Voice-controlled system with a liquid handling device and a computer device

Abstract

The invention relates to a system for carrying out voice control of at least one liquid handling device using a computer device and a microphone device, the liquid handling device and the computer device as well as associated methods and computer program codes.

Description

[0001] The invention relates to a system for voice control of at least one liquid handling device using a computer device and a microphone device, the liquid handling device and the computer device, as well as associated methods and computer program codes.

[0002] A liquid handling device is a laboratory instrument used to dispense, transfer, or otherwise manipulate small quantities of liquid with high accuracy and precision. These devices are commonly used in laboratory work across many scientific fields and related laboratories, such as chemistry, biology, and medical research, for tasks like pipetting, diluting, mixing, and dispensing samples and reagents. There are various types of liquid handling devices, ranging from hand pipettes and multi-channel pipettes to automated liquid handling systems capable of handling multiple samples simultaneously, particularly when using a robotic system to manipulate samples and sample containers during complex automated experiments.Programmable liquid handling devices offer one or more of the following functions: programmability, dosing automation, and liquid level sensing, which can improve the efficiency and reproducibility of experiments.

[0003] A programmable pipette is a type of liquid handling device designed to automate pipetting tasks. Programmable pipettes, particularly handheld pipettes, are typically equipped with a data interface and / or a user interface through which the user can define pipetting parameters that determine the pipetting process to be performed automatically by the programmable pipette. Pipetting parameters typically include the volume of liquid to be dispensed, the number of dispensing steps, the rate of aspiration and / or dispensing of the sample into / out of the pipette tip, and other parameters related to mixing and dilution, for example. Programmable pipettes can be used in a variety of applications, including sample preparation, assay development, and high-throughput screening.Furthermore, programmable pipettes can be integrated into laboratory automation systems, enabling improved throughput and enhanced experimental control.

[0004] A known bottleneck in improving the efficiency of handheld programmable pipettes is the limited functionality of the user interfaces of existing pipettes. A user interface is essential for operating the pipette, particularly for setting pipetting parameters. Since the pipette is often handheld, requiring one hand to operate the device, using that same hand for the interface is challenging. As the other hand is frequently needed for other manual tasks during a laboratory experiment, such as holding a sample container, there is a need for an efficient and ergonomic user interface that can be used under such conditions.

[0005] A speech recognition system is used to recognize and interpret spoken language and convert it into text or commands that a computer can understand. Such systems are not to be confused with a "voice recognition system," a technology known for being designed to identify and authenticate a person based on their unique voiceprint. Such speech recognition systems are frequently used in security systems, for example, in voice-activated access controls or biometric identification systems, where the system needs to verify the speaker's identity. In the context of the present invention, both terms "speech recognition" and "voice recognition" are used to describe technologies for converting spoken language into text or commands.

[0006] The use of a speech recognition system to operate laboratory equipment is well known. For example, document EP1438138B1 describes a pipette that incorporates a speech recognition system (VRS) which decodes verbal commands and outputs the result to a central processing unit of the pipette, which in turn generates electronic control signals. To implement such technology, the data processing power of the pipette must be high, resulting in an expensive pipette with relatively high energy consumption. Another disadvantage of a pipette with a VRS, according to the inventors of the present invention, is that an operator tends to turn their mouth towards the pipette's microphone to optimize speech recognition quality.It was found that this increases the risk of contamination of the pipette and the samples treated by the pipette, which is a clear disadvantage when working with biological analytes.

[0007] Speech recognition systems are commonly used in voice-activated smart home devices and in automotive voice control interfaces. Such systems use continuous acoustic monitoring to detect voice commands in a room by constantly listening to and collecting audio data, which raises privacy concerns. While the data is typically stored locally on a device or in the cloud, there have been instances of unauthorized access or compromise. Such technology is unacceptable in research labs where any verbal exchange of highly sensitive information must remain confidential.

[0008] The aim of the present invention is therefore to find a system and suitable components for it in order to implement voice control for at least one liquid handling device, in particular a hand pipetting device, which offers safe and efficient operation.

[0009] The system and its components according to the invention, including the pipetting device and the computer device, combine the ease of use of speech recognition with a secure and field-proven activation function. The microphone device and the activation device (the actuation device) can be positioned in different locations, since the microphone device is not part of the pipetting device, which contains the activation device. Users are thus relieved of the obligation to speak directly into the pipette. Furthermore, integrating the actuation device as part of the pipetting device allows for secure activation of the speech recognition system under manual control by an operator. This eliminates the need for continuous monitoring of voice commands in a room, thereby improving the privacy of spoken information in a laboratory setting.Since the computer running the speech recognition process is not part of the pipetting device, the costs and energy consumption of the pipetting device can be kept low, resulting in longer operating time for the battery-powered pipetting device and an extended lifespan for the device itself. The computer can be a dedicated and optimized device for running the speech recognition process. It can be located on the premises of the pipetting device owner, either in the same room or even in a separate room. This approach improves data privacy compared to a cloud-based solution where speech recognition takes place in the cloud, i.e., on an external server, which may be located on the premises of a service provider offering high-quality speech recognition.However, the use of a cloud-based speech recognition system / method is also possible and preferred for the implementation of the invention.

[0010] In addition to the description of liquid handling devices in further sections of this patent application, a liquid handling device may be a hand pipette. A hand pipette is a portable laboratory instrument designed to be held and operated by hand, enabling users to manually aspirate and dispense precise quantities of liquid. A hand pipetting device may typically comprise one, several, or all of the following components: a pipette body, in particular with a grip area for holding the pipette by hand, and / or with a connection area, in particular a working cone, for attaching a pipetting container, in particular a pipette tip or a dispensing tip, to the pipette or to a pipetting channel of the pipette; a movable rod and / or a piston; a mechanical device for driving the movable rod / piston;an electric motor for moving the piston; a volume adjustment mechanism that can be mechanically and / or electronically controlled; an actuating element, for example a toggle switch, for manually triggering a pipetting process, in particular with a finger of the hand, especially for aspirating or dispensing a liquid sample into / from a pipette or dispenser tip; one or more actuating elements for detecting user input and activating a function of the pipetting device or for enabling user input into the pipetting device; a connecting section, in particular a working cone, for coupling a pipetting container, in particular a pipette tip or a dispenser tip, to the pipetting channel; a tip ejector; an electrically operated user interface, in particular with a display, in particular a touchscreen;A battery and / or secondary battery for powering at least one electrically operated part of the hand pipette, wherein this part is, for example, the electrical control device of the hand pipette, one or more actuating elements for providing user input, or the electric motor. The electrical control device is preferably a programmable device.

[0011] The volume adjustment mechanism of mechanical pipettes is often located near the top of the pipette, allowing users to set and adjust the desired volume of liquid to be aspirated and dispensed. Depending on the specific model of mechanical pipette, this may be a rotary knob or another mechanism. The volume adjustment mechanism of a programmable electronic pipette is typically implemented by programming the pipette's electrical control device, which controls the electric motor and thus the moving rod / piston.

[0012] A hand pipetting device is typically configured to pipette sample volumes of varying sizes, from microliters (µL) to milliliters (mL). They offer a precise and controlled solution for handling liquids in tasks such as pipetting samples, reagents, or buffers in laboratory experiments, sample preparation, and various other applications.

[0013] A liquid handling device can be a robotic pipette, which is typically a programmable pipette. A robotic pipette is an automated system for moving a pipette tip / dispensing tip using one or more electric motors, instead of moving it manually. The robotic pipette is usually operated under program control, with an electronic control device capable of controlling the one or more electric motors. The robotic pipetting device can be configured to automatically move the pipette tip / dispensing tip along at least one or any direction of a Cartesian coordinate system (x, y, z axes). Alternatively or additionally, it can be configured to move one or more sample containers relative to the pipette tip / dispensing tip.A robotic pipetting device can be an automated pipetting system that includes automation features such as robotic arms to perform pipetting tasks. It can process multiple samples simultaneously and is often compatible with other laboratory equipment. A robotic pipetting device can also be a liquid handling workstation, which is a modular system that combines various liquid handling modules such as pipetting units, plate handling devices, and other functions like heating / cooling or shaking. A liquid handling workstation is typically used for complex pipetting workflows.

[0014] The actuating device is configured, in particular, so that a user can activate the voice control input mode, especially by manually actuating the actuating device. The actuating device is also configured, in particular, so that a user can initiate the speech recognition process by causing the electrical control device to issue the start signal, especially by manually actuating the actuating device.

[0015] The actuating device is preferably configured so that it can be activated with a finger of a user's hand, in particular with the same hand holding a hand pipette, which is useful for the preferred embodiment of the hand pipette for one-handed operation. The actuating device is preferably arranged on the housing of a liquid handling device, in particular on the pipette body.

[0016] The actuating device is preferably a dedicated device. A dedicated device refers to a physical device, in particular a button, on a fluid handling device that is specifically assigned or intended for a particular function or feature. Unlike general-purpose buttons, which can be used for multiple functions or actions depending on the context or input mode of the fluid handling device, a dedicated device is intended for a specific purpose and remains consistent in its functionality. A dedicated device, in particular a button, is useful for providing quick access to the frequently used function of a voice control input mode. A dedicated device offers convenience and ease of use, as users can directly trigger a specific function without having to navigate through menus or on-screen controls.

