Patent Application
20210298334
This invention relates generally to the food preparation field, and more specifically to a new and useful system and method in the food preparation field.
Typically, during food preparation (e.g., cooking), a user or computing system sets an amount of time to cook the food for. This process can lead to nonideal food preparation such as overcooking foodstuff, undercooking foodstuff, and/or other nonidealities. Thus, there is a need in the food preparation field to create a new and useful system and method. This invention provides such new and useful system and method.
The following description of the preferred embodiments of the invention is not intended to limit the invention to these preferred embodiments, but rather to enable any person skilled in the art to make and use this invention.
As shown in
As shown in
Embodiments of the technology preferably function to estimate when a foodstuff 15 will reach a desired state (e.g., finished temperature, target temperature, target quality, “doneness,” etc.) also referred to as ‘completion’; however, the technology can be used to identify when a given object and/or space has reached a target temperature, and/or can serve any suitable function. The foodstuff is preferably animal based (e.g., meats, eggs, dairy, etc.); however, embodiments of the technology can be used for any suitable foodstuff (e.g., vegetables, carbohydrates, fruits, etc.).
Variations of the technology can confer several benefits and/or advantages.
First, variants of the technology can provide an estimate for time until foodstuff cooking completion. Specific variants can measure the temperature for the foodstuff during cooking and update the estimated amount of time before the food will reach a target temperature. Specific examples can estimate the remaining time to foodstuff completion based on the current foodstuff temperature. In one example, as shown in
Second, variants of the technology can correct for noisy data (e.g., noisy foodstuff parameter data). Specific variants can enable accurate estimates of the remaining time to completion using noisy temperature readings of the foodstuff (e.g., from low quality probes, from sensors moving within the foodstuff, etc.) by leveraging estimation techniques (e.g., a Kalman filter) to estimate the true values for inputs to the time-to-completion determination. Other variants can selectively update the remaining time estimate, and use the wall clock to count down the remaining time between estimate updates (example shown in
Third, variants of the technology can be robust to changes in the temperature of a foodstuff appliance. In specific examples, the technology can generate accurate estimates of the time to completion despite changes to the foodstuff appliance (e.g., opening and closing, adding new foodstuffs to the foodstuff appliance, changing foodstuff appliance settings, etc.).
Fourth, variants of the technology can be used for any suitable foodstuff. In specific examples, the technology can estimate the amount of time until the foodstuff is cooked to a desired state for a variety of different foodstuffs such as beef, pork, fowl (e.g., duck, chicken, turkey, emu, ostrich, goose, pigeon, etc.), mutton, camel, goat, eggs, bison, rabbit, wild game, seafood, vegetables, candy, bakery (e.g., bread, cake, etc.), and/or any suitable foodstuff.
Fifth, variants of the technology can perform processing steps quickly and require minimal storage capacity. In specific examples, the technology can be configured to run efficiently on microprocessors requiring a small amount of storage and a small number of computations to perform one or more calculations. In nonlimiting examples, storage can be minimized by storing tables of coefficients (e.g., for a given curve identifier or temperature curve parameter; for a given set of temperature ranges, etc.) and/or a set of equations, instead of storing complex state spaces, multidimensional curves, and/or neural networks. In nonlimiting examples, processing power can be reduced (and/or calculation speed can be increased) by approximating the polynomial curves as a set of piecewise linear curves; this can reduce the requisite processing power because linear equations can be faster to calculate than polynomial equations.
However, variants of the technology can confer any other suitable benefits and/or advantages.
The system 10 preferably functions to perform the method, but can optionally: heat (e.g., cook) the foodstuff 15, and/or perform any suitable function. For example, the system can estimate when the foodstuff will be finished (e.g., when the foodstuff will reach a target temperature, a target doneness, etc.) and provide an estimated time to completion (ETC) (e.g., notify one or more users when the foodstuff is estimated to finish).
The foodstuff appliance 10 (e.g., cooking system, cooking appliance, foodstuff system, etc.) preferably functions to change the temperature of a foodstuff (e.g., cook foodstuff, reheat foodstuff, etc.). The cooking system can cook the food based on a set of operation instructions (e.g., operation settings), wherein the operation instructions can be: automatically determined, manually determined (e.g., set by a user), and/or otherwise determined or controlled. The cooking system is preferably configured to hold foodstuff (e.g., including on racks, on spits, etc.; with an inner volume that the foodstuff can fit inside; before cooking; during cooking; while the foodstuff is resting; etc.). The cooking system preferably includes one or more heating elements. The heating elements can be arranged above, below, beside, insertable into, a combination of the above locations, and/or in any suitable position relative to at least part of the foodstuff (e.g., a cavity where foodstuff is placed within the cavity). Example heating elements can include: heaters (e.g., resistive heaters), magnetrons, burners, conductive heating elements, inductive heating elements, and/or any other suitable heating element. The cooking system can optionally include one or more cooling elements (e.g., to keep foodstuff cold until the foodstuff is ready to be cooked, to preserve the food after cooking is complete, etc.). The cooking system can apply direct heat (e.g., such as via conduction; heating element is in contact with the foodstuff) and/or indirect heat (e.g., such as via convection, radiation; heating element is not in contact with the foodstuff) to foodstuff. The heat can be constant and/or variable. In specific examples, the cooking system can be an oven, a range (e.g., gas, electric, etc.), a microwave, a pressure cooker, a rotisserie, a fryer, a toaster, a fire, a grill, a sous vide, and/or any suitable cooking system can be used.
The foodstuff appliance is preferably connected (e.g., via a wired or wireless connection) to the computing system, but can additionally or alternatively be connected to the user interface, the sensors, and/or any other suitable component.
