , such sensors would be damaged i f directly exposed to hot air within the cavity or sucked from the cavity 3 , due to the high temperature . The suggested fan arrangement is advantageous for avoiding such high temperatures , thereby enabling use of temperature sensitive sensors not suitable for ordinary temperatures in oven cavities .

[0147] Regarding the underlying invention, it is based on the fact that certain types of food, during cooking, release certain substances (Volatile Organic Compounds , VOC ) . Some sensor assemblies , e . g . MOS sensors , are sensitive to concentrations of such substances in a gas mixture . It has been found, that even i f the air coming from the cavity is partially mixed with the air coming from the service compartment 8 , the sensor can still ef fectively detect the presence of certain VOC, in particular i f placed and arranged according to embodiments of the present disclosure .

[0148] Mixing the air from the cavity 3 with the air coming from the service embodiment 8 ensures that the temperature of the airstream 19 rests below the maximum tolerable temperature of the sensor electronics .

[0149] The sensor signal ( s ) may be processed in order to recognize a response pattern in the signal output from the sensor assembly. Such response pattern and / or an associated cooking level or cooking degree may be identified by means of a machine learning model (e.g. SVM classifier) .

[0150] It has been found that a certain response pattern may be indicative of a state of doneness of the food. In particular a corresponding method proves to be effective on foods that develop a crust such as bread or bakery products .

[0151] In an example operation of a cooking appliance,

[0152] • A user may select a cooking program, such as a "bread baking" function

[0153] • The program or function may be started, and, for example, when preheating is completed, the user may put the food, such as dough, inside the cavity 3, close the door, and the cooking cycle may be started.

[0154] • The sensor assembly 12 (e.g. a VOC-sensor-arrangement ) for detecting VOC and / or images may be activated.

[0155] • Responsive to the sensor assembly 12 detecting a certain response pattern, a pop-up may be shown in a user interface (of the oven or of another, e.g. mobile, device including a display) , e.g. "food is ready", or "done" (or similar) .

[0156] • Responsive to that, the cooking function may be then terminated or the heating elements may be activated / deactivated etc. All the above is possible without the need for the user to set the cooking time at the beginning of the function .

[0157] In possible embodiments , after detecting the certain response pattern, the user may be asked i f the cooking time should be extended ( e . g . "add 5 minutes" option) . I f , for example , within a certain time period the user will not select anything, the oven will stop the cooking, avoiding burning of the food .

[0158] Compared to known solutions , the underlying invention provides a solution for cooking status prediction with gas sensors that avoids substantial modi fication to the oven architecture , or the addition of complex features . The sensor positions identi fied and described above and further above have been identi fied through extensive testing . Such positions have been proven optimal or ideal e . g . for an MOS sensor array, because it such positions , in accordance with the present disclosure , are impinged by a constant air flow coming from the cavity and spread by a fan, the airflow, even though diluted, mirroring the gas concentration of the cavity . The air flow is cool enough not to cause any problems to oven and sensor electronics , because it will get cooler along the path and it is mixed with air coming from the service compartment .

[0159] A further advantage compared to known solutions is that the sensor assembly suggested herein is able to detect a doneness level independently from the food si ze or quantity, without requiring, for example , additional sensors such as weight sensors etc . This disclosure and the embodiments described herein in particular relate to cooking appliances, in particular domestic cooking appliances, particularly ovens and cavity-based cookers, aiming to enhance automation and precision in cooking, in particular home cooking.

[0160] Based on the sensor assembly and sensor unit(s) , knowledge of the system is leveraged, by integrating the available information obtained through the sensor units. In particular the sensors for detecting airborne substances, in particular gas sensors, and optical or visual sensors, i.e. computer vision, and also other sensors, may be used for recognizing cooking phases, estimate doneness, and adjust cooking parameters, at least in part, automatically. Thus, consistent cooking results may be obtained, e.g. independently from food colors (camera limitation) or amount of ingredients variety (computational complexity for the electronic nose) . Enhanced cooking precision may be obtained, and real-time monitoring and adaptive control for the cooking, or more general for the thermal food preparation process, is possible .

[0161] The idea described aims to create customer experience improvements in the usage of home appliances (mostly ovens / cavity cooking) , by automatizing the cooking process and recognizing the key stages of the cooking phases and the end of cooking, allowing the customer the possibility to obtain consistently highly appreciated cooking results, with minimal efforts. Through the use of a camera and an electronic / chemical nose (e.g. gas sensor (s) , humidity sensor (s) , temperature sensor(s) etc.) , and / or food sensor(s) (e.g. food probe (s) , single or multipoint) , and / or other sensors disclosed, the cooking appliance or a corresponding control unit is provided access to relevant cooking data that are relevant for the cooking process and allow an, at least in part, preferably fully, automatic recognition of different cooking phases and enable cooking automatization .