[0017] Preferably, the actuating device is marked, e.g., by a contrasting color, and / or provided with a label or symbol to indicate its assigned function, in particular by a microphone symbol. This enables intuitive learning of how to use the functions of a voice control input mode.

[0018] The actuating device may include a push button. A push button comprises a mechanism in which pressing the button closes an electrical circuit, thereby triggering a specific action or function. The actuating device may include a tactile button. A tactile button provides physical feedback when pressed, giving the user a sense of confirmation. A tactile button typically has a short travel distance and a distinct click or tactile feedback. The actuating device may include a membrane button. A membrane button consists of a thin, flexible membrane, usually marked with a printed symbol or icon. When pressed, the membrane deforms and comes into contact with an underlying conductive layer, thus closing the electrical circuit.

[0019] The actuating device may include a toggle switch. A toggle switch has a tilting mechanism that allows the user to switch between two or three states. The switch tilts back and forth between two states and indicates the current state, preferably visually or through tactile feedback. The actuating device may include a slide switch. A slide switch has a sliding mechanism that moves along a rail or groove to change the position or state of the switch. The actuating device may include a toggle switch. A toggle switch has a lever or a flip mechanism that toggles between two states: on / off or open / closed. The actuating device may include a rotary encoder. A rotary encoder is an input device that allows users to turn a knob or wheel to make inputs.

[0020] An actuating device, in particular a push button or a toggle switch, can be spring-loaded to return to its initial position when the switch is released by the user. Preferably, the electrical control device is configured / programmed such that the voice control input mode remains activated as long as the actuating device remains activated by the user action, and that the voice control input mode is terminated when the actuating device is released by the user, in particular when the spring returns the actuating device to its initial position.

[0021] The actuating device is preferably a touch-sensitive element. The touch-sensitive element can be a dedicated device or implemented by a multi-purpose device and / or a touchscreen, in particular a capacitive sensor screen. Preferably, the electrical control device of the liquid handling device is programmed to display a graphical representation of the actuating device on the touchscreen, wherein the area of ​​the graphical representation is programmed to serve as an input field for detecting at least one touch by a user's finger, in particular for detecting the start, activation, and / or end of the voice control input mode.

[0022] The touch-sensitive element can also be a device that uses one of the following methods: capacitive touch detection, resistive touch detection, inductive touch detection, optical touch detection, acoustic touch detection, piezoelectric touch detection, force sensing resistor (FSR).

[0023] Preferably, the actuating device is arranged on / in the microphone device, in particular attached or integrated, or connected to the microphone device, in particular via a cable connection.

[0024] The actuating device can also be a standalone device that can be wirelessly connected to the computer device. The standalone device can, in particular, have its own power supply, especially a battery, and / or its own microcontroller. For example, the actuating device can be a portable device or be configured to be worn on a hand, for example, by implementation in a glove device, a wristband, or a finger ring. In such an embodiment, a user can hold the liquid handling device in one hand and simultaneously activate the actuating device with that hand or the other hand, even if that hand is already holding the liquid handling device or another object, such as a sample container.The operating device can be configured to be operated with the same hand that wears the wearable device or with the other hand. Furthermore, the wearable device can be configured to be worn on any part of the body where it is to be attached, such as the head, an arm, or a hand. The wearable device can be one of the following: glasses, clothing, a necklace, a belt, a watch, an adhesive bandage for the skin or other surface, or a body implant.

[0025] The actuating device can also be a handheld device, which is different from the liquid handling device.

[0026] Preferably, the electrical control device of the liquid handling device, also referred to as the "electronic control device", is configured and / or programmed to provide the user with a feedback signal to indicate the activation of the voice control input mode.

[0027] Preferably, the system, the microphone device, the computer device and / or the liquid handling device, in particular the hand pipetting device, or any other device of the system includes a loudspeaker; preferably, the feedback signal is provided as an audio signal output by the loudspeaker.

[0028] Preferably, the system, the microphone device, the computer device, and / or the liquid handling device, in particular the hand pipetting device, or any other device of the system, includes a visual output device, for example, a light source; preferably, the feedback signal is provided as a visual signal output by the visual output device. The visual output device may include a display or an output area within the display. The visual output device may include an electrochromic polymer, wherein a change in the voltage applied to the electrochromic polymer alters its optical properties. The visual output device may be integrated into the actuating device, for example, by using a button that functions as a light source.

[0029] The liquid handling device's wireless network adapter is configured for data exchange via a wireless connection. The liquid handling device's wireless network adapter is configured to connect the liquid handling device to a computer's wireless network adapter. The wireless network adapter is hardware, including integrated circuits, configured for data exchange within a wireless network, specifically a WLAN network, particularly a WiFi network, implementing, for example, WiFi STA, WiFi AP, or WiFi STA+AP.The wireless network adapter can be a standalone module that can be connected to a liquid handling device or a computer device, or it can be integrated into a computer device, in particular into a microcontroller of a liquid handling device or computer device. Preferably, a wireless network adapter includes an antenna for transmitting / receiving data by radio data transmission at suitable radio frequencies.

[0030] The wireless network adapter is capable of transmitting the start signal to the computer device via the wireless connection to initiate the speech recognition process and of receiving control data. The control unit of the liquid handling device is programmed, and the wireless network adapter of the liquid handling device is capable of receiving the control data and controlling at least one function of the liquid handling device based on this data. The liquid handling device, or multiple liquid handling devices, can form a wireless network with the computer device and, in particular, the microphone device, which may contain a wireless network adapter.

[0031] A wireless network is a computer network that uses wireless data connections between network nodes. Wireless networks are a way for data processing entities to avoid the costly and often impractical use of cables. Examples of wireless networks include cellular networks, wireless local area networks (WLANs), and ad-hoc networks. A wireless LAN (WLAN) is a wireless computer network that connects two or more computer devices via wireless communication to form a local area network (LAN) within a limited area such as a laboratory, campus, or office building. This allows users to move freely within the area while remaining connected to the network. A WLAN can also connect to the internet via a gateway. Wireless LANs (WLANs) are typically based on the IEEE 802.11 standards for wireless local area networks.These are commonly referred to as Wi-Fi, which is a trademark of the Wi-Fi Alliance. They are also used for home and small office networks that connect laptops, printers, and smartphones to a wireless router, which then connects them to the internet. The wireless network can be provided by an access point device, which has a wireless network adapter and is separate from the computer device, or by the computer device itself.

[0032] The computer device and / or the at least one liquid handling device participating in the same wireless network preferably constitute a so-called service set. In the IEEE 802.11 standards for wireless local area networks (including Wi-Fi), a service set is a group of wireless network devices that share a service set identifier (SSID)—usually the natural language name that users see as the network name. For example, all devices that together form and use a Wi-Fi network called "company-net" are a service set. A service set forms a logical network of nodes that operate with common link-layer network parameters; they form a logical network segment. The service set identifier (SSID) defines a service set or an extended service set. An SSID is preferably customizable.These SSIDs can preferably have a length of zero to 32 octets (32 bytes) and, for the sake of simplicity, are preferably written in a natural language, such as English.

[0033] The computer device can be programmed to provide a wireless network identified by a network identifier. The network identifier can uniquely identify the wireless network among a possible multitude of other wireless networks. The network identifier can be any code suitable for identifying the wireless network.

[0034] The computer device is preferably programmed to automatically connect a liquid handling device to the wireless network using the network identifier and, in particular, a passphrase, without requiring any user interaction to establish the connection. The control device of the liquid handling device is preferably programmed to automatically connect the liquid handling device to the wireless network provided by the computer device using the network identifier and, in particular, a passphrase, without requiring any user interaction to establish the connection. Such a system configuration is suitable for enabling automated exchange of service data between the liquid handling device and the computer device.

[0035] The computer device is preferably programmed to authenticate the at least one liquid handling device whose network identifier is stored in a data memory of the liquid handling device and which connects to the wireless network using the network identifier and in particular using a passphrase to enable the connection of the liquid handling device to the wireless network.

[0036] In the context of the invention, authentication is a process used to verify that a liquid handling device has the right to access the wireless network connected to the computer device. This is preferably a single-factor authentication. With this authentication system, the liquid handling device must know the network identifier and preferably the corresponding passphrase (password) in order to access the wireless network. If the correct network identifier is unknown or the password is incorrect, the liquid handling device is denied access to the wireless network. A password is a secret string of characters known only to the liquid handling device accessing the wireless network and the access point device that verifies this access.

[0037] The network identifier may or may not be announced by the computer device or access point device. In the case of a service set, the computer device or access point device can announce its presence several times per second by sending beacon frames containing the network identifier (SSID). Fluid handling devices can passively discover a computer device or access point device by listening for beacons, or they can send probe frames to actively search for a computer device or access point device with the desired SSID. Once the fluid handling device finds a computer device or access point device with the appropriate name, it can send an associate request frame containing the desired SSID. The computer device or access point device can respond with an associate response frame, also containing the SSID.

[0038] The computer device or access point device can use any authentication type and employ several simultaneously, for example: open authentication to the access point; shared-key authentication to the access point; EAP authentication to the network; MAC address authentication with " " to the network; a combination of MAC-based, EAP, and open authentication; use of CCKM for authenticated clients; use of WPA key management. Preferably, shared-key authentication is used to comply with the IEEE 802.11b standard.