The computing system 110 preferably functions to process sensor data to determine an estimated time to foodstuff completion. However, the computing system can additionally or alternatively perform any other suitable function. In variants, the computing system can perform calculations (e.g., determine temperature curves, determine the amount of time until foodstuff is cooked, etc.), transmit information (e.g., to the user such as via the user interface), control the cooking system operation, and/or perform any suitable function. The computing system is preferably integrated into the cooking system (e.g., mounted to a housing shared by the computing system and the cooking system); however, the computing system can be remote (e.g., a server, a cloud, etc.), distributed, and/or arranged in any suitable manner. In variants where the computing system shares a housing with the cooking system, the computing system is preferably mounted to an external surface of the cooking system, but can be mounted to an internal surface of the cooking system and/or otherwise be integrated into the cooking system.
When the computing system is remote from the cooking system, the computing system can additionally receive foodstuff measurements transmitted from the cooking system(s) (e.g., via a wired or wireless connection). The computing system is preferably coupled (e.g., wired, wirelessly) to the user interface. The computing system preferably includes one or more processors and memory; however, the computing system can include any suitable components. Examples of processors that can be used include: microprocessors, ASICs, CPUs, GPUs, TPUs, and/or any other suitable processing system. Examples of memory that can be used include: flash, RAM, or any other suitable digital or analog memory. In specific examples, the microprocessor(s) are configured to efficiently perform multiplication, addition, and search operations; however, the computing system can be configured in any suitable manner.
The computing system is preferably operable to process the sensor data to determine a foodstuff state. The foodstuff state preferably includes a (current and/or estimated) foodstuff parameter and a (current) temperature curve parameter. However, the foodstuff state can additionally or alternatively include a cooking system state, an error estimate for the current foodstuff parameter, an error for the temperature curve parameter, a previous foodstuff parameter, a future foodstuff parameter, a previous temperature curve parameter, a future temperature curve parameter, and/or any suitable information. The foodstuff state is preferably determined using a state estimator. Examples of state estimators can include a Kalman filter, a modified Kalman filter (e.g., a Kalman filter where the state transition model is modified to reflect or based on a temperature curve form), an extended Kalman filter, a particle filter, Monte Carlo, Benes filter, machine learning, neural networks, and/or any suitable state estimator. However, the foodstuff state can be otherwise determined. Inputs to the state estimator can include: a previous foodstuff state (e.g., only an immediately previous foodstuff state, an initial foodstuff state, foodstuff states estimated prior to the current time, etc.), a measured foodstuff parameter, a state-transition model, an observation model, covariance of noise (e.g., process noise, observation noise), and/or any suitable inputs. The computing system preferably only retains the current foodstuff state (e.g., at time t) and an immediately prior foodstuff state (e.g., at time t−1). When a new foodstuff parameter is measured (e.g., at time t+1) and/or a new foodstuff state is determined, the prior foodstuff state can be removed (e.g., deleted from memory, overwritten, etc.) and/or otherwise modified. However, the computing system can retain all prior foodstuff states, a predetermined number of foodstuff states, and/or any suitable foodstuff states.
In some variants, the foodstuff states can be combined (e.g., stitched together) to form a foodstuff parameter profile (e.g., a temperature profile) that relates the foodstuff parameter to time.
The computing system 110 preferably functions to store and/or receive a temperature curve database 115 (e.g., a temperature curve database) including one or more temperature curves that can be used to determine the estimated time to completion. The temperature curves can depend on the foodstuff identity (e.g., a different database of temperature curves for different classes or types of foodstuff, a different subset of temperature curves within a database associated with a different class or type of foodstuff, etc.) and/or be universal (e.g., the same set of temperature curves can be used for all foodstuff, the same set of temperature curves can be used for two or more types or classes of foodstuff, etc.).
The temperature curves (e.g., cooking curves, heating curves, cooling curves, etc.) are preferably a set of one or more curves that define a relationship between foodstuff temperature (e.g., past foodstuff temperature, current foodstuff temperature, future foodstuff temperature, measured foodstuff temperature, estimated foodstuff temperature, etc.) and time (e.g., that foodstuff is cooking, that foodstuff is predicted to cook for, etc.). The method preferably selects a different temperature curve set (set of temperature curves) for each foodstuff type or foodstuff class (e.g., white meat chicken, dark meat chicken, steak, ground beef, etc.), but can additionally or alternatively select the set of temperature curves based on: foodstuff size (e.g., weight, thickness, shape, etc.), foodstuff composition (e.g., fat, muscle, collagen, fiber, bone, etc.), cooking system (e.g., type), cooking system parameters (e.g., temperature; size; fill percentage; thermal properties such as heat capacity, emissivity, etc.; appliance state, such as open, closed, convection, broil, etc.; etc.), initial foodstuff temperature, target foodstuff temperature, reference foodstuff temperature, foodstuff temperature trajectory, foodstuff temperature change over time, foodstuff probe (e.g., location of foodstuff probe such as relative to foodstuff such as center, surface, depth; thermocouple size; etc.), and/or any suitable parameter. The temperature curves can be raw (e.g., unprocessed data such as the temperature profile) and/or processed (e.g., smoothed, fit, denoised, scaled, etc.).
Each temperature curve set preferably includes one or more temperature curves. Each temperature curve of the set of temperature curves is preferably associated with a single temperature curve identifier or parameter (e.g., temperature curve parameter, ‘α,’ such as a different value of α for each temperature curve); however, each temperature curve can correspond to more than one parameter. a is preferably inversely proportional to temperature; however, α can be directly proportional to temperature, independent of temperature, be related to the cooking rate, be related to the foodstuff thickness, be related to the foodstuff heat transmissivity, and/or have any suitable relationship to temperature. During cooking, foodstuff can remain on a single temperature curve, switch between temperature curves, and/or follow more than one temperature curve.