[0162] The methods and systems suggested herein enable consistent cooking results by: enabling high-quality cooking outcomes by accurately detecting and responding to cooking phases. Further, user effort may be reduced by: automating the cooking process thereby minimizing the need for user intervention, making cooking easier and more convenient.

[0163] In addition, enhanced cooking precision can be obtained by: Combining for example different sensor types, such as of visual and chemical sensors ad corresponding sensor data, which provides a more comprehensive understanding of the cooking process, leading to better control and precision. Specifically, using an "electronic nose" (e.g. implemented by the sensor unit(s) configured for detecting volatile substances) in (domestic) cooking appliances opens new opportunities for improving cooking experience.

[0164] In particular embodiments, where a camera (optical sensor) is integrated, high-resolution cameras may be used for food recognition and monitoring. An electronic nose as suggested herein may include gas and humidity sensors etc., wherein such a sensor may be further integrated with a temperature sensor in the same device or sensor unit.

[0165] Other information relevant to the cooking may be obtained or retrieved from: food sensors such as core temperature inside the food etc.. Through the use of mathematical models, the temperature at the core of the food can be measured by indirectly interpolating the available information (camera and other sensors or status of the oven, for example "door open" x "time".

[0166] The method and systems enable automated, or at least semiautomated control: The system may adjust cooking parameters (temperature, time, heating levels etc.) based on available information and data (e.g. measured data and / or estimated information, the status of the oven, i.e. activated actuators, status of valves for climatic regulation, door status (e.g., the door open for a certain time would drop the humidity and the temperature in the cavity) . The absorbed power by elements or components of the appliance can also be an information retrieved from the appliance and used for controlling the cooking process .

[0167] Appliances may include a user interface, which may be intuitive and may enable, amongst others, recipe selection and customization.

[0168] The present disclosure also considers connectivity aspects, wherein remote monitoring and control via remote and / or mobile devices, e.g. smartphones etc., and other devices may be provided.

[0169] Reliable in low-visibility conditions: Works even when the cake and tray are both dark and indistinguishable by camera .

[0170] Sensitive to chemical changes: Detects VOCs from Maillard reactions, caramelization, and chocolate aroma compounds.

[0171] Effective in smoke or condensation: Not affected by visual obstructions .

[0172] In embodiments including sensors for detecting volatile compounds or substances (VOC) and image / video sensors, an algorithm may comprise:

[0173] Initialization: preheat, then capture baseline VOC levels and identify a "neutral" signature; Pattern: obtain VOC "profile" vs expected profile (food type may either be pre-selected by the user or automatically recognized by the camera) based on the models;

[0174] Cooking phases: e.g. Maillard reactions (ketones, aldehydes) , caramelization, end of cooking determination and cooking control (e.g. "baking markers") obtained from the trained model .

[0175] For example, if a VOC pattern matches a trained "end-of- bake" fingerprint, "ready" signal may be triggered.

[0176] The electronic nose and camera (and also other sensor types used in combination) may provide synergy effects, wherein :

[0177] Readiness or doneness of food is often not only depending on surface changes but also on changes that happen inside the food - e.g. forming of structure inside of a cake.

[0178] Internal processes can be hard to detect. Camera can track volume changes of food over time, but has poor results with low visible food (e.g. chocolate cake) , whereas an electronic nose may detect information that can be linked to the surface of the food but not the inside state. Data from the electronic nose may support e.g. camera segmentation methods for volume change tracking by giving cues of surface changes happening. Hence the accuracy of camera-based segmentation models may be increased. For example, when certain reaction products are recognized by electronic nose, the segmentation models get additional information to look for areas that changed accordingly.

