EP4352654A1 - Computerimplementiertes verfahren und computerprogramm zum maschinellen lernen temporaler zusammenhänge in einem oder mehreren messsignalen und verfahren und computerprogramm zum bestimmen einer messgrösse - Google Patents
Computerimplementiertes verfahren und computerprogramm zum maschinellen lernen temporaler zusammenhänge in einem oder mehreren messsignalen und verfahren und computerprogramm zum bestimmen einer messgrösseInfo
- Publication number
- EP4352654A1 EP4352654A1 EP22729534.2A EP22729534A EP4352654A1 EP 4352654 A1 EP4352654 A1 EP 4352654A1 EP 22729534 A EP22729534 A EP 22729534A EP 4352654 A1 EP4352654 A1 EP 4352654A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- machine learning
- filter
- ann
- learning algorithm
- intx
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01K—MEASURING TEMPERATURE; MEASURING QUANTITY OF HEAT; THERMALLY-SENSITIVE ELEMENTS NOT OTHERWISE PROVIDED FOR
- G01K13/00—Thermometers specially adapted for specific purposes
- G01K13/04—Thermometers specially adapted for specific purposes for measuring temperature of moving solid bodies
- G01K13/08—Thermometers specially adapted for specific purposes for measuring temperature of moving solid bodies in rotary movement
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01K—MEASURING TEMPERATURE; MEASURING QUANTITY OF HEAT; THERMALLY-SENSITIVE ELEMENTS NOT OTHERWISE PROVIDED FOR
- G01K7/00—Measuring temperature based on the use of electric or magnetic elements directly sensitive to heat ; Power supply therefor, e.g. using thermoelectric elements
- G01K7/42—Circuits effecting compensation of thermal inertia; Circuits for predicting the stationary value of a temperature
- G01K7/427—Temperature calculation based on spatial modeling, e.g. spatial inter- or extrapolation
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F16—ENGINEERING ELEMENTS AND UNITS; GENERAL MEASURES FOR PRODUCING AND MAINTAINING EFFECTIVE FUNCTIONING OF MACHINES OR INSTALLATIONS; THERMAL INSULATION IN GENERAL
- F16H—GEARING
- F16H57/00—General details of gearing
- F16H57/01—Monitoring wear or stress of gearing elements, e.g. for triggering maintenance
Definitions
- the invention relates to a computer-implemented method and computer program for machine learning of temporal relationships in one or more measurement signals.
- the invention also relates to a method and computer program for determining a measured variable
- recurrent neural networks require high computing capacities.
- recurrent neural networks require high computing capacities, both during a training phase and during a deployment phase of the trained recurrent neural network.
- Such high computing capacities can only be provided to a limited extent by means of embedded systems, for example microcontrollers, for example automotive control units, also known as electronic control units, ECUs, domain/zone controllers or data concentrator units, DCU.
- microcontrollers for example automotive control units, also known as electronic control units, ECUs, domain/zone controllers or data concentrator units, DCU.
- ECUs electronice control units
- DCU data concentrator units
- the invention provides a computer-implemented method for machine learning of temporal relationships in one or more measurement signals.
- the procedure includes the steps
- the machine learning algorithm predicts at least one measurement variable that correlates with at least one measurement signal.
- This method is a learning or training method.
- the invention provides a computer program for machine learning of temporal relationships in one or more measurement signals.
- the computer program comprises instructions which cause a computer to carry out the steps of the learning method according to the invention when the computer program is run by the computer.
- a machine learning algorithm can be trained with this computer program.
- the invention provides a method for determining at least one measured variable. At least one measurement signal correlating with the at least one measurement variable is converted into a measurement signal according to the invention Learning method learned machine learning algorithm entered and the Maschi nenrivalgorithm outputs the at least one measured variable.
- This method uses a machine learning algorithm that has already been trained.
- the invention provides a computer program for determining a measured variable.