[0039] The computer device or access point device is preferably programmed to enable the transmission of the start signal and control data between the liquid handling device and the computer device / access point device.

[0040] The electronic control device (also referred to as electrical / electronic control device or control device) is part of the liquid handling device and is preferably a programmable device. It may include a data processing device, in particular a computer unit (CPU) for processing data and / or a microcontroller, or the control device itself may be a data processing device. A control device may be part of a microcontroller or include a microcontroller. The control device is configured and programmed, in particular, to control functions of the liquid handling device, especially the automatic or semi-automatic execution of a pipetting process defined by at least one program parameter.The control device is configured and programmed, in particular, to define at least one program parameter or several program parameters using the control data received from the computer device. The liquid handling device is configured so that a user can define at least one program parameter using the voice control input mode and the speech recognition method. The control device may include a data storage device. The data storage device is configured, in particular, for the non-volatile (permanent) storage of data. A data storage device may also be configured for volatile data storage.

[0041] The liquid handling device may include a user interface, preferably a display, preferably a touchscreen, and / or at least one input device for sensing user input, such as a button, knob, switch, contact sensor, or keyboard; electronics and / or electronic circuits, the actuating device. While the actuating device is preferably or exclusively used to start / activate the voice-controlled input mode, the other parts of the user interface may be used to define at least one program parameter or to manually control a pipetting operation (i.e., the aspiration and / or release / dispensing of the sample from a pipetting container attached to the liquid handling device), e.g., to start and / or stop it.

[0042] A computer device is, in particular, a programmable device for the automatic execution of functions that are controllable by (program) software. A computer device generally includes, in particular, a data processing device, especially a processing unit (CPU) for data processing and / or a microprocessor, or the computer device itself may be a data processing device. A computer device may be part of a microcontroller or contain a microcontroller. The computer device may include a data storage device. The data storage device is, in particular, configured for the non-volatile (permanent) storage of data. A data storage device may also be configured for volatile data storage.

[0043] The microphone device is preferably capable of recognizing the user's speech, wherein the user's speech contains information for controlling at least one function of the liquid handling device, wherein the recognition is capable of executing the speech recognition method, which is a program-controlled method and which - records the user's speech using the microphone device, wherein the user's speech contains information for controlling at least one function of the liquid handling device, - recognizes at least one function command from the user language and - Control data depending on which outputs at least one function command.

[0044] The microphone device preferably comprises a data storage and / or data processing device that can be programmed to record and / or store in the data storage a data file containing the sound wave representing the user speech captured by the microphone. However, it is also preferred that the recording of the sound wave is performed by the computer device.

[0045] The microphone device is preferably configured for capturing audio in room environments, especially in a laboratory.

[0046] The microphone device preferably comprises at least one microphone. The microphone can be one of the following: a condenser microphone; a dynamic microphone; a boundary microphone, also known as a PZM (Pressure Zone Microphone); a shotgun microphone; an array microphone consisting of several microphone capsules arranged in an array configuration; a wireless microphone.

[0047] The microphone device can be connected to the wireless network. The microphone device can be connected to the computer device via a wired or wireless connection.

[0048] The approach according to the invention even allows the use of more than one, in particular several, microphone devices, each of which can be located in different positions. This can be used to capture more than one sound wave of the same spoken information.

[0049] Programming the liquid handling device: The electronic control device is programmed to - to detect an actuation of the actuating device caused by a user action, wherein the user action in the case of a pressure-sensitive actuating device includes applying pressure to the actuating device, in the case of a touch-sensitive actuating device (e.g. in the case of capacitive detection) includes touching a sensor field of the actuating device, and in the case of a mechanical actuating device includes moving a movable element of the actuating device. The electronic control device is programmed to - to activate a voice control input mode in response to the detection of actuation, wherein the voice control input mode is a state during which a speech recognition procedure can be activated on the computer device, at least when a sound wave is recorded (in the case of silence or a sound wave that is too short or otherwise faulty, no speech recognition procedure can be activated), to detect a voice input for controlling at least one function of the liquid handling device, and during which the control device is able to receive control data from the remote computer device.

[0050] The term "voice control input mode" refers to a state of the fluid handling device and its control device in which voice recording is possible and in which control data can be received from the computer device and in which the control data can be used by the control device to define a program parameter or part of the program parameter.

[0051] The electronic control device is programmed to - to provide a start signal in response to the detection of the actuation and to send the start signal to the computer device via the wireless connection in order to start the speech recognition process on the computer device.

[0052] The start signal can be start data transmitted over a wireless data connection, or another signal that can be transmitted wirelessly, such as a radio transmission of an analog signal. The start signal, particularly the start data or analog signal, can be uniquely defined so that the computer device can unambiguously receive the information that triggers the speech recognition process and potentially any related processes that support it. The related process might include a program to detect whether the sound wave is damaged, too short, or otherwise defective; a normalization process to adjust the amplitude level of the sound wave or other audio characteristics of the sound wave; and / or a program to combine different sound waves received by multiple microphones to optimize sound quality.

[0053] Programming the computer device: The computer device is programmed to * to receive the start signal via wireless connection, * to be triggered by the start signal in order to start the speech recognition process.

[0054] The speech recognition method is a computer program or a program-controlled method that - records the user's speech using at least one microphone device, wherein the user's speech contains information for controlling at least one function of the liquid handling device, - recognizes at least one function command from the user language and - depending on which at least one function command outputs control data.

[0055] Speech recognition technology: Speech recognition technology, also known as automatic speech recognition (ASR) or voice recognition, is a technology that converts spoken language into written text or commands. The technology is widely known and involves a series of steps to accurately transcribe and understand spoken words.

[0056] The speech recognition process can be programmed to include at least one or all of the following programmed steps: - Audio recording: The process can begin by recording the audio input, which contains the spoken language, as a sound wave using the microphone device. - Preprocessing: The captured audio file can then be preprocessed to remove background noise, normalize the volume, and improve the quality of the speech signal. This step helps to improve the accuracy of the speech recognition system. - Feature extraction: In this step, the preprocessed audio signal is converted into a sequence of acoustic features that represent different aspects of the speech signal. Commonly used features are Mel-frequency-cepstral coefficients (MFCCs) or filter banks. - Acoustic modeling: In acoustic modeling, a statistical model is trained that assigns the extracted acoustic features to phonetic units or sub-word units. This model captures the relationship between the acoustic properties of the language and the corresponding linguistic units. - Language modeling: Language modeling focuses on the statistical analysis of language patterns and probabilities. It helps the speech recognition system understand and predict the most likely word sequence, taking context into account. Language models are created using large text corpora. - Decoding: In the decoding phase, the acoustic and linguistic models are used together to find the most probable word sequence that matches the input language. To do this, the acoustic features are compared with the models, and the most probable word sequence is selected. - Post-processing: After the decoding process is complete, post-processing techniques can be applied to refine the recognized text. This can include grammar checking, spell checking, and language-specific post-processing rules. - Output: The final result of the speech recognition process is the recognized text or commands derived from the spoken input, which form the control data.

[0057] The recognized text can be automatically converted into a suitable command, which can be done using a data table that correlates the recognized text with a desired command. This table can be stored in or connected to a data memory of the computer device. The desired command can be suitable for programmatically calling a specific function on the control device, for example. - a command used by the control device to toggle the screen content displayed on a display of the liquid handling device, and / or a command to navigate through the pages of a user interface of the liquid handling device; - a command used by the control device to start or stop a pipetting operation or to turn off the power supply to the liquid handling device; - a command used by the control device to call a predefined or user-defined and pre-stored method; - a command used by the control device to eject a pipetting tip attached to a working cone of the liquid handling device.

[0058] According to a preferred embodiment, a voice control input mode is activated in response to the detection of actuation. Upon activation of the control input mode, a speech recognition method is activated on a remote computer device to detect / listen to a voice input for controlling at least one function of the liquid handling device. During the detection / listening phase, the control device is able to receive control data from the remote computer device.

[0059] According to a preferred embodiment, the speech recognition method uses a recognition algorithm to recognize at least one function command and to output the control data depending on at least one function command, wherein the recognition algorithm may include solutions such as one or more neural networks, transformer architectures, hidden Markov machines, other probabilistic prediction models, or any combination of the above.

[0060] According to a preferred embodiment, the speech recognition method is programmed to recognize a function command from the group of preselected function commands, including: - Setting a pipetting parameter, which can be one of the following parameters: a pipetting volume, a pipetting volume to be aspirated, a pipetting volume to be dispensed, a number of pipetting steps, a pipetting rate, a number of pre-wetting steps, a pre-wetting volume, a number of mixing steps, ... - Setting a digit, including any digit from 0, 1, ...9 - Setting a number, including multi-digit numbers - Switching a liquid handling device - Ejection of a pipette tip from a working cone of the liquid handling device - Recording a spoken note in an electronic laboratory journal, which may be a program of the computer device or another computer device of the system, to record the user's experiment (where the experiment is defined as the pipetting processes performed by the liquid handling device under the user's control, including the definitions of program parameters, including any sound waves and / or control data and / or the values ​​extracted from the control data to define a program parameter).