Each temperature curve within the temperature curve set is preferably a piecewise smooth function (e.g., piecewise linear functions, piecewise quadratic, piecewise polynomial, formed from a set of linear functions that are defined over a number of temperature or time intervals of equal or varying size, etc.), but can additionally or alternatively be: piecewise continuous, piecewise functions, linear, nonlinear (e.g., be a polynomial function, an exponential function, a logarithmic function, splines, rational functions, logistic functions, hyperbolic functions, sinusoidal functions, Bessel functions, sigmoid functions, etc.), a combination thereof, and/or otherwise constructed.
In specific variants, each temperature curve can be split into a set of temperature windows (e.g., temperature ranges). The temperature windows (e.g., number, size, starting temperature, etc.) can be determined heuristically, using a machine learning algorithm (e.g., neural networks), using an optimized parameter search, manually, and/or can be chosen in any suitable manner. Each temperature window is preferably independent of the other temperature windows; however, any temperature window can depend on the other temperature windows. The temperature windows are preferably nonoverlapping; however, any temperature windows can overlap other temperature windows. The temperature windows are preferably different for each temperature curve (e.g., within a set, across sets); however, the temperature windows can be shared across temperature curves. In a specific example, the temperature curve can be divided into temperature windows corresponding to temperatures: less than −18° C., −18° C. to 4° C., 4° C. to 25° C., 25° C. to 45° C., 45° C. to 55° C., 55° C. to 60° C., 60° C. to 65° C., 65° C. to 70° C., 70° C. and above. However, any suitable number and ranges for temperature windows can be used. In a specific example, each temperature curve can be a set of linear functions (e.g., piecewise smooth functions) having the functional form: t=CT+DT*(1/α), where t is the time (e.g., time relative to the initial time, instantaneous time, etc.), where CT=a0+a1*T+a2*T2 and DT=a0+a1*T+a2*T2, where T is the temperature (e.g., instantaneous temperature, initial temperature for the temperature window, final temperature within the temperature window, mean temperature for the temperature window, median temperature for the temperature window, temperature relative to the initial temperature, etc.), and a0, a1, a2, b0, b1, b2, are fit parameters associated with a temperature window. However, the temperature curves can be configured in any suitable manner.
The sensor preferably functions to measure foodstuff parameters. The computing system can be connected (e.g., via a wired or wireless connection) to one or more sensors. Each cooking system can be connected to or include one or more sensors. The sensor can optionally be connected to one or more dataloggers that function to store the timeseries of sensor data. The foodstuff parameter is preferably the current foodstuff temperature (e.g., at the center of the foodstuff, at the surface of the foodstuff, etc.); however, the foodstuff parameter can include the foodstuff identity (e.g., type of foodstuff such as animal, cut, variety, etc.), the size of the foodstuff, a time parameter (e.g., how long has the foodstuff been in the cooking system, how long has the foodstuff been cooking, etc.), weight of the foodstuff, foodstuff composition (e.g., amount of fat, bone, collagen, etc.; distribution of fat, bone, collagen, etc.; etc.), foodstuff color (e.g., browning and/or blackening, caramelization, etc.), foodstuff optical characteristics (e.g., an image of the foodstuff), foodstuff noise (e.g., amplitude, frequency, frequency spectrum, etc.), and/or any suitable foodstuff parameter. The sensor is preferably configured to interface with the foodstuff before, during, and after cooking (e.g., inserted into the foodstuff, attached to the foodstuff surface, line of sight to the foodstuff, foodstuff on top of the sensor, etc.), but can interface with the foodstuff at any suitable time(s). The sensor is preferably coupled (either directly or indirectly) to the computing system and, optionally, to the user interface.
The sensor 120 is preferably a foodstuff probe, more preferably a thermometer 125; however, the sensor can additionally or alternatively include one or more of: an optical sensor (e.g., a camera, an infrared camera, infrared temperature probe, Raman temperature probe, etc.), an acoustic sensor (e.g., microphone), a spectroscopic sensor (e.g., colorimetric, gas chromatography, mass spectrometer, etc.), an electronic nose, a pressure sensor (e.g., scale), a timer, a cooking system sensor (e.g., detect cooking system opening and closing, detect amount of time cooking system remains open, etc.), structural sensor (e.g., structured light, etc.), and/or any suitable sensor or set thereof. In a specific example, the sensor can be a thermometer (e.g., thermocouple) that is inserted into the foodstuff (e.g., center of the foodstuff; near a foodstuff feature such as bones, fat, etc.; supported within the foodstuff such that it does not move; etc.). However, the sensor can be a surface thermometer, a cooking cavity thermometer, and/or any suitable sensor. The same sensor is preferably used for different foodstuff; however, a different sensor can be used for each foodstuff (e.g., type of foodstuff, size of foodstuff, foodstuff identity, etc.).
The sensor can sample the foodstuff parameter at predetermined times, at a predetermined frequency, responsive to a trigger (e.g., a request to measure the foodstuff parameter), at random times, at a timing the depends on the foodstuff parameter, with a timing that depends on the target foodstuff parameter, and/or with any suitable timing.
The user interface 130 preferably functions to communicate with a user (e.g., receive inputs such as foodstuff identification, foodstuff preferences such as doneness, target temperature, etc. from a user; send outputs to a user such a foodstuff temperature, estimated time to foodstuff completion, etc.; etc.). The user interface is preferably attached to the cooking system (e.g., on an outside surface of the cooking system), but can additionally or alternatively be connected to (e.g., mounted to, communicably connected to) the computing system, an auxiliary system, a user device (e.g., a smartphone, a tablet, etc.), or otherwise configured. The user interface can include one or more: touchscreens, knobs, dials, displays, buttons, applications (e.g., on a user device such as a smart phone, mobile phone, computer, etc.), speakers, and/or any other suitable user interface. In a first example, the user interface is a touchscreen mounted to a door of the appliance. In a second example, the user interface is an application executing on a user's mobile device (e.g., smartphone). In a third example, the user interface is mechanically and electrically connected to the temperature probe, and can be separate from or integrated within the appliance.