- the computer program includes instructions that cause a microcontroller to perform the steps of the inventive deployment method when the computer program is executed by the microcontroller.
- temporal relationships in data for example measurement signals
- recurrent networks also known as recurrent networks
- temporal relationships in data are learned by machine using artificial neural networks with feedback, also known as recurrent networks, and are determined by a trained recurrent network when used.
- an artificial neural network comprising feedbacks
- outputs of neurons are connected to inputs of at least one neuron of a previous layer.
- the artificial neural network can use direct feedback, in which a neuron’s own output is used as an additional input, and/or lateral feedback, in which a neuron’s output is connected to another neuron in the same layer will include.
- Time series can be processed by means of feedback, for example indirect feedback.
- a recurrent neural network is a feedback neural network comprising direct, indirect and/or lateral feedback.
- the recurrent layers require a relatively large amount of computing and/or storage capacity in both training and use.
- the temporal relationships are shown as a function of transfer functions of filters.
- the temporal relationships can be determined by means of feedback-free machine learning algorithms, that is to say in particular the processing of recurrent layers is eliminated.
- This saves computing and/or storage capacity. For example, only measurement signals and filtered signals are fed into a feedforward network. This means that at least the trained machine learning algorithm can be run on a microcontroller with comparatively limited computing and/or storage capacities.
- a further advantage is that feedback, which is not set in a comprehensible manner in the known prior art, for example in recurrent networks using backpropagation-through-time algorithms, is replaced by the transfer functions of the filters, including frequency responses, impulse responses, step responses.
- the filter output signals contain information from the past.
- a memory based on the physical properties of the filters is thus implemented.
- the learning and application method according to the invention is thus based on data and physical knowledge and is therefore easier to explain and/or understand.
- the determination of the measured variable from the at least one correlating measured signal according to the invention has the further advantage that a sensor for measuring the measured variable can be saved.
- the measured variable is determined automatically, ie by processing a data-driven algorithm, in the application without sensory measurement of the measured variable.
- the at least one measurement signal can be, for example, a measurement signal that is required for many applications.
- a sensor for measuring the measured variable is saved.
- a temperature can be determined without using a temperature sensor for this purpose.
- the machine learning algorithm can be, for example, an artificial neural network, a support vector machine or a random forest classifier. In one aspect, the machine learning algorithm is an artificial neural network.
- the filters can be analog or digital filters.
- the filter is a component comprising at least one resistor and at least one capacitor.
- the filters are implemented in a digital signal processor using software that is processed sequentially.
- the filters are implemented in digital circuits by means of hardware programming and/or parallel data processing, for example field programmable gate arrays, FPGAs for short.
- the commands of the computer program according to the invention include machine commands, source text or object code written in assembly language, an object-oriented programming language, such as C++, or in a procedural programming language, such as C.
- the computer program is a Flardware-independent application program, for example is provided via the data carrier according to the invention or a data carrier signal, according to an aspect by means of software over the air technology, for any flardware, for example via a middleware.
- the computer program is a Flardware-dependent program, for example firmware for the control module according to the invention.
- the data carrier includes volatile memory, for example RAM, DRAM, SRAM, and non-volatile memory, for example ROM, flash EEPROM.
- volatile memory for example RAM, DRAM, SRAM, and non-volatile memory, for example ROM, flash EEPROM.
- non-volatile memory for example ROM, flash EEPROM.
- the data carriers are, for example, flash memory cards, USB sticks. According to one aspect of the inventions, the data carrier is connected to an in/out system of a microcontroller of a driving system and transfers the computer program to the microcontroller.
- the microcontroller for example, is based on several central processors, for example on up to three independent 32-bit CPUs. This meets the highest safety standards and at the same time significantly increases performance.
- the cores of the microcontroller have a clock frequency of 200 MHz, for example.
- the microcontroller includes, for example, a 4 MB flash memory. This means that the microcontroller has reduced complexity, optimized power consumption and saves costs.