[0061] According to a preferred embodiment, a method is provided which includes the computer-implemented steps performed by the computer device to * receive a start signal via wireless connection, * is triggered by the start signal to initiate a speech recognition process, which is a program-controlled process and which - records the user's speech via a microphone device connected to the computer device, wherein the user's speech contains information for controlling at least one function of the liquid handling device, - recognizes at least one function command from the user language and - Control data depending on which outputs at least one function command, * sends the control data to the liquid handling device via a wireless connection in order to control at least one function of the liquid handling device depending on the control data.

[0062] In general, a speech recognition method suitable for use with the present invention can be based on a commercially available software solution. A commercially available speech recognition system with which the speech recognition method mentioned in the claims can be easily implemented can be described as follows: A speech recognition system can be implemented on a computer device using an (Internet) connection between the computer device and an external computing server (cloud-based solution). Novel speech and audio classification systems primarily use deep learning (DL) methods such as recurrent neural networks (RNNs) [Swedia, ER, Mutiara, AB, Subali, M., Ernastuti: Deep Learning Long-Short Term Memory (LSTM) for the recognition of Indonesian speech digits using LPC and MFCC features. In: International Conference on Informatics and Computing (ICIC). pp. 1-5. IEEE (2018)], and convolutional neural networks (CNNs) [Becker, S., Ackermann, M., Lapuschkin, S., Müller, K., Samek, W.: Interpreting and explaining deep neural networks for classification of audio signals. CoRR (2018), http: / / arxiv.org / abs / 1807.03418] or transformer networks [Araabi, A., Monz, C.: Optimizing transformer for low-resource neural machine translation. In: Scott, D., Bel, N., Zong, C. (eds.) Proceedings of the 28th International Conference on Computational Linguistics. pp. 3429-3435. International Committee on Computational Linguistics (2020), https: / / doi.org / 10.18653 / v1 / 2020.coling-main.304] for feature extraction and development. Such DL solutions utilize the computing power of external servers [Devlin, J., Chang, M., Lee, K., Toutanova, K.: BERT: Pre-training of deep bidirectional transformers for language understanding. In: Burstein, J., Doran, C., Solorio, T. (eds.) Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. pp. 4171-4186. Association for Computational Linguistics (2019), https: / / doi.org / 10.18653 / v1 / n19-1423] [Dhar, P.: The carbon impact of artificial intelligence. Nat. Do. Intellectual 2(8), 423-425 (2020), https: / / doi.org / 10.1038 / s42256-020-0219-9].

[0063] However, a speech recognition system / method can also be adapted using a computer device that is not connected to the internet (offline solution). Such offline solutions can be specifically adapted for operation with limited computing power. The document "Implementation of Speech-based Digit-Recognition for Resource-Constrained Environments" (Implementation of speech-based digit recognition for resource-constrained environments), Tim Kraienhorst, 2022, Institute for Microsystems Engineering (www.tu-harburg.de / mst), Eppendorf AG (now: Eppendorf SE), describes a prototype implementation for classifying spoken digits on CPUs, which represents a first step towards implementing a classification of spoken commands for changing device settings.Hardware- and chip-architecture-independent reduction techniques were used to port digit classification to a dual-core Cortex-A7 processor of an STM32MP157C. A spoken digit classification algorithm was proposed using a CNN framework and a MobileNetV3 backbone, along with optimization techniques to increase its robustness and reduce resource requirements. This enabled classification with the ability to scale its capacity via a single parameter α. Furthermore, high accuracy was maintained primarily through the application of efficient quantization and knowledge distillation. Additionally, data magnification was used to improve classification accuracy on the embedded device.This approach reduced the accuracy of the test set by 0.7 percentage points while simultaneously reducing the size of the CNN by more than 34 times to 107.78 kB compared to a standard small MobileNetV3.

[0064] In general, a speech recognition system, also known as an automatic speech recognition (ASR) system, is a technology that uses machine learning or other probabilistic models with natural language processing techniques to recognize spoken language and transcribe it into text. The goal of speech recognition is to accurately convert spoken words and sentences into written text that can then be further processed and analyzed by a computer. Several technical approaches can be used to implement automatic speech recognition systems, such as Hidden Markov Models (HMMs), Deep Neural Networks (DNNs), Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Gaussian Mixture Models (GMMs).

[0065] There are several commercial products that use one or more of the aforementioned speech recognition techniques. Such products can be used to implement speech recognition algorithms in the computer device of the present invention. Examples of commercially available products include: Kaldi (https: / / kaldi-asr.org) is an open-source speech recognition toolkit that allows users to create custom speech recognition models. Written in C++ and licensed under the Apache License v2.0, Kaldi includes a variety of acoustic and speech models that can be used to recognize speech in various languages, as well as tools for creating and training new models. - TensorFlow Speech Recognition (https: / / www.tensorflow.org): This is an open-source project that provides a set of pre-trained models for speech recognition using TensorFlow. It includes models for recognizing speech in a wide variety of languages, as well as tools for building and training new models. TensorFlow can even build speech recognition systems that run on relatively low-powered computing devices, such as a Raspberry Pi.

[0066] TensorFlow Speech Recognition uses a combination of Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) to extract features from the audio input and perform sequence modeling to recognize spoken words.

[0067] The basic pipeline of TensorFlow speech recognition includes the following steps: 1. Audio preprocessing: The raw audio input is preprocessed to extract features relevant for speech recognition, such as the spectrogram and Mel frequency-cepstral coefficients (MFCCs). These features are then normalized to improve the performance of the neural network. 2. Convolutional Neural Network (CNN): The preprocessed audio features are fed into a CNN, which learns to extract higher-level features relevant for speech recognition. The CNN output is a sequence of feature maps that capture the temporal and spectral structure of the audio. 3. Recurrent Neural Network (RNN): The sequence of feature maps from the CNN is fed into an RNN, for example, a Long Short-Term Memory (LSTM) network, which learns to model the temporal dynamics of the speech signal. The output of the RNN is a sequence of hidden states that capture the context and history of the speech input. 4. Fully connected layers: The output of the RNN is then fed into one or more fully connected layers that learn to map the hidden states to a sequence of output probabilities via the vocabulary of words. 5. Decoding: The final result is a probability distribution of the possible words that could be spoken. This probability distribution is then decoded using a decoding algorithm, such as the beam search algorithm, to obtain the most probable word sequence that corresponds to the input language.

[0068] The pre-trained models provided by TensorFlow Speech Recognition can also be adapted to new datasets or customized for any use case, such as controlling a pipetting device. The project also offers tools for data preparation, model training and evaluation, as well as examples of how to use the models in real-world applications.

[0069] TensorFlow Speech Recognition makes it possible to create an ASR system that can run offline on a local computer, provided the necessary hardware resources are available. This would require training a custom model on a local dataset and deploying it for inference on the local machine.

[0070] There are also other ASR libraries and frameworks designed for offline execution on local devices with limited resources, such as the Kaldi toolkit and the Mozilla DeepSpeech project based on Tensorflow Lite.

[0071] The computer device is programmed to send the control data to the liquid handling device via a wireless connection, and the control device of the liquid handling device is programmed to receive the control data and control at least one function of the liquid handling device depending on the control data.

[0072] If the control data contains one or more numbers, the control device is programmed to define a program parameter using the one or more numbers, for example, a sample volume to be aspirated, a sample volume to be released or dispensed, a number of dispensing steps, a pipetting rate, a time, an identification number for selecting a pre-programmed method or for calling a pre-stored method, a method that contains the definition of one or more pipetting operations in a sequence.

[0073] Embodiments and further subject matter of the invention: The computer device is located remotely from the at least one liquid handling device, which means that both the computer device and the at least one liquid handling device are separate devices, i.e., independent devices that are not connected by a wired data link and have their own power supply.

[0074] The microphone device is preferably not part of the at least one liquid handling device. The microphone device is preferably arranged remotely from the at least one liquid handling device. The microphone device can be a portable device, for example, a headset. However, in embodiments, it is also possible and preferred that the microphone device is part of the liquid handling device.

[0075] Preferably, the liquid handling device has a data storage device containing device identification data suitable for uniquely identifying the liquid handling device (or its control device) within a group of many liquid handling devices (and their respective control devices). The device identification data may include a pipette identification number, in particular a serial number of the liquid handling device or a pipette device. The device identification data can be used to address the liquid handling device if several liquid handling devices are within range of wirelessly connectable devices of the computer device or within the wireless network.The use of the device identification data can enable the control data to be uniquely transmitted to the same fluid handling device that transmitted the start signal to the computer device, thus enabling the recognition of the control data. The control device can be programmed to receive and process the control data if, and preferably only if, the device identification data contained in the control data matches the device identification data of the control device or the fluid handling device that contains the control device. The control device can be programmed to receive the control data and discard any processing of the control data if, and preferably only if, the device identification data contained in the control data does not match the device identification data of the control device or the fluid handling device that contains the control device.the fluid handling device which contains the control device.

[0076] Preferably, the control device is programmed to provide the start signal in the form of start data containing the device identification data.

[0077] Preferably, the computer device is programmed to identify the liquid handling device by extracting the device identification data from the start data and to send the control data, which is associated with user voice input and extracted from the user voice by performing the speech recognition procedure, to the liquid handling device depending on the device identification data. Preferably, the device identification data is included in the control data, which means that the device identification data is transmitted to the control device together with the control data.

[0078] Preferably, the speech recognition method uses a recognition algorithm to recognize the at least one function command and to output the control data depending on the at least one function command, wherein the recognition algorithm preferably comprises a neural network or other programmed algorithms. The recognition algorithm may include providing the final output of the speech recognition method, which is the recognized text or the recognized commands derived from the spoken input and constitute the control data.