However, the system can additionally or alternatively include any other suitable component or set thereof.
The method 200 preferably functions to estimate the time to completion (ETC) for a foodstuff cooking session (e.g., ETC until foodstuff cooking completion). Completion can be associated with (e.g., determined based on achieving) a target temperature, a target doneness, a target color, a target sound, a target texture, a target time (e.g., a target amount of time that the foodstuff is maintained at, above, or below a threshold temperature), and/or based on any suitable metrics of the foodstuff. Completion can be a predetermined value (e.g., a target temperature based on a foodstuff, a setting as selected by a chef, etc.), a user preference (e.g., a stored user preference such as a preferred doneness for a steak of rare, medium rare, medium, medium well, or well done, etc.), a user selection (e.g., a selection input by the user, example shown in
The method can include identifying a foodstuff S210, changing a foodstuff temperature S220, determining a temperature curve S230, determining the time to foodstuff completion S240, and optionally, transmitting information to a user S250, but can additionally or alternatively include any suitable steps. The method preferably uses a set of temperature curves (e.g., predetermined temperature curves); however, the method can use any other suitable temperature curve generated manually or in any suitable manner. The method and/or steps associated with the method are preferably performed automatically; however, the method can be performed semi-automatically, manually, and/or in any suitable manner. The method (and/or any suitable steps associated with the method) can be performed at: a predetermined frequency (e.g., 1 ms, 10 ms, 100 ms, 1 s, 10 s, 30 s, 1 min, 2 min, 5 min, 10 min, etc.), at predetermined temperature intervals (e.g., 0.1° C., 0.5° C., 1° C., 2° C., 5° C., 10° C., 20° C., 50° C., etc.), in response to new sample measurement receipt, for each measurement timestep, for every N measurements or timesteps, performed with variable frequency, performed once, performed with random timing, and/or performed with any suitable timing.
The estimated time to completion (ETC) is preferably an estimate of the amount of time until the foodstuff is finished cooking (e.g., when the foodstuff will reach a target temperature, when the foodstuff will reach a target temperature range, when the foodstuff will be safe to eat, when the foodstuff will reach a target doneness, etc.); however, the ETC can be a minimum amount of time before the foodstuff will be finished cooking (e.g., 15 min, 30 min, 45 min, 1 hour, 2 hours, etc.), an estimated amount of time to reach a temperature below the target temperature (e.g., to account for resting, to alert the user that the food is nearly ready, based on user statistics such as how close to foodstuff completion they normally remove the foodstuff from the cooking system, etc.), a maximum amount of time to completion, and/or any suitable time. In a specific example, if the estimated time to completion is greater than a threshold duration (e.g., one hour), the estimated time to completion can be “at least one hour;” however, the ETC can be any suitable amount of time.
The ETC can be a specific time (e.g., a wall clock time, such as 08:05, 08:07, 09:11, 09:00, 19:02, 18:58, etc.), a duration (e.g., 1 s, 2 s, 5 s, 10 s, 30 s, 1 min, 2 min, 5 min, 10 min, 15 min, 30 min, 45 min, 1 hour, 2 hours, 4 hours, 8 hours, etc.), an estimated duration (e.g., to the nearest minute, to the nearest second, to the nearest 5 minutes, to the nearest 10 minutes, to the nearest 30 seconds, to the nearest hour, etc.), an estimated time (e.g., 17:00, 17:05, 17:10, 17:15, etc.), and/or any suitable time and/or timespan.
The ETC is preferably based on the target foodstuff parameter (e.g., final temperature); however, the ETC can be based on a target doneness (e.g., blue rare, rare, medium rare, medium, medium well, well done, etc.), target temperature change from an initial foodstuff temperature, texture (e.g., firmness, flakiness, crunchiness, etc.), appearance (e.g., color), weight, mass, past foodstuff parameter(s), cooking system settings, amount of time (e.g., amount of time the foodstuff has been in the cooking system), and/or any suitable parameters (e.g., foodstuff parameters). The accuracy of the ETC preferably improves as the cooking time progresses; however, the accuracy of the ETC can be independent of cooking time, and/or be related to cooking time in any suitable manner. In a specific example, the ETC is the difference in time between the instantaneous time and the predicted time for the food stuff to reach a target temperature; however, the ETC can be determined in any suitable manner. A cooking session can have one or more ETC. In an illustrative example, a cooking session can include an a first ETC associated with an estimated time to interact with the foodstuff (e.g., flip the foodstuff, wrap the foodstuff, season the foodstuff, add one or more ingredients, etc.) and a second ETC associated with an estimated time to foodstuff doneness. However, a cooking session (and/or foodstuff) can have any suitable number of ETCs (e.g., intermediate ETC, final ETC).
The ETC (e.g., the ETC displayed to the user) can be updated when a new ETC is calculated, at a predetermined frequency, when the foodstuff parameter reaches a predetermined foodstuff parameter (e.g., an update temperature threshold, an update time threshold, etc.), according to a timer (e.g., a countdown timer), after a predetermined number of ETC updates, when the ETC changes by a threshold time change, after N ETC estimates within a time window agree, after the next checkpoint in the temperature curve is passed (e.g., next temperature checkpoint), and/or with any suitable timing.
In some embodiments, the method and/or steps thereof can include and/or be performed as disclosed in U.S. patent application Ser. No. 16/380,894 filed 10 Apr. 2019 entitled ‘CONNECTED FOOD PREPARATION SYSTEM AND METHOD OF USE’, which is incorporated in its entirety by this reference. However, the method and/or steps thereof can be performed in any suitable manner.