[0179] List of Reference Signs

[0180] 1 cooking appliance

[0181] 2 chas s i s

[0182] 3 cavity

[0183] 4 door

[0184] 5 food

[0185] 6 heating means

[0186] 7 cavity interior

[0187] 8 service compartment

[0188] 9 control unit

[0189] 10 control panel

[0190] 11 fan

[0191] 12 sensor assembly

[0192] 13 insulation layer

[0193] 14 exhaust port

[0194] 15 exhaust duct

[0195] 16 fan housing

[0196] 17 impeller

[0197] 18 drive motor

[0198] 19 airstream

[0199] 20 exit opening

[0200] 21 first input port

[0201] 22 second input port

[0202] 23 exhaust duct wall section 24 sensing area

[0203] 26 convection fan

[0204] 27 output port

[0205] 28 flowline

[0206] 29 outer circumference of the impeller

[0207] 30 sensor

[0208] 31 outer housing

[0209] 32 sniffer opening

[0210] 33 central base

[0211] 34 blade

[0212] 35 outer frame

[0213] 36 glass pane / window

[0214] 37 door handle

[0215] 131-134 method steps

[0216] 135 sensor data

[0217] 136 input data

[0218] 137 trained model

[0219] 138 output data

[0220] 139 cooking parameter

[0221] 140 notification data

[0222] 141 operating parameter or control parameter

[0223] A axis of rotation

[0224] Al air sucked from the cavity A2 air sucked from the service compartment

[0225] DI first distance

[0226] D2 second distance

[0227] El first end E2 second end

[0228] LI first fictitious line

[0229] L2 second fictitious line

[0230] M air mixture

[0231] P virtual plane SI . . S4 section

[0232] TP1 first point

[0233] TP2 second point

[0234] Vt tangential velocity

Claims

1. Claims1. A method for supervising a thermal preparation process of food (5) , such as baking, cooking roasting, grilling or steaming, performed by a cooking appliance (1) , the method comprising: receiving (131) sensor data (135) , the sensor data (135) captured by a sensor arrangement (12) in connection with the thermal preparation process of the food (5) , the sensor arrangement (12) comprising one or more sensor units (30) associated with the cooking appliance (1) , and the sensor data (135) comprising sensor data (135) captured by at least one multivariate sensor assembly (12) arranged and configured to detect airborne compounds present in vapors released during the thermal food preparation process; providing (132) the received sensor data (135) as input data (136) to a trained model (137) , the model 8137) trained based on a plurality of training data sets, each providing, for respectively one or more food items, associations between at least one cooking parameter related to the one or more food items, and associated compounds released by the respective one or more food items or generated during the thermal preparation process of the respective one or more food items; applying (133) the trained model (137) to the input data (136) ; and determining (134) , based on output data (138) obtained by applying the trained model (137) to the input data (136) , one or more cooking parameters ofthe food (5) undergoing the thermal preparation process; and generating (135) or updating (135) , based on the determined one or more cooking parameters, one or more of: notification data (140) for providing a notification representative of the one or more cooking parameters to a user of the cooking appliance (1) , and one or more operating parameters (141) for controlling one or more subsystems of the cooking appliance, wherein the one or more subsystems are at least temporarily involved in the thermal preparation process the food.

2. The method of claim 1, the sensor arrangement (12) comprising the one or more sensor units (30) further comprises at least one further sensor unit comprising at least one of: one or more optical sensors configured for capturing images and / or video data of the food at least one of prior to and during the thermal preparation process; one or more temperature sensors for detecting temperatures associated with the thermal preparation process of the food (5) , one or more humidity sensors for detecting humidity levels associated with the thermal preparation process of the food, one or more acoustic sensors for detecting cooking sounds generated by the food during the thermal cooking process, one or more ultrasonic sensors for detecting one or more physical parameters of the food (5) , detecting means for detecting an actual power consumption; the method further comprising:receiving further sensor data captured by at least one of the least one further sensor unit , providing the received further sensor data as additional input data to the trained model , applying the trained model also to at least a part or a subset of the additional input data, the model further trained with regard to sensor data associated with the at least one further sensor unit , based on a plurality of further training data sets , each providing, for respectively one or more food items , associations between the at least one cooking parameter related to the one or more food items and sensor data speci fic to a respective further sensor unit .3 . The method of any of claims 1 or 2 , wherein the trained model ( 137 ) comprises ( i ) a uni fied model architecture that j ointly processes the input parameters to generate the output data, or ( ii ) a collection of separate models , each configured to process a subset of the input data and respectively tailored to process a particular type of combination of selected types of sensor data associated with a type of sensor unit or types of sensor units of the one or more sensor units of the sensor arrangement .4 . The method of claim 3 , wherein the trained model is implemented according to ( ii ) , and the method further comprises aggregating, fusing, concatenating, or merging, at least in part , outputs of the separate models to determine or generate at least one of : the one or more cooking parameters , display data, and operating parameters .

5. The method of any of claims 1 to 4, further comprising controlling the thermal preparation process of the food (5) by applying at least one of the one or more operating parameters to at least one of one or more subsystems of the cooking appliance (1) , instructing a user interface (10) associated with the cooking appliance (1) and an associated user to issue a notification representative of at least one of the one more cooking parameters.

6. The method of any of claims 1 to 5, wherein the one or more operating parameters (141) comprise at least one of: one or more power levels or heating levels of one or more heating subunits of the cooking appliance, an estimated remaining cooking time, a predicted doneness endpoint, an operating level of a cooling fan (11) or of a convection fan (26) , an operating level of a vaporizer, an operating level of a vapor extraction system or steam exhaust system.

7. The method of any of claims 1 to 5, wherein the method is carried out, at least in part, by at least one of: a control unit (9) of the cooking appliance (1) and a remote computing device communicatively coupled with the cooking appliance.