[0079] The recognition algorithm can include automatically converting the recognized text into a suitable command, which can be done using a data table that correlates the recognized text with a desired command. This table can be stored in or connected to a data memory of the computer device. The desired command can be suitable for programmatically calling a specific function on the control device, for example. - a command used by the control device to toggle the screen content displayed on a display of the liquid handling device, and / or a command to navigate through the pages of a user interface of the liquid handling device; - a command used by the control device to start or stop a pipetting operation or to turn off the power supply to the liquid handling device; - a command used by the control device to invoke a predefined or user-defined and previously stored method; - a command used by the control device to eject a pipetting tip attached to a working cone of the liquid handling device.

[0080] Preferably, the computer device is part of a pipette manager device that has a user interface with a display, in particular a touchscreen, and is configured to receive user input for setting at least one pipetting parameter and send the at least one pipetting parameter to the at least one liquid handling device when the voice control input mode is inactive. The pipette manager device may incorporate the functionality of already known pipette manager devices, e.g., the Visionize® pipette manager, commercially available from Eppendorf SE, Hamburg, Germany.

[0081] Preferably, the computer device is a dedicated computer or server. Preferably, the computer device is part of a smart device, e.g., a tablet PC or a smartphone. Preferably, the computer device is part of another laboratory device for the treatment and / or analysis of at least one liquid sample, e.g., part of a liquid handling device that is not the liquid handling device according to the invention.

[0082] Preferably, the computer device is programmed to terminate the speech recognition process by analyzing the user speech data containing the user's speech. Preferably, the computer device is programmed to terminate the speech recognition process by analyzing whether, during a predetermined period, any sound exceeding a predetermined volume or amplitude is detected by the microphone device.

[0083] Preferably, the at least one fluid handling device is configured to detect the end of manual user interaction with the actuating device, and the computer device is programmed to terminate the speech recognition process and the voice control input mode based on the detection of the end of manual user interaction. For example, the voice control input mode can be active as long as the actuating device is being actuated by the user, for example, by pressing a spring-loaded activation button / switch to define the duration of the voice control input mode.

[0084] Preferably, the at least one liquid handling device includes a signal output device, and the control device is programmed to output a signal to the user indicating whether the voice control input mode is active and / or inactive. The signal output device can be configured to provide the feedback signal.

[0085] Preferably, the actuating device comprises one of the following devices: - a special key, especially a spring-loaded key, - a button located at a specific position on the housing of a hand pipetting device, wherein this position is particularly located on the handle formed by the housing of the hand pipetting device.

[0086] Preferably, the speech recognition method is programmed to recognize a function command from the group of preferred function commands, including: - Setting a pipetting parameter, which can be one of the following: a pipetting volume, a pipetting volume to be aspirated, a pipetting volume to be dispensed, a number of pipetting steps, a pipetting speed, a number of pre-wetting steps, a pre-wetting volume, a number of mixing steps, ... - Setting a digit, including any digit from 0, 1, ...9 - Setting a number, including multi-digit numbers - Switching a liquid handling device - Ejection of a pipette tip from a working cone of the liquid handling device.

[0087] The invention also relates to the liquid handling device, which is suitable as a liquid handling device of the system according to the invention, in particular a hand pipette for handling liquid samples in a laboratory, comprising * an actuating device for electrically detecting a manual user action, * a wireless network adapter for exchanging data over a wireless connection, * an electronic control device programmed to - to detect an actuation of the operating device caused by a user action, - to activate a voice control input mode in response to the detection of the action, during which a speech recognition procedure is activated on a remote computer device to capture a speech input for controlling at least one function of the liquid handling device, and during which the control device is able to receive control data from the remote computer device, - to provide a start signal in response to the detection of the actuation, which is suitable for initiating the speech recognition process, - Sending the start signal via the wireless connection to the remote computer device, which is connected to a microphone device, in order to start the speech recognition process on the remote computer device, - preferably: Waiting for a period of time that is the maximum period during which the voice control input mode is active before it is deactivated at the end of the period in order to receive control data, - in response to the sending of the start signal, and in particular during the aforementioned period, receive the control data from the remote computer device via a wireless connection and - Controlling at least one function of the liquid handling device depending on the control data.

[0088] The invention also relates to the computer device, which is suitable as a computer device of the system according to the invention and serves to execute a programmed speech recognition method, comprising * a wireless network adapter for exchanging data over a wireless connection, wherein the computer device is programmed to * receive a start signal via wireless connection, * is triggered by the start signal to initiate the speech recognition process, which is a program-controlled process and which - records a user voice using a microphone device connected to the computer device, wherein the user voice contains information for controlling at least one function of the liquid handling device, - recognizes at least one function command from the user language and - Control data depending on which outputs at least one function command, * wherein the computer device is programmed to send the control data via a wireless connection to a liquid handling device in order to control at least one function of the liquid handling device depending on the control data.

[0089] The invention also relates to a method for performing voice control of at least one liquid handling device, in particular a hand pipette in a laboratory, using a remote computer device connected to a microphone, wherein the at least one liquid handling device comprises an actuating device for electrically detecting a manual user action, a wireless network adapter for exchanging data via a wireless connection, and an electronic control device, wherein the method comprises the following computer-implemented steps, which are performed by the electronic control device, to - Detecting an actuation of the operating device caused by a user action, - to activate a voice control input mode in response to the detection of the action, during which a speech recognition procedure is activated on a remote computer device to capture a speech input for controlling at least one function of the liquid handling device, and during which the control device is able to receive control data from the remote computer device, - to provide a start signal in response to the detection of the activation, which is suitable for starting the speech recognition procedure, - Sending the start signal via the wireless connection to the computer device to start the speech recognition process on the computer device, * wherein the method comprises the computer-implemented steps performed by the computer device to * Receiving the start signal via the wireless connection, * is triggered by the start signal to initiate the speech recognition process, which is a program-controlled process that - records the user's speech using the microphone device, wherein the user's speech contains information for controlling at least one function of the liquid handling device, - recognizes at least one function command from the user language and - Control data depending on which outputs at least one function command, * send the control data to the liquid handling device via a wireless connection, * wherein the method comprises the computer-implemented steps performed by the control device to * receive the tax data and * to control at least one function of the liquid handling device depending on the control data.

[0090] The invention also relates to a method for controlling at least one function of a liquid handling device, in particular a hand pipette, wherein the liquid handling device comprises an actuating device for electrically detecting a manual user action, a wireless network adapter for exchanging data via a wireless connection and an electronic control device. * wherein the procedure comprises the computer-implemented steps performed by the control unit to - Detecting an actuation of the operating device caused by a user action, - to activate a voice control input mode in response to the detection of the action, during which a speech recognition procedure is activated on a remote computer device to capture a speech input for controlling at least one function of the liquid handling device, and during which the control device is able to receive control data from the remote computer device, - to provide a start signal in response to the detection of the activation, which is suitable for starting the speech recognition procedure, - Sending the start signal via the wireless connection to the remote computer device, which is connected to a microphone device, in order to start the speech recognition process on the remote computer device, - preferably: Wait for a period of time equal to the maximum period during which the voice control input mode is active before deactivating it at the end of that period in order to receive control data, - in response to the sending of the start signal and in particular during the aforementioned period, receive the control data from the remote computer device via a wireless connection and - Controlling at least one function of the liquid handling device depending on the control data.

[0091] The invention also relates to a method for controlling at least one function of a liquid handling device, in particular a hand pipette, using a speech recognition method executed by a computer device, wherein the method comprises the following steps performed by the computer device * Receiving a start signal via the wireless connection, * is triggered by the start signal to initiate a speech recognition process, which is a program-controlled process and which - records a user voice using a microphone device connected to the computer device, wherein the user voice contains information for controlling at least one function of the liquid handling device, - recognizes at least one function command from the user language and - Control data depending on which outputs at least one function command, * sends the control data to the liquid handling device via a wireless connection in order to control at least one function of the liquid handling device depending on the control data.

[0092] The invention also relates to program code for implementing the method according to claim 14 or claim 15 or claim 16. The invention also relates to a computer program [product] comprising instructions which, when the program is executed by a computer, cause the computer to execute the methods or the steps of the method according to claim 14 or claim 15 or claim 16.

[0093] The invention also relates to a computer-implemented method for performing a speech recognition process by determining at least one functional command from speech data, comprising the following steps: • Capturing speech data containing a digitized sound wave received through a microphone device, • Determining function command data that contains information about at least one function command for controlling at least one function of a liquid handling device, wherein the function command data are determined by inputting the speech data into an algorithm that establishes the relationship between speech data and function commands, wherein the algorithm is part of a learning algorithm, in particular a neural network or a convolutional neural network (CNN) that has been trained to establish this relationship.

[0094] Programming parameters to be defined via the voice control input mode

[0095] The voice-controlled input mode is useful for a user when defining program parameters, which can then define a more or less complex pipetting method that follows a predefined operating mode or is a completely user-defined method. In particular, defining digits of a number or numbers by speech recognition saves a great deal of time, especially when many program parameters need to be defined. The invention enables the implementation of a powerful speech recognition method that runs on a high-performance computer device separate from the liquid handling device.