Identifying the foodstuff S210 preferably functions to determine what foodstuff is present in the cooking system (e.g., meat such as animal, type, cut, etc.; vegetable; bakery; etc.). Identifying the foodstuff preferably occurs before changing the foodstuff temperature; however, identifying the foodstuff can occur at the same time as and/or after changing the foodstuff temperature. Identifying the foodstuff can be performed manually (e.g., by a user at a user interface), semi-automatically (e.g., user confirms foodstuff identity selected by the computing system), and/or automatically (e.g., using computer vision, using one or more sensors, using the temperature measurements, etc.). Identifying the foodstuff can optionally include determining foodstuff parameters (e.g., size, mass, shape, number of pieces of foodstuff, foodstuff orientation such as relative to heating elements, etc.).
In a first example, identifying the foodstuff can be performed as disclosed in U.S. patent application Ser. No. 17/014,932 filed 8 Sep. 2020 and entitled “CONNECTED FOOD PREPARATION SYSTEM AND METHOD OF USE,” which is incorporated in its entirety by this reference.
In a second example, the foodstuff can be identified by classifying an image of the foodstuff.
In a third example, S210 can be performed after S220, S230, S240, and/or at any suitable timing. In these variants, the temperature curve that the foodstuff is following and/or the set of foodstuff curves that the foodstuff temperature is following can be used to identify and/or refine the identification of the foodstuff. For example, after determining the foodstuff state and/or temperature curves a predetermined number of times, a foodstuff identity and/or a probability that the foodstuff belongs to a specific class can be determined.
In a fourth example, S210 can be performed after several iterations of S220, S230 and/or S240. This example can include: comparing the history of temperature measurements for the cooking session against the temperature curves associated with known foodstuff classes and identifying the foodstuff as the foodstuff class with curve(s) that best match the cook session's temperature history. The cook session's temperature history can be matched against the curves by: calculating curve parameters based on the temperature history and finding the most similar set of known curve parameters; comparing the temperature history to a set of reference temperatures for each known curve; and/or otherwise comparing the temperature history against the known curves. The most similar set of known curve parameters can be determined based on a calculated distance measure (e.g., sum of square differences, sum of absolute differences, Minkowski difference, Chebyshev distance, etc.) between the temperature history and the known temperature curves, a classification approach (e.g., machine learning based curve fitting, genetic algorithm, graph-search, neural networks, etc.), and/or otherwise determined.
However, identifying the foodstuff can be performed in any manner.
S210 preferably includes determining target foodstuff parameters (e.g., to determine when the foodstuff reaches completion). The target foodstuff parameters can be selected manually (e.g., be input by a user) and/or automatically (e.g., generated according to recipe instructions; generated based on the foodstuff identity; based on user preferences; based on a label mapping, such as “medium rare” is 130° F. to 135° F.; etc.). The target foodstuff parameters can be set once and/or changed during one or more subsequent steps of the method.
Changing foodstuff temperature S220 functions to heat (and/or cool) one or more foodstuffs to a temperature (e.g., a common temperature, a finished temperature, a target temperature, a doneness temperature, etc.). The target temperature (and/or common temperature) preferably depends on the foodstuff (e.g., foodstuff type, foodstuff identity, etc.); however, the target temperature can be a predetermined value, a selected temperature, can depend on the user (e.g., user preferences), a minimum temperature (e.g., a minimum safe temperature), a maximum temperature, a recommended temperature (e.g., a chef recommended temperature), and/or can be any suitable temperature. Changing foodstuff temperature is preferably performed by a cooking system (e.g., by operating heating and/or cooling elements, by creating an environment with an ambient temperature that differs from the food stuff temperature, etc.); however, changing foodstuff temperature can be performed by a computing system (e.g., a cooking simulation) and/or any suitable system and/or component. Changing a foodstuff temperature S220 preferably occurs before determining a set of temperature curves; however, changing a foodstuff temperature can occur at the same time as and/or after determining a set of temperature curves. Changing a foodstuff temperature S220 preferably occurs before estimating the time to completion; however, changing a foodstuff temperature can occur at the same time as and/or after estimating the remaining cook time.
Changing foodstuff temperature preferably includes cooking the foodstuff; however, changing the foodstuff temperature can include simulating cooking the foodstuff (e.g., using a computing system to estimate the temperature distribution within a foodstuff), cooking a proxy foodstuff (e.g., heating a material that has thermal properties that mimic a foodstuff), and/or any suitable substitute.
The foodstuff is preferably cooked according to a set of operating instructions (e.g., generic for the foodstuff, specific to the foodstuffs characteristics, etc.), but can additionally or alternatively be or otherwise cooked. The operating instructions can be: manually set, automatically set (e.g., by a recipe, set of cooking instructions, etc.), or otherwise determined. In a specific example of automatic cooking system control, the cooking system follows the set of operating instructions, but allows the cooking time to vary (e.g., be controlled by a user, by the method, etc.). In a second example of automatic cooking system control, the recipe can be adjusted (e.g., during the cooking session, before the cooking session, etc.) based on: the altitude, the foodstuff parameters (e.g., thickness, volume, density, starting temperature, marbling, etc.), a target temperature, a target doneness, and/or otherwise adjusted.
Cooking the foodstuff preferably includes maintaining the cooking system at substantially a constant temperature (e.g., cooking system temperature varies by <5° C., <10° C., <20° C., etc.); however, the cooking system can be operated according to any suitable temperature profile (e.g., ramp up temperature, ramp down temperature, hold temperature, etc.), the temperature of the cooking system can vary (e.g., have a variance >5° C., >10° C., >20° C., etc.; the cooking system set temperature can differ from the cooking system true temperature; etc.), the cooking system can be controlled to operate at a predetermined setting (e.g., heat output setting, such as high/low/medium; fuel consumption setting), be uncontrolled, and/or be operated in any suitable manner.
S220 preferably includes measuring foodstuff parameters S225. In some variants, the foodstuff parameters can be used to generate a foodstuff parameter profile for the cooking session (e.g., a temperature profile such as a relationship between foodstuff temperature and time). However, the measured foodstuff parameters can be used independent of each other and/or have any suitable dependence and/or relationship.