- only part of the runtime and/or the memory of the microcontroller is reserved for executing the learned machine learning algorithm. The remaining part is reserved for other applications running on the microcontroller, such as transmission control software.
- the method is performed with a plurality of measurement signals, which are all or partially filtered and fed forward into the machine learning algorithm.
- a number of measurement variables, each of which correlates with one or more of the measurement signals can be predicted and/or determined.
- a measured variable that correlates with a number of the measured signals can be predicted and/or determined more precisely.
- At least one measurement signal is input into a plurality of filters comprising filters with different transfer functions.
- filters comprising filters with different transfer functions.
- the machine learning algorithm is fed forward with the multiple filter outputs.
- the different transfer functions are weighted differently. Due to the different weighting, the machine learning algorithm learns, as a function of the at least one measurement signal, to determine the filter or filters whose transfer function optimizes a memory of the machine learning algorithm.
- the different filters may be filters with different time constants, including, for example, relatively dynamic filters and relatively slow filters.
- At least one measurement signal is input into cascading filters.
- the machine learning algorithm is fed forward with the filter output of the last of the cascading filters and/or with multiple filter outputs of respective filters of the cascading filters.
- filter parameters using known Optimization methods for machine learning algorithms, complex temporal relationships can be mapped with a fraction of the resource requirements of a classic implementation with feedback, such as a recurrent implementation.
- a delay is added to the at least one filter output signal or to a plurality of the filter output signals and the machine learning algorithm is fed forward with the respective delayed filter output signals.
- a delay is added to the filter output signal of the last of the cascading filters and/or a plurality of filter output signals of respective filters of the cascading filters.
- the time delay can be in the millisecond or second range, for example.
- the machine learning algorithm is an artificial neural feedforward network that monitors the prediction of the at least one measured variable and learns from actual data comprising measured values of the at least one measured variable and target data comprising the predicted values of the at least one measured variable.
- the filter serves as a memory for the corresponding layer and can represent temporal relationships according to the selected transfer function, for example the frequency response.
- the supervised learning is based on a gradient based aspect using back propagation.
- the actual data is provided, for example, using at least one reference sensor that measures the at least one measured variable.
- the actual data is also called ground truth.
- the at least one measurement signal or the measurement signals include an engine speed, a transmission speed and/or a torque converter signal.
- the machine learning algorithm predicts an oil temperature as a measured variable, for example the temperature of a transmission oil.
- a special temperature sensor for Measuring the temperature of the transmission oil is thus replaced by the machine learning algorithm, or generally by an artificial intelligence.
- the machine learning algorithm in particular the learned machine learning algorithm, represents a sensor substitute.
- the engine speed, the transmission speed and/or the torque converter signal are filtered and/or weighted using the cascading filters, for example with respective delays.
- the filter or filters comprise at least one PTi element, at least one filter with a finite impulse response and/or at least one filter with an infinite impulse response.
- a PTi element has the transfer function
- the at least one measurement signal is filtered sequentially by cascading PT-i elements.
- a delay is added to the last filter output signal in the cascading PTi elements.
- a finite impulse response filter produces an impulse response of finite length and is characterized by its stability.
- the finite impulse response filter is implemented digitally.
- an infinite impulse response filter can provide an infinitely long lasting impulse response.
- signal distortions are usually modeled as a finite filter and measured accordingly using test signals.
- an infinitely long filter must be used, or a good approximation of it. This means that signal distortion can be better compensated for.
- the infinite impulse response filter is a digital filter.
- an oil temperature for example a temperature of a transmission oil, is determined by means of the application method according to the invention.
- At least one torque converter signal a signal from an electric machine or a starting element, for example if the transmission is constructed without a converter, is input into the machine learning algorithm.
- the machine learning algorithm outputs the oil temperature.
- a temperature sensor that measures the oil temperature is used.