[0096] The following describes the operating modes and their preferably associated operating parameters, each preferably provided by or used by the liquid handling device. The term "operating parameter" is also used interchangeably with "program parameter" or "pipetting parameter": Preferably, an operating parameter is provided that defines a volume to be pipetted. An operating parameter may be provided that defines an intake volume to be drawn in during the aspiration step, and / or an operating parameter may be provided that defines a delivery volume to be dispensed during a delivery step.

[0097] Preferably, at least one operating parameter is provided with which the number of directly successive or indirectly successive pipetting volumes is determined, preferably at least one operating parameter with which the number of aspiration steps and / or dispensing steps and preferably also the respective associated pipetting volumes, the respective associated pipetting velocities and / or accelerations and / or the respective associated time intervals between the steps are determined.

[0098] Preferably, one operating mode of the liquid handling device, in particular the hand pipetting device, relates to the "dispensing" (DIS) of a sample. Associated operating parameters preferably include: the volume of each individual sample relative to the pipetting volume during one of several dispensing steps; the number of dispensing steps; the rate during sample intake; and the rate during sample dispensing. The dispensing function is particularly suitable for the rapid filling of a microtiter plate with a liquid reagent and can be used, for example, to perform an ELISA.

[0099] Preferably, an operating mode relates to the "automated dispensing" (ADS) of a sample. Associated operating parameters preferably include: the volume of each individual sample relative to the pipetting volume during one of several dispensing steps; the number of dispensing steps; the duration of the time interval after which the dispensing steps are automatically performed sequentially at constant intervals – the time interval can define these intervals or, for example, the delay between the end and the beginning of successive dispensing steps; the rate during sample uptake; and the rate during sample dispensing. This dispensing function is more convenient for filling a microtiter plate because the user does not need to repeatedly trigger a release step by actuating it, for example, by pressing a button; instead, the release occurs automatically after the automated dispensing has started.Like all other operating programs of an operating mode, automated dosing can also be performed if the corresponding program is executed for at least one continuous actuation of a control element, such as a continuously pressed button. This is advantageous, for example, for long series of dosing operations or reactions where a precise time window must be adhered to. The automated dosing function is particularly convenient for filling a microtiter plate, as in this scenario the user does not need to trigger a single dispensing step by pressing a button; the dispensing occurs automatically. This can be used, for example, for performing an ELISA.

[0100] Preferably, one operating mode relates to the "pipetting" (pip) of a sample. Associated operating parameters are preferably: the volume of the sample to be pipetted; the rate during sample intake; the rate during sample dispensing.

[0101] Preferably, an operating mode relates to the "pipetting followed by mixing" (P / Mix) of a sample. Associated operating parameters are preferably: the volume of the sample to be aspirated and / or dispensed; the mixing volume; the number of mixing cycles; the rate during sample aspiration; and the rate during sample dispensing. The "pipetting followed by mixing" function is recommended, for example, for pipetting very small volumes. If a dispensing volume < 10 µl is selected, it is recommended to flush this volume into the corresponding reaction fluid. This can be achieved by automatically initiating a mixing motion after the liquid is dispensed. The mixing volume and the number of mixing cycles are predefined. One application of this operating mode is, for example,The dispensing of a liquid that, due to its physical properties, is more difficult to dose than water, whereby its residues in the pipetting container, particularly in the pipette tip, are flushed out with the existing liquid from the container or pipette tip. Another application could be the immediate mixing of the dispensed liquid with the existing liquid. This mode of operation is advantageous, for example, when DNA is added to a PCR mixture.

[0102] A preferred operating mode relates to the "repeated aspiration" of a sample, also referred to as "inverted dosing" or "ASP" for aspiration. The associated operating parameters are preferably: the volume of the sample(s) to be aspirated; the number of samples; the aspiration rate; and the dispensing rate. The function serves to repeatedly aspirate a quantity of liquid and dispense the total quantity. Refilling the pipetting chamber in a single operation is not required. The rate is the same for all samples. During operation, the following preferably occurs: Starting from the initial position, the pipetting device aspirates a partial volume by actuating the first type of control device.After the last partial volume has been aspirated, the pipetting device preferably issues a warning, which the user must acknowledge, preferably by actuating the second type of control device. Upon subsequent actuation of the second type of control device, the total volume is dispensed again. For actuating the first or second type, the control device preferably comprises at least two operating buttons: one for inputting a control signal of the "first type" to the control device and one for inputting a control signal of the "second type" to the control device. The control device may, in particular, include a toggle switch that is pivotable about an axis perpendicular to the longitudinal axis of the pipetting device between a first signal release position ("toggle switch up") for actuating the first type and a second signal release position ("toggle switch down") for actuating the second type.

[0103] Preferably, one operating mode relates to the "dilution" (Dil) of a sample. Associated operating parameters are preferably: the sample volume, the air bubble volume, the volume of the diluent, the rate of intake, and the rate of dispensing. The maximum volume of the diluent = nominal volume - (sample + air bubble). This function serves to take in a sample and a diluent, separated by an air bubble, and to dispense the total quantity. The rate is the same for all partial volumes. During operation, the following preferably occurs: Starting from the initial position, the pipetting device first takes in the volume of the diluent, then an air bubble, and finally the sample. Each intake is preferably triggered separately by actuating the control device of the first type. Subsequently, the total quantity is dispensed completely.

[0104] One operating mode preferably relates to the "sequential dispensing" (SeqD) of samples. The associated operating parameters are preferably: the number of samples (preferably up to a maximum number Nmax of preferably 5 ≤ Nmax ≤ 15, preferably Nmax = 10); the individual volumes of each sample; the aspiration rate; and the dispensing rate. This function serves for the sequential dispensing of freely selectable volumes Nmax, preferably without refilling the pipetting container multiple times. The rate is the same for all samples. The number of samples is preferably the primary parameter for inputting the individual volumes. The pipette must preferably check with each volume input that the maximum volumes of the pipetting device are not exceeded; if necessary, a warning is issued.After all parameters have been entered, the pipetting device takes in the total volume after actuating the first type of control device and dispenses a single volume after each actuation of the second type of control device. All further processes are preferably carried out as in normal dispensing.

[0105] One operating mode preferably relates to the "sequential pipetting" (SeqP) of samples. The associated operating parameters are preferably: the number of samples (preferably up to a maximum number Nmax of preferably 5 ≤ Nmax ≤ 15, preferably Nmax = 10); individual volumes of the individual samples; aspiration rate; dispensing rate. This function is used to pipette a maximum number Nmax of freely selectable volumes, which are programmed before starting and whose sequence is fixed. The rate is preferably the same for all samples to allow for easy handling of this operating mode. The rate can also be set variably. The execution of the function corresponds to the execution of pipetting. The previously entered volumes are processed in the programmed sequence. After dispensing, the activation of a control element, e.g.,By pressing a button, the user can choose whether to take up the next sample, or whether to perform a "blow-out" (i.e., a complete and safe blow-out of the samples still contained in the pipetting container) by means of an overstroke before taking up the next sample, and / or whether to change the pipetting container.

[0106] Preferably, one operating mode relates to the reverse pipetting (rPip) of samples. Associated operating parameters are preferably: the volume of the individual sample; the aspiration rate; the dispensing rate; and the activation of the counters. For this rPip function, more volume than is to be dispensed is aspirated. This is achieved by lowering the piston before the liquid is aspirated, by means of a second type of actuation, i.e., by pressing a button or a rocker switch downwards, into the lower position of a blow-out process, i.e., an overstroke of the piston that exceeds the piston's position in a single pipetting stroke. At the start of the volume aspiration, the pipetting device aspirates the dispensed volume and the selected volume. To account for the backlash of the actuator in the triggering direction, the pipetting device performs an additional stroke, which is immediately triggered again.This is similar to dispensing, but preferably done with an automated triggering of the dispensing stroke at maximum speed.

[0107] During the execution of the "rPip" operating mode, the following preferably occurs: 1. The piston of the pipetting device automatically moves to the blow-out position and remains in the lower position. 2. The first type of control device is activated: The piston moves upwards by the blow-out stroke and the pipetting volume stroke. 3. The second type of control device is activated: The piston moves downwards by the pipetting volume stroke and stops before blowing out. 4. The second type of control device is activated: The piston performs the blow-out and remains in the lower position. Alternatively to the fourth type, the first type of control device is activated: The piston moves upwards by the pipetting stroke. The "rPip" mode is particularly suitable for pipetting plasma, sera, and other liquids with a high protein content.The "Pipetting" mode is particularly suitable for aqueous solutions. The "rPip" mode is especially suitable for solutions containing surfactants to minimize foaming when dispensing into the target container. The liquid is drawn up with a certain amount of excess volume (blowout volume). This excess volume is generally not part of the dispensed volume and is preferably not dispensed into the target container. Particularly if the same sample is to be used again for pipetting, the excess volume can remain in the pipette tip. If a different liquid is used, the excess volume and / or preferably the pipetting container is discarded.

[0108] A set of operating parameters preferably controls a control program for executing the desired pipetting process. The control program can be provided in the form of electrical circuits of the control device and / or by executable program code suitable for controlling a control device that is controllable and preferably programmable by program code.

[0109] The system, the liquid handling device, the computer device and the methods according to the invention are applicable, for example, in research or industry in the fields of pharmaceuticals, biomedicine, biotechnology, in particular plant biotechnology, microbiology, chemistry or nanotechnology.