The foodstuff parameter profile can be determined for a series of one or more time windows. The duration of each time window is preferably any suitable value and/or range between 1 second to 30 minutes such as 5 minutes; however, the duration of each time window can be any suitable value or range thereof between 1 second to 1 day and/or any suitable duration. The duration of each time window can be longer than, shorter than, or the same length as the previous and/or subsequent time window. For each time window, the foodstuff parameter can be measured at a given frequency (e.g., every 1 second, 5 sec, 10 sec, 30 sec, 1 min, 5 min, 10 min, etc.), at a variable frequency, at a frequency that depends on the computing system (e.g., matched to an operation frequency), at a frequency that depends on an uncertainty (e.g., a temperature uncertainty, a sensor accuracy, etc.), and/or at any suitable frequency. The frequency of measurement in a specific time window can depend on the frequency for other time windows (e.g., previous, subsequent, etc.; such as the same frequency, greater frequency, lesser frequency, etc.), can depend on an uncertainty (e.g., sensor accuracy, estimated statistical error in measurement, etc.), and/or be determined in any suitable manner. However, the foodstuff parameter can be measured at a predetermined frequency, at a variable frequency (e.g., at a frequency that depends on the difference between the current foodstuff parameter and the target foodstuff parameter, at a frequency that depends on the current ETC, at a frequency that depends on the amount of time the food stuff has been cooking, at a frequency that depends on the recipe, at a frequency that depends on the foodstuff, etc.), responsive to a trigger (e.g., when a foodstuff parameter measurement is requested), and/or with any suitable timing.
In specific variants, the measured foodstuff parameters (and/or foodstuff parameter profile) are preferably referenced (e.g., normalized) to a reference foodstuff parameter (e.g., a foodstuff parameter measured or determined at a common time such as at the starting time, at the finishing time, etc.; at a common temperature such as the starting temperature, the finished temperature, a reference temperature of the set of temperature curves, etc.; etc.), wherein the temperature curve determines normalized foodstuff parameter values; however, the measured foodstuff parameters can be raw, scaled, and/or used in any suitable manner. The reference temperature can depend on the measured foodstuff parameter and/or be independent of the measured foodstuff parameter (e.g., depend on the temperature curve set, depend on the foodstuff identifier, etc.). For example as shown in
In a specific example, the measured foodstuff temperature is normalized to the initial foodstuff temperature by dividing the measured foodstuff temperature by the initial foodstuff temperature. In a variation of the first specific example, the measured foodstuff temperature can be normalized to the target foodstuff temperature. In a second specific example, the measured foodstuff parameters can be referenced to the initial temperature by subtracting the initial temperature from each measured foodstuff parameter (e.g., as it is measured such as by taring or calibrating the sensor; digitally subtracting the reference temperature; etc.). In a third specific example that is particularly, but not exclusively, beneficial for foodstuff that is cooked according to a “low and slow” cooking recipe and/or requires long (e.g., greater than 1 hour, greater than 2 hour, greater than 3 hour, greater than 4 hour, greater than 6 hour, etc.) cook times, the measured foodstuff parameter can be referenced to a reference temperature associated with the foodstuff and/or recipe by subtracting a reference temperature associated with the temperature curves from the measured foodstuff parameter. In a fourth specific example, when a measured foodstuff parameter is less than or equal to a threshold foodstuff parameter (e.g., temperature, time, etc.), the measured foodstuff parameter can be referenced to the initial value of the foodstuff parameter. In the fourth specific example, when the measured foodstuff parameter is greater than or equal to the threshold foodstuff parameter, the foodstuff parameter can be referenced to a reference foodstuff parameter such as a reference foodstuff parameter associated with a temperature curve. However, the measured foodstuff parameters can be used raw, and/or processed or referenced in any suitable manner.
In variants, when an error occurs or is detected (e.g., no foodstuff parameter can be measured; temperature of the foodstuff changes in an unexpected manner such as increases faster than expected, decreases quickly, etc.; etc.), an error message can be generated for the user. However, errors can be handled in any suitable manner.
Determining the foodstuff state S230 functions to determine the current state of the foodstuff (e.g., the state of the foodstuff associated with the measured foodstuff parameter). Determining the foodstuff state preferably occurs after measuring current foodstuff parameters; however, determining the foodstuff state can occur at the same time and/or before measuring current foodstuff parameters. Determining the foodstuff state can be performed before, during, and/or after calculating the ETC. The foodstuff state is preferably determined using a state estimator, but can be determined in any manner. The state estimator can be a Kalman filter, an extended Kalman filter, a modified Kalman filter, a particle filter, a Monte Carlo simulator, Benes filter, hidden Markov model, and/or any suitable state estimator. The inputs to the state estimator are preferably the measured foodstuff parameter(s) and/or the previous foodstuff state. For example, as shown in
In a first variant, S230 can include: calculating a rate of temperature change based on at least a first and second sensor measurement; determining an α based on the rate of temperature change (e.g., wherein the α can be calculated based on the rate of temperature change); and selecting the temperature curve associated with the determined α, wherein the selected temperature curve (and/or parameters thereof) can be used to estimate the ETC.
In a second variant, S230 can include: estimating the current α based on at least the measured foodstuff parameter, wherein α can be estimated as a state of the system (e.g., using a state estimator; using a filter, such as a Kalman filter; etc.). The temperature curve (and/or hyperparameters thereof) associated with the estimated current α can then be used to estimate the ETC.