- the learned machine learning algorithm is executed on a microcontroller.
- the learned machine learning algorithm is executed on a microcontroller.
- 3b shows an exemplary embodiment of a further measurement signal and filter output signals for this measurement signal
- 3d shows an exemplary embodiment of a measured variable predicted according to the invention with an associated reference
- 3e shows an exemplary embodiment for deviations between the prediction according to the invention and the reference
- 4 shows an exemplary embodiment of a learning method according to the invention
- FIG. 5 shows an exemplary embodiment of an application method according to the invention.
- the arrangement of FIG. 1 feeds measurement signals In1, In2. InX into a machine learning algorithm ANN.
- the measurement signals In1, In2. InX correspond to one or more measured variables.
- the machine learning algorithm ANN learns to determine the one or more measured variables from the measured signals In1, In2, InX, without these having to be measured by sensors.
- the machine learning algorithm ANN predicts a measured variable Out and outputs it.
- the ANN machine learning algorithm for example, is an artificial neural feedforward network that is taught in a monitored manner.
- the measurement signals In 1 , In2, ... ln(X-1) are fed unfiltered into the machine learning algorithm ANN.
- the measurement signal InX is, without loss of generality, according to the invention in filters F1, F2. F5, comprising, for example, the cascading filters F4,
- the filters differ in their transfer functions.
- the respective filter outputs are fed forward into the machine learning algorithm ANN.
- a delay D1 is added to the filter output signal of the filter F1 and the delayed filter output signal is fed forward into the machine learning algorithm ANN.
- the filter output signal of the filter F1 is an input signal of the cascaded filter F4.
- the filter output signal of the filter F4 is fed forward into the machine learning algorithm ANN without delay and with a delay D2.
- the delay-free filter output signal of the filter F4 is an input signal of the cascaded filter F5.
- FIG. 2 shows a microcontroller pC on which a machine learning algorithm ANN learned according to the learning method according to the invention is used.
- the corresponding deployment method is shown in FIG.
- the microcontroller pC receives, for example, the measurement signal InX as an input signal, which correlates with the measurement variable Out.
- the trained machine learning algorithm ANN determines the measured variable Out without having to measure it using sensors. Since the machine learning algorithm ANN according to the invention does not include any feedback, for example it can be a simple artificial neural feedforward network, it can be executed according to the invention with the limited resources of the microcontroller pC.
- 3a to 3e show, as an application of the solution according to the invention, the determination of a temperature of a gear oil by means of the machine learning algorithm ANN, for example, without the temperature having to be measured with a sensor.
- a temperature sensor is replaced by the machine learning algorithm.
- the plotted values are based on real measurements.
- 3a shows the measurement signals, for example
- n_ICE combustion engine speed
- n_ShaftOut speed of the output gear plotted over time, with one point corresponding to 10ms.
- SH50s_SeqFilt3_TrqConWork sequentially filtered PT1 value (Signal -> Pt1 -> Pt1 ->Pt1 -> ...) + 50s delay input to output;
- the filters that receive the filter output signals Filt1_pwr_TrqConvWork and Filt7_pwr_TrqConvWork are, for example, PTi elements with different transmission functions.
- ICE_TrqAct Torque of the combustion engine with interventions from the gearbox in each case plotted against time.
- the signals shown in FIGS. 3a, 3b, 3c are the input signals of the machine learning algorithm ANN.
- the different filter output signals contain information from the past, with the filter-specific information going back different distances in the past based on the different transfer functions.
- a memory is thus trained, by means of which the machine learning algorithm ANN determines a temporal relationship in the measurement signals and from this predicts the measurement variable, as shown in FIG. 3d, viz
- AI_pred_INT output signal, ie predicted measured variable, for example the temperature of a gear oil
- C_TrsmOil target size, ground truth;
- the sensor that measures this temperature is located in the oil sump and is surrounded by oil;
- C_Chip [°] and C_CPU [°C] respective temperature values, measured with other temperature sensors in the ECU that are not located in the oil sump. These temperature values are used as additional input variables to predict C_TrsmOil. 3d shows a deviation of the prediction from the reference, viz
- Pred_error error between prediction of machine learning algorithm and ground truth
- Pred_error_lnt Error between prediction of machine learning algorithm and ground truth rounded to whole numbers.