[0110] Further preferred configurations of the system, the computer device, and the fluid handling device according to the invention, and of the methods according to the invention, will become apparent from the following description of the exemplary embodiments in conjunction with the figures and their description. Unless otherwise described or evident from the context, identical components of the exemplary embodiments are essentially identified by the same reference numerals. In detail: Fig. Figure 1 shows a liquid handling device, which is a hand pipette according to an embodiment of the invention. Fig. Figure 2a shows a system according to an embodiment of the invention. Fig. Figure 2b shows a system according to an embodiment of the invention. Fig. Figure 3 shows a system according to an embodiment of the invention. Fig. Figure 4 schematically shows an example of the method according to the invention using the system according to the invention.

[0111] Fig. Figure 1 shows a perspective view of the liquid handling device 1 for implementing voice control of at least one liquid handling device, in particular a hand-held pipetting device, according to an embodiment of the invention, which is a hand-held electric piston-stroke pipette 1. In the pipette 1, the stroke of the piston is electrically driven. The activation of the stroke in the various operating modes of the pipette is electrically controlled by an electronic control device 17 with an attached storage device within the pipette 1. The control device 17 includes a wireless network adapter 18, a radio module (WiFi), to transmit data to an external computer device (see Figure 1). Fig. ) via a Wi-Fi connection.

[0112] The operating parameters (program parameters) and other settings of the pipette can be controlled by the user via the user interface or the control device and the pipette's display. To detect a user action and to activate a voice control input mode in response to the detection of the actuation, the pipette 1 includes the actuation device 50, which is a spring-loaded button 50 on the front of the pipette housing.

[0113] Pipette 1 stores several electrically controlled pipetting programs, with each operating mode preferably being assigned a pipetting program corresponding to a pipetting procedure. A pipetting program can be uniquely defined by a set of operating parameters. Once defined, the pipetting program can be triggered by the user and is automatically started by the pipette. The pipetting program specifically includes the execution of a pipetting procedure.

[0114] The pipette 1 comprises a base body 2, which includes a lower shaft section 3 and an upper section 4, which in particular includes the display 5 and the controls. The control section 3 runs parallel to the longitudinal axis A of the pipetting device, while the upper section 4 is inclined to axis A and runs parallel to axis B. The inclined arrangement of the upper section 4 allows for very ergonomic use of the display.

[0115] The pipette 1 comprises a handle section 7 with a retaining flap 6, which rests on the user's index finger when the pipette 1 is held by the user as intended, while the handle section 7 rests in the user's palm. The thumb can, in particular, reach the ejection button 8, which, when pressed along axis A, moves the spring-loaded ejection sleeve 9 downwards and ejects the pipette tip 10 from the nose cone 11 of the pipetting device to which it is attached. The ejection mechanism can also be electronically controlled. The pipette 1 includes a metallic contact projection 19 on each side of the upper section 4, which serves to charge the integrated battery that forms the energy storage device of the electric pipette. The thumb can also, in particular, reach the actuating device 50.

[0116] The operating unit (12; 13; 14a; 14b) (also referred to as user interface device / equipment) comprises a rotary knob 12, a toggle switch 13, a first control button 14a, a second control button 14b, and the actuating device 50. The electronic control device is located inside the base body 2 and is connected to the operating unit, in particular to the actuating device 50.

[0117] The electronic control device is programmed to - to detect an actuation of the actuating device 50 caused by a user action, - to activate a voice control input mode in response to the detection of the action, during which a speech recognition procedure is activated on the computer device to recognize a speech input for controlling at least one function of the liquid handling device, and during which the control device is able to receive control data from the remote computer device, - in response to the detection of the actuation, provide a start signal and send the start signal to the computer device via the wireless connection in order to start the speech recognition process on the computer device, - after completion of the speech recognition process, receive control data from the computer device and - define at least one program parameter using the control data.

[0118] The disc-shaped adjusting wheel 12 is rotatably mounted on the base body 2, in particular parallel to the substantially flat front surface of the upper section 4. The adjusting wheel 12 comprises a number of detents corresponding to the number of selectable positions of the adjusting wheel. The detents are designed such that a marking 12a for indicating the set position of the rotary knob 12 can be aligned with the marking 15, which is attached to the front surface of the upper part 4 on the base body 2.

[0119] The color display 5 serves as the central information element for the user. Specifically, it displays the various operating modes of the pipette 1 and the parameter values ​​of the operating parameter. In each of the two areas 5a and 5b, information is displayed that tells the user which function is assigned to the first control button 14a or the second control button 14b on the currently displayed screen, provided a function is assigned to each button on that screen. Each control button is thus designed as a control element with variable functionality and, in combination with the displayed function, is referred to as a "softkey".

[0120] The actuating device 50 is a dedicated button and, in the present embodiment, not a "softkey". The user can observe on the display 5 whether their user voice input has been correctly recognized, for example, to form the desired digit if voice recognition is used to define the digits of a number that defines the value of a program parameter, for example, the value of the user parameter "sample volume to be aspirated".

[0121] Preferably, the pipetting device is designed to switch between the different functions of a softkey when a specific operating mode of pipette 1 is selected. This can be achieved, for example, by double-clicking the softkey or by holding the softkey for a minimum time, e.g., for 2 seconds.

[0122] Preferably, a display page is provided for each operating mode of the pipette 1, which is shown on the display with the layout specific to that operating mode. A display page may also be provided for defining at least one pre-wetting step. If adjustable operating parameters or other changeable entries are provided on the display page, these can be highlighted with the control rocker switch 13 and, in particular, selected with the control button 14a. In this case, the control button 14a has the "selection" function, and the text is displayed at position 5a. Changing the parameter values ​​of an operating parameter or changing the selection or an entry is done by actuating the rocker switch 13.

[0123] The toggle switch 13 is arranged on the base body such that it can pivot about an axis perpendicular to the longitudinal axis A. When the user presses the upper section 13a, a first function of the toggle switch 13 is activated; when the user presses the lower section 13b, a second function of the toggle switch 13 is activated. The toggle switch is positioned so that no function is triggered when it is not pressed. In particular, in a manual operating mode of the pipette, the toggle switch 13 serves to aspirate the sample to be pipetted into the pipette tip 10 while the user presses the upper section 13a, and to dispense the sample from the pipette tip 10 while the user presses the lower section 13b.

[0124] The Pipette 1 can be operated in various modes, which have been explained in detail above. A first set of operating modes can be selected directly via the rotary knob 12; a second set of operating modes can be selected via a display page labeled "special" or "Spc," where each entry describes an operating mode. Within each of the operating modes set with the rotary knob 12, the actuator / button 50 can be used to start the voice control input mode and to define one or more program parameters.

[0125] Fig. Figure 1 schematically shows a system 1000 according to an embodiment of the invention. The system comprises the pipette 1, the computer device 100, and a microphone device 80, all of which can be located in the same laboratory space but separately from one another. The pipette 1 and the computer device 100, as well as the microphone device 80 and the computer device 100, can be connected via a WLAN 200. The WLAN can be established by the computer device 100, which can function as a WLAN access point. Alternatively, the pipette 1 and the computer device 100, as well as the microphone device 80 and the computer device 100, can be wirelessly connected in another way. The microphone device 80 and the computer device 100 can alternatively be connected by a cable connection.

[0126] Fig. Figure 2b schematically shows a system 1000' according to an alternative embodiment of the invention, wherein the liquid handling device 1 is connected to the computer device 100 via a router 90 and an external network 250, for example, the Internet 250. The system comprises the pipette 1, the computer device 100, and a microphone device 80, wherein the pipette 1 and the microphone device 80 can be located in the same laboratory room but separately from each other. The pipette 1 and the router 90, as well as the microphone device 80 and the router 90, can be connected via a WLAN 200. The router can comprise a computer device. The router can be part of a computer device, for example, part of a pipetting manager.

[0127] The WLAN can be set up using the router 90, which can act as a WLAN access point. The microphone device 80 can be part of the computer device 90 or the router 90.

[0128] The system 1000 or 1000' is configured for voice control of at least one liquid handling device 1, in particular a hand pipetting device 1, in a laboratory and comprises: * which includes at least one liquid handling device 1, * a microphone device 80, * a computer device 100 connected to the microphone device, wherein the computer device 100 and the microphone device 80 are arranged remotely from the at least one liquid handling device, wherein the at least one liquid handling device 1 comprises * an actuating device 50 for electrically detecting a manual user action, * a wireless network adapter 18 for exchanging data over a wireless connection, * an electronic control device 17 which is programmed to - to detect an actuation of the actuating device 50 caused by a user action, - to activate a voice control input mode in response to the detection of the action, during which a speech recognition procedure is activated on the computer device 100 to capture a speech input for controlling at least one function of the liquid handling device 1, and during which the control device 17 is able to receive control data from the remote computer device 100, - to provide a start signal in response to the detection of the actuation and to send the start signal via the wireless connection 200 to the computer device 100 in order to start the speech recognition procedure on the computer device 100, wherein the computer device 100 includes a wireless network adapter 101 for exchanging data via a wireless connection and is programmed such that it * receive the start signal via the wireless connection 200, * to be triggered by the start signal to initiate the speech recognition process, which is a program-controlled process that - records the user's speech using the microphone device, wherein the user's speech contains information for controlling at least one function of the liquid handling device, - recognizes at least one function command from the user language and - Control data depending on which outputs at least one function command, wherein the computer device 100 is programmed to send the control data to the liquid handling device 1 via a wireless connection 200, and wherein the control device 17 of the liquid handling device 1 is programmed to receive the control data and to control the at least one function of the liquid handling device 1 depending on the control data.