8. The method of any of claims 1 to 6, wherein one or more of the sensor units of the sensor arrangement (12) are (i) built-in sensor units or integral sensor units of the cooking appliance, and / or (ii) sensor units external to the cooking appliance (1) and communicatively coupled to the cooking appliance (1)and / or a processing device for processing the sensor data .

9. The method of any of claims 1 to 7, wherein the cooking parameter is one or more of: a type, category or composition of food (5) , a level of doneness, a finishing endpoint, a browning level, an inner food temperature, outer or ambient temperature, a surface temperature of the food (5) , a humidity level, a food weight, a food volume, a food geometry, a food distribution, a food position, a rack position, a crust level, a surface browning, a surface texture, a Maillard level, a cooking stage, a change of rate of sensor data, a change velocity of senor data.

10. The method of any of claims 1 to 8, comprising: initiating a thermal food preparation process; obtaining first sensor data captured by one or more sensor units (30) of the sensor arrangement (12) in a timespan covering a pre-phase and an initial phase of the thermal preparation process ; processing the first sensor data, preferably by feeding the sensor data as input data to the trained model, and determining at least one of: a food property or food characteristic, and sensor base level signals for the sensor units; based on the food-related properties: automatically selecting or suggesting for selection by a user at least one of a food type, food category, and a food composition, and / or correcting a program for thermal food preparation; and based on the sensor base levelsignals, setting respective sensor base levels; performing the thermal food preparation process and continuously or repeatedly performing the steps according to claim 1.

11. The method of claim 1, further comprising, predicting, by applying the trained model to the sensor data, an endpoint for the thermal preparation process of the food, and, if the endpoint is reached, generating a notification to a user and / or stopping or initiating a final phase of the thermal cooking process.

12. A control unit (9) comprising at least one processing unit programmed or configured such that, when operated, the processing unit carries out the steps of the method according to any of claims 1 to 11.

13. A cooking appliance (1) for thermal food preparation of food (5) or a cooking system comprising the cooking appliance (1) for thermal food preparation of food (5) , comprising a sensor arrangement including at least one multivariate sensor assembly (12) configured to detect airborne compounds present in vapors released during the thermal food preparation process from the food (5) , and a control unit according to claim 12.

14. A cooking appliance (1) , in particular according to claim 13, comprising: a cooking cavity unit comprising a cooking cavity (3) having an exhaust port (14) a cooling fan assembly (11, 17, 18, 15) or vapor extraction assembly comprising a vaporextraction fan; and a multivariate sensor assembly (12) ; wherein : the fan (11) is air-fluidly connected to the exhaust port (14) for sucking air from the cooking cavity (3) , and an exhaust duct (15) with one or more walls enclosing an exhaust channel that extends between the fan (11) and an exit opening (20) of the exhaust duct (15) , the exhaust duct (15) being air-fluidly associated with the fan (11) such that, in operation, an airstream (M, 19) generated by the fan (11) and including air sucked from the cooking cavity (3) through the exhaust port (14) passes through the exhaust channel and is exhausted via the exit opening (20) ; the multivariate sensor assembly (12) is arranged in a section (23) of a duct wall of the exhaust channel (23) , with a sensing area (24) of the sensor assembly (12) directly facing the interior of the exhaust channel such that the airstream (M, 19) passes the sensing area (24) before exiting the exit opening (20) .

15. The cooking appliance or system (1) according to claim 13 or 14, wherein the multivariate sensor assembly(12) is configured to detect, in the airstream (M, 19) generated by the cooling fan (11) in the exhaust channel, at least two, three or more organic substances released by the food (5) during the thermal food preparation process into the cavity (3) .

16. The cooking appliance (1) or system according to any of claims 13 to 15, wherein the multivariate sensor assembly (12) is mounted near the exit opening (20) .

17. The cooking appliance (1) or system according to any of claims 13 to 16, wherein the sensor assembly (12) comprises at least one metal-oxide-semiconductor (MOS) sensor .

18. The cooking appliance (1) or system according to any of claims 13 to 17, wherein the multivariate sensor assembly (12) comprises two or more separate sensor units (30) or an array including two or more sensor units, respectively configured for detecting at least one specific organic substance19. The cooking appliance (1) or system (1) according to any of claims 13 to 18, wherein the sensor assembly comprises at least one optical sensor, in particular an image capturing device and / or a video camera, preferably implemented with the cooking appliance.

20. The cooking appliance (1) or system according to claim 19, wherein at least one of the at least one optical sensor is implemented with the cooking appliance, wherein the at least one optical sensor is positioned such that image or video data of the food positioned in or inserted into cavity can be captured, wherein the optical sensor is located at or on a handle of a door of the appliance, or at or on an inner, preferably upper, corner or edge of the cavity.

Images