- the error ranges between -2°C and +2°C and is acceptable for determining the oil temperature.
- a method step L1 at least one measurement signal Int1, Int2, . . . IntX is input into at least one filter F1, F2, F3 and at least one filter output signal is obtained.
- a feedback-free machine learning algorithm ANN is fed forward with the at least one filter output signal.
- IntX learned as a function of a transfer function of the at least one filter F1, F2, F3.
- the machine learning algorithm ANN based on the learned temporal relationship, predicts at least one measurement variable Out correlating with at least one measurement signal Int1, Int2, IntX.
- step E1 at least one measurement signal Int1, Int2,
- the machine learning algorithm ANN outputs the at least one measured variable Out.
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Control Of Transmission Device (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
Description
Claims
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| DE102021205835.3A DE102021205835A1 (de) | 2021-06-10 | 2021-06-10 | Computerimplementiertes Verfahren und Computerprogramm zum maschinellen Lernen temporaler Zusammenhänge in einem oder mehreren Messsignalen und Verfahren und Computerprogramm zum Bestimmen einer Messgröße |
| PCT/EP2022/063165 WO2022258318A1 (de) | 2021-06-10 | 2022-05-16 | Computerimplementiertes verfahren und computerprogramm zum maschinellen lernen temporaler zusammenhänge in einem oder mehreren messsignalen und verfahren und computerprogramm zum bestimmen einer messgrösse |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| EP4352654A1 true EP4352654A1 (de) | 2024-04-17 |
Family
ID=82020142
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| EP22729534.2A Withdrawn EP4352654A1 (de) | 2021-06-10 | 2022-05-16 | Computerimplementiertes verfahren und computerprogramm zum maschinellen lernen temporaler zusammenhänge in einem oder mehreren messsignalen und verfahren und computerprogramm zum bestimmen einer messgrösse |
Country Status (3)
| Country | Link |
|---|---|
| EP (1) | EP4352654A1 (de) |
| DE (1) | DE102021205835A1 (de) |
| WO (1) | WO2022258318A1 (de) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102023207540A1 (de) * | 2023-08-07 | 2025-02-13 | Zf Friedrichshafen Ag | Verfahren zur Ermittlung einer Ausgabetemperatur eines Prüfsystems |
| DE102024200747B3 (de) | 2024-01-29 | 2025-03-27 | Zf Friedrichshafen Ag | Steuerung eines Schaltgetriebes |
-
2021
- 2021-06-10 DE DE102021205835.3A patent/DE102021205835A1/de active Pending
-
2022
- 2022-05-16 EP EP22729534.2A patent/EP4352654A1/de not_active Withdrawn
- 2022-05-16 WO PCT/EP2022/063165 patent/WO2022258318A1/de not_active Ceased
Non-Patent Citations (1)
| Title |
|---|
| WAN ERIC A: "Finite Impulse Response Neural Networks for Autoregressive Time Series Prediction", 1 December 1992 (1992-12-01), pages 1 - 22, XP093232999, Retrieved from the Internet <URL:https://www.researchgate.net/publication/2677848_Finite_Impulse_Response_Neural_Networks_for_Autoregressive_Time_Series_Prediction> [retrieved on 20250313] * |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2022258318A1 (de) | 2022-12-15 |
| DE102021205835A1 (de) | 2022-12-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3662418B1 (de) | Verfahren und vorrichtung für maschinelles lernen in einer recheneinheit | |
| DE102016011520B4 (de) | Produktionsausrüstung mit Maschinenlernsystem und Montage-und Prüfeinheit | |