In an example of the second variant, determining the foodstuff temperature curves can include filtering the measured foodstuff parameters. Filtering the measured foodstuff parameters preferably functions to reduce the noise (e.g., measurement noise, readout noise, etc.) in the measured foodstuff parameters and estimate the temperature curve parameter associated with the foodstuff; however, filtering the measured foodstuff parameters can perform any suitable function. The measured foodstuff parameters are preferably filtered in real-time (e.g., as the measurements are recorded); however, the measurements can be filtered at any suitable time. The state estimator is preferably initialized using the measured foodstuff parameter and a temperature curve parameter of 1; however, the state estimator can be initialized randomly, to a manually entered value, to a value determined based on the foodstuff class, and/or using any suitable values.
In a variation of this example, the measured foodstuff parameter is filtered using a Kalman filter; however, any suitable sensor fusion algorithm, smoothing algorithm, and/or filtering method can be used. In a specific example of this variant, a modified Kalman filter can be used. The modified Kalman filter can be defined as: {circumflex over (x)}k=Fk{circumflex over (x)}k−1 where
presents the true (e.g., denoised, current) temperature of the foodstuff, a is a temperature curve parameter associated with the temperature curve the foodstuff is currently estimated to be following, {circumflex over (x)}k−1 represents the previous estimated foodstuff temperature and previous a for the prior timestep,
represents the state transition model with K representing a temperature rise (and/or temperature change rate). In a specific example, K can be determined according to (and/or defined as):
where δ is the time interval between measurements (e.g., frequency of measurements), and a1, a2, b1, and b2 are constants (e.g., global constants, constants specified by the temperature curve associated with at−1, constants calculated based on at−1 and/or t, constants specified by the temperature curve associated with at, constants defined by the functional form of the temperature curves, etc.); however, K can be otherwise determined. In this specific example, each new measurement can be used to update the current temperature and current α; however, each new measurement can be used in any suitable manner. In this specific example, α can be initialized to 1 (e.g., to seed the filter for the initial estimate such as for the first time point); however, α can be initialized to any suitable value. In this specific example, Tcurrent can be initialized to the first temperature measurement, to an ambient temperature value, to 0° C., to a reference temperature, and/or to any other suitable value. In a second specific example, K can be determined according to (and/or defined as):
where f(a;t) is the temperature curve corresponding to temperature curve parameter a, time t, and temperature T. However, K can be otherwise defined and/or the measurements can be filtered in any suitable manner.
In a third specific variant, the temperature curve parameter can be determined by comparing the current foodstuff parameter measurement and a previous foodstuff parameter measurement (e.g., immediately preceding measurement, measurements within a prior time window, etc.). In a specific example, α can be estimated by finding the slope between the current foodstuff parameter and the previous foodstuff parameter; however, α and/or the current foodstuff temperature curve can be determined in any suitable manner.
Determining the foodstuff state preferably includes determining (e.g., selecting) the foodstuff parameter curve S235, which functions to determine the current foodstuff temperature curve for use in ETC calculation. The foodstuff parameter curve is preferably determined based on an analysis α, but can be determined based on the foodstuff identifier or otherwise determined. The analysis α can be the current α, a previous α, an average or median of α from a prior time window or set of measurements, and/or other α.
The foodstuff parameter curve is preferably determined (e.g., retrieved, accessed, etc.) from the foodstuff parameter curve database (and/or memory). The foodstuff parameter curve is preferably a single parameter curve, but can be a set of foodstuff parameter curves for a foodstuff class; a subset of the foodstuff parameter curves (e.g., associated with the analysis α), and/or other curve or set thereof. The foodstuff parameter curve is preferably a temperature curve, but can alternatively be a color curve, a noise curve, size curve, weight curve, and/or any other suitable curve describing a foodstuff parameter change over time.
However, the foodstuff parameter curve can be modelled, derived, and/or otherwise determined and/or selected.
In a specific embodiment, the temperature curve is preferably the temperature curve from the set of temperature curves with a value of α closest to the current α (e.g., as determined in the foodstuff state). The closest α can be the α that has the smallest magnitude difference from the current α, the value that the current α is rounded to (e.g., rounding up, rounding down, standard rounding rules, etc.), the determined value of α (e.g., the current α), and/or any suitable value of α. However, the temperature curve can be derived and/or interpolated between two or more temperature curves (e.g., when α is between temperature curve parameters that are stored in the database), extrapolated (e.g., when α is outside the range of α represented in the temperature curve database), be modelled, and/or otherwise be determined.
In a specific example as shown in
Determining the estimated time to completion S240 preferably functions to estimate the amount of time (e.g., ETC) before the foodstuff reaches a target (e.g., a target temperature, a target mass, a target moisture, target doneness, etc.). Determining the ETC preferably occurs simultaneously with changing the foodstuff temperature; however, determining the ETC can occur before and/or after changing the foodstuff temperature. Determining the ETC is preferably performed by a computing system (e.g., one or more microprocessors of the computing system); however, any suitable system and/or component can be used.
The ETC can be determined (e.g., updated) for each measurement (e.g., each measurement of the foodstuff parameters); however, the ETC can be determined and/or updated for a subset of measurements (e.g., every 5, 10, 20, 30, 100, etc. measurements), at a predetermined frequency (e.g., every 1 s, 5 s, 10 s, 30 s, 60 s, 5 min, 10 min, etc.), once (e.g., at the start of cooking, once the temperature reaches a threshold temperature, etc.), responsive to a request for ETC, when a predetermined number of consecutive a match, and/or be determined at any suitable time.
Determining the ETC can include calculating the ETC based on the temperature curve, the current foodstuff state, the instantaneous time, and/or any suitable data. ETC is preferably calculated by subtracting an estimated completion time from the instantaneous time; however, ETC can be calculated in any suitable manner. The estimated completion time is preferably determined based on the current temperature curve. In a specific example, the foodstuff temperature curve can be determined based on α. The estimated time for the foodstuff to reach the target temperature can be determined by assuming that the foodstuff will stay on the same temperature curve (e.g., ttarget=CTtarget+DTtarget*1/αcurrent). In this example, the ETC is the time at which the foodstuff will reach the target temperature minus the current time (e.g., tcurrent=CTcurrent+DTcurrent*1/αcurrent, wherein Tcurrent and αcurrent are determined by the modified Kalman filter; tcurrent can be the current duration of the cooking session; etc.), such that ETC=ttarget−tcurrent. In a second specific example, the ETC can be determined by extrapolating from the previous foodstuff temperature. In this example, the previous (e.g., previously measured, previously determined foodstuff temperature such as from the modified Kalman filter, etc.) foodstuff temperatures can be fit to a curve (e.g., temperature curve, logistic function, exponential function, piecewise smooth function, etc.). The fit curve can be extrapolated to determine when the foodstuff will reach the completion temperature. However, the ETC can be determined in any suitable manner. In variants, the ETC can be determined in more than one manner. In these variants, the actual ETC can be selected from the ETCs based on a score (e.g., quality) of the ETC. In specific examples, an aggressive (e.g., earliest, shortest), a conservative (e.g., longest, latest), a mean (e.g., average, weighted average), a median, a most common, a highest scored (e.g., highest probability), and/or any suitable ETC can be used as the actual ETC. However, the ETC can be selected in any suitable manner.
Presenting information S250 functions to inform the user of the ETC, of information regarding the processing (e.g., cooking progress) of the foodstuff, of the foodstuff state, and/or otherwise functions. Transmitting information can additionally or alternatively function to transmit information such as errors, foodstuff parameters, cooking system parameters, and/or any suitable information and/or data to the user. Transmitting information to the user can include: transmitting the information to a user device; presenting the information on a user interface (e.g., of the foodstuff appliance); sending a notification to a user device; playing a notification on an audio user device; or otherwise transmitting information associated with the ETC to the user. Presenting the information to the user is preferably performed by the user interface (e.g., of the computing system, a user device, the foodstuff system, etc.); however, any suitable component can be used. Presenting information preferably occurs after determining the ETC; however, transmitting information can occur before and/or during determining the ETC.
The information can include: the ETC (e.g., number of minutes, such as “15 minutes left”); clock time, such as “3:04 pm”), a minimum time until completion (e.g., “more than 1 hour left”), a maximum wait time (e.g., “less than 3 hours left”), errors, foodstuff parameters, foodstuff state, cooking system parameters, and/or any other suitable information. Information can be transmitted every time there is a change in the information (e.g., ETC updates, ETC increases, ETC decreases, etc.), every time information is determined (e.g., ETC is calculated), at a user-specified frequency, at a predetermined frequency (e.g., every 5 min, 10 min, 30 min, etc.), when a predetermined threshold is met (e.g., ETC is 5 min, 10 min, 30 min, etc.; after foodstuff has been in the cooking system for 5 min, 10 min, 30 min, etc.; after foodstuff temperature reaches a threshold value such as 30° C., 40° C., 50° C., 60° C., 70° C., 80° C., 90° C., 100° C., etc.; as shown for example in
In some variants, the information can be updated according to a countdown, which allows the system to continue updating a displayed ETC by counting down according to a unit of time (e.g., seconds, minutes, hours, etc.), even when the displayed ETC is not updated using the determined ETC. In these variants, the information can be continually updated according to the countdown (e.g., a separate clock that continually updates the ETC based on an elapsed time). In an illustrative example, the displayed ETC can be updated with the determined ETC when the update event occurs, and the displayed ETC can be updated based on a wall clock in the interim (e.g., example shown in
In some embodiments, the information, including but not limited to the ETC, presented to the user can be processed. Processing the information can function to decrease the amount of discontinuities within the presented information, decrease the chance that inaccurate information is presented to the user, and/or otherwise improve a user experience with the presented information. Examples of information processing can include: smoothing information, averaging information, detecting outliers and/or inliers, removing outliers and/or inliers, and/or otherwise processing the information. The information can be processed based on information calculated from preceding foodstuff states, based on information from predicted future foodstuff states, based on modelled information, and/or based on any suitable information. In an illustrative example, the ETC can be smoothed and/or averaged with ETCs determined from one or more preceding foodstuff states (e.g., the preceding 1, 2, 3, 5, 10, 20, 30, 50, 100, etc. ETCs; ETC back to the initial ETC estimate; etc.). However, the information can be otherwise processed.
In some embodiments, no information can be presented to a user until the error in the information is within a predetermined threshold and/or a predetermined amount of determined information is in agreement (e.g., three or more pieces of consecutively determined information are the same and/or differ by at most a predetermined threshold). In an illustrative example, the first time the ETC is presented to the user can be after the determined cooking curve has not changed after a predetermined number (e.g., 3, 5, 10, 20, etc.) of foodstuff state calculations. However, approximate information can be presented, the determined information can be presented (e.g., with a confidence interval, without additional details, etc.), and/or any suitable information can be presented at any suitable time.
In a specific example of the method, as shown in
Embodiments of the system and/or method can include every combination and permutation of the various system components and the various method processes, wherein one or more instances of the method and/or processes described herein can be performed asynchronously (e.g., sequentially), concurrently (e.g., in parallel), or in any other suitable order by and/or using one or more instances of the systems, elements, and/or entities described herein.
As a person skilled in the art will recognize from the previous detailed description and from the figures and claims, modifications and changes can be made to the preferred embodiments of the invention without departing from the scope of this invention defined in the following claims.
This application is a divisional of U.S. patent application Ser. No. 17/100,046 filed 20 Nov. 2020, which claims the benefit of U.S. Provisional Application No. 62/937,996, filed 20 Nov. 2019, each of which is incorporated in its entirety by this reference.
| Number | Date | Country | |
|---|---|---|---|
| 62937996 | Nov 2019 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 17100046 | Nov 2020 | US |
| Child | 17345125 | US |