| WO2018104280A1 (de) | Verfahren und steuereinrichtung zum steuern eines technischen systems | |
| EP2106576A1 (de) | Verfahren zur rechnergestützten steuerung und/oder regelung eines technischen systems | |
| EP4352654A1 (de) | Computerimplementiertes verfahren und computerprogramm zum maschinellen lernen temporaler zusammenhänge in einem oder mehreren messsignalen und verfahren und computerprogramm zum bestimmen einer messgrösse | |
| WO2014187828A1 (de) | Methode zur ermittlung eines modells einer ausgangsgrösse eines technischen systems | |
| EP3081914A1 (de) | Überwachung einer Maschine mit einer Rotierenden Maschinenkomponente | |
| DE102022120560B4 (de) | Verfahren zur kalibrierung eines drehmomentsensorsystems für ein getriebe | |
| EP4414795B1 (de) | Verfahren und system zum betrieb eines technischen geräts mit einem modell auf basis künstlicher intelligenz | |
| DE102021109169A1 (de) | Verfahren zum Trainieren eines neuronalen Netzes | |
| DE102013206292A1 (de) | Verfahren und Vorrichtung zum Erstellen eines datenbasierten Funktionsmodells | |
| DE102019207342A1 (de) | Konzept für eine Datenverarbeitung für ein zumindest teilautomatisiertes Führen eines Kraftfahrzeugs | |
| DE112020007081T5 (de) | Numerische Steuerungsvorrichtung und Lernvorrichtung | |
| DE102020207449A1 (de) | Verfahren, Computerprogramm und Vorrichtung zum Verarbeiten von Signalen | |
| EP3855390A1 (de) | Verfahren zum erzeugen einer information aus mindestens einem kamerabild | |
| DE102019123447A1 (de) | Verfahren zum Betrieb von Fahrzeugkomponenten | |
| DE102022002242A1 (de) | Virtueller Sensor und Verfahren zum Betrieb eines virtuellen Sensors | |
| DE102023206032A1 (de) | Verfahren und Vorrichtung zum verbesserten Auswerten von Messsignalen eines Sensors | |
| AT527911B1 (de) | Verfahren zum Bestimmen eines Alterungszustands und/oder eines Innenwiderstands einer Batterie | |
| DE102024106117B3 (de) | Computerimplementiertes Verfahren zum Ermitteln einer Steuerfunktion | |
| WO2022161729A1 (de) | Steuereinrichtung zum steuern eines technischen systems und verfahren zum konfigurieren der steuereinrichtung | |
| WO2022078781A1 (de) | COMPUTERIMPLEMENTIERTES VERFAHREN UND VORRICHTUNG ZUR BESTIMMUNG EINES KONFIDENZMAßES EINER MODELLIERTEN PHYSIKALISCHEN GRÖßE EINES VIRTUELLEN SENSORS | |
| EP4506765B1 (de) | Modul zur sequenziellen datenerfassung mit integrierter ki-auswertung | |
| EP4075338B1 (de) | Schneller und energiesparender iterativer betrieb künstlicher neuronaler netzwerke | |
| DE102022205715A1 (de) | Verfahren zum Bereitstellen von Trainingsdaten und Testdaten zum Trainieren eines Algorithmus des maschinellen Lernens |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: UNKNOWN |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE |
|
| PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE |
|
| 17P | Request for examination filed |
Effective date: 20231204 |
|
| AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR |
|
| DAV | Request for validation of the european patent (deleted) | ||
| DAX | Request for extension of the european patent (deleted) | ||
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: EXAMINATION IS IN PROGRESS |
|
| 17Q | First examination report despatched |
Effective date: 20250319 |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE APPLICATION IS DEEMED TO BE WITHDRAWN |
|
| 18D | Application deemed to be withdrawn |
Effective date: 20250722 |