[0129] Fig. Figure 3 is a further schematic representation of an embodiment of the system 1000 according to the invention, which uses the same reference numerals as in Figure 3. Fig. 2a is used. To implement the microphone device 80, several microphones 80a, 80a', 80a" can be provided here, wherein the microphones are arranged, for example, at different locations in a laboratory to ensure the recording of a high-quality sound wave that is particularly suitable for carrying out the speech recognition procedure.

[0130] Fig. Method 300 shows for carrying out voice control of at least one liquid handling device 1, in particular a hand pipetting device 1 in a laboratory, using a remote computer device 100 connected to a microphone device 80, wherein the at least one liquid handling device 1 comprises an actuating device 50 for electrically detecting a manual user action, a wireless network adapter 18 for exchanging data via a wireless connection and an electronic control device 17. * wherein the method 300 comprises the computer-implemented steps performed by the electronic control device 17 to - Detecting an actuation of the actuating device 50 caused by a user action; (Step 301) - to activate a voice control input mode in response to the detection of the action, during which a speech recognition procedure is activated on a remote computer device to capture a speech input for controlling at least one function of the liquid handling device, and during which the control device is able to receive control data from the remote computer device; (Step 302) - in response to the detection of the actuation, provide a start signal suitable for initiating the speech recognition process; (Step 303) - Sending the start signal via the wireless connection to the computer device to start the speech recognition process on the computer device (Step 304) * wherein the method comprises the computer-implemented steps performed by the computer device 100 to * Receiving the start signal via the wireless connection (Step 305) * triggered by the start signal to initiate the speech recognition process (step 306), which is a program-controlled process that - records the user's speech using the microphone device 80, wherein the user's speech contains information for controlling at least one function of the liquid handling device (step 307) - recognizes at least one function command from the user language (step 308) and - Control data depending on which outputs at least one function command (step 309), * sends the control data to the liquid handling device via a wireless connection (step 310) * wherein the method comprises the computer-implemented steps performed by the control device 17 to * receive the tax data (step 311) and * to control at least one function of the liquid handling device depending on the control data (step 312). QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] EP 1438138B1

[0006] Cited non-patent literature

[0000] International Conference on Informatics and Computing (ICIC). pp. 1-5. IEEE (2018

[0062] Becker, S., Ackermann, M., Lapuschkin, S., Müller, K., Samek, W.: Interpreting and explaining deep neural networks for classification of audio signals. CoRR (2018), http: / / arxiv.org / abs / 1807.03418

[0062] Araabi, A., Monz, C.: Optimizing transformer for low-resource neural machine translation. In: Scott, D., Bel, N., Zong, C. (Hrsg.) Proceedings of the 28th International Conference on Computational Linguistics. S. 3429-3435

[0062] International Committee on Computational Linguistics (2020), https: / / doi.org / 10.18653 / v1 / 2020.coling-main.304

[0062] Devlin, J., Chang, M., Lee, K., Toutanova, K.: BERT: Vorabtraining tiefer bidirektionaler Transformatoren für das Sprachverständnis. In: Burstein, J., Doran, C., Solorio, T. (Hrsg.) Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. S. 4171-4186. Association for Computational Linguistics (2019), https: / / doi.org / 10.18653 / v1 / n19-1423

[0062] Dhar, P.: The carbon impact of artificial intelligence. Nat. Mach. Intell. 2(8), 423-425 (2020), https: / / doi.org / 10.1038 / s42256-020-0219-9

[0062] www.tu-harburg.de / mst

[0063] https: / / kaldi-asr.org

[0065] https: / / www.tensorflow.org

[0065]

Claims

System (1000; 1000') for performing voice control of at least one liquid handling device (1), in particular a hand pipette, in a laboratory, comprising: * the at least one liquid handling device (1), * a microphone device (80), * a computer device (100) connected to the microphone device, wherein the computer device and preferably the microphone device are arranged remotely from the at least one liquid handling device, wherein the at least one liquid handling device comprises: * an actuating device (50) for electrically detecting a manual user action, * a wireless network adapter (18) for exchanging data via a wireless connection (200), * an electronic control device (17) programmed to detect an actuating device caused by a user action,- to activate a voice control input mode in response to the detection of the actuation, during which a speech recognition procedure is activated on the computer device to detect a voice input for controlling at least one function of the fluid handling device, and during which the control device is able to receive control data from the remote computer device; - to provide a start signal in response to the detection of the actuation and to transmit the start signal to the computer device via the wireless link to initiate the speech recognition process on the computer device, wherein the computer device (100) includes a wireless network adapter (101) for exchanging data via a wireless link and is programmed to * receive the start signal via the wireless link, * be triggered by the start signal to initiate the speech recognition process,which is a program-controlled process which: - records the user's speech using the microphone device, wherein the user's speech contains information for controlling at least one function of the liquid handling device; - recognizes at least one function command from the user's speech; and - outputs control data depending on the at least one function command, wherein the computer device (101) is programmed to transmit the control data to the liquid handling device via a wireless connection (200); and wherein the control device of the liquid handling device (1) is programmed to receive the control data and control the at least one function of the liquid handling device depending on the control data. System according to claim 1, wherein the liquid handling device comprises a data storage device containing device identification data suitable for uniquely identifying the liquid handling device within a group of many liquid handling devices, wherein the control device is programmed to provide the start signal in the form of start data containing the device identification data, and wherein the computer device is programmed to identify the liquid handling device by extracting the device identification data from the start data and sending the control data to the liquid handling device depending on the device identification data. System according to one of the preceding claims, wherein the speech recognition method uses a recognition algorithm for recognizing the at least one function command and for outputting the control data depending on the at least one function command, wherein the recognition algorithm comprises a neural network. System according to one of the preceding claims, wherein the computer device is part of a pipette manager device comprising a user interface device with a display, in particular a touchscreen, and configured to receive user input for setting at least one pipetting parameter and sending the at least one pipetting parameter to the at least one liquid handling device when the voice control input mode is inactive. System according to one of the preceding claims, wherein the computer device is a dedicated server, part of an intelligent device, e.g. a tablet PC or a smartphone, or part of another laboratory device for the treatment and / or analysis of at least one liquid sample. System according to one of the preceding claims, wherein the computer device is programmed to complete the speech recognition process by analyzing the user speech data containing the user speech. System according to one of the preceding claims, wherein the computer device is programmed to terminate the speech recognition process by analyzing that no noise exceeding a predetermined volume is detected by the microphone device during a predetermined period of time. System according to one of the preceding claims, wherein the at least one fluid handling device is configured to detect the end of manual user interaction with the actuating device, and the computer device is programmed to terminate the speech recognition process and the voice control input mode based on the detection of the end of manual user interaction. System according to one of the preceding claims, wherein the at least one liquid handling device comprises a signal output device and the control device is programmed to output a signal to the user indicating whether the voice control input mode is active and / or inactive. System according to one of the preceding claims, wherein the actuating device comprises one of the following elements: - a special button, in particular a spring-loaded button, - a button located at a specific position on the housing of a hand pipetting device. System according to one of the preceding claims, wherein the speech recognition method is programmed to recognize a function command from the group of preferred function commands, including: - setting a pipetting parameter, which may be one of the following parameters: a pipetting volume, a pipetting volume to be aspirated, a pipetting volume to be dispensed, a number of pipetting steps, a pipetting rate, a number of pre-wetting steps, a pre-wetting volume, a number of mixing steps, ... - setting a digit, including any digit from 0, 1, ... 9 - setting a number, including multi-digit numbers - switching a liquid handling device - ejecting a pipette tip from a working cone of the liquid handling device. Liquid handling device, in particular hand pipetting device, for handling liquid samples in a laboratory, comprising: * an actuating device for electrically detecting a manual user action, * a wireless network adapter for exchanging data via a wireless connection, * an electronic control device programmed to: - detect actuation of the actuating device caused by a user action, - activate a voice control input mode in response to the detection of the actuation, during which a speech recognition method is activated on a remote computer device to detect voice input for controlling at least one function of the liquid handling device, and during which the control device is able to receive control data from the remote computer device, - provide a start signal in response to the detection of the actuation.which is suitable for starting the speech recognition procedure, - sending the start signal via the wireless connection to the remote computer device, which is connected to a microphone device, in order to start the speech recognition procedure on the remote computer device, - preferably: waiting for a period of time corresponding to the maximum period of time during which the speech control input mode is active before it is deactivated at the end of this period of time in order to receive control data, - receiving the control data from the remote computer device via a wireless connection in response to the sending of the start signal and in particular during the aforementioned period of time, and - receiving the control data from the remote computer device via a wireless connection in response to the sending of the start signal and in particular during the aforementioned period of time,and - controlling at least one function of the liquid handling device depending on the control data. A computer device for performing a speech recognition method, comprising a wireless network adapter for exchanging data via a wireless connection, wherein the computer device is programmed to receive a start signal via the wireless connection, be triggered by the start signal to initiate the speech recognition method, which is a program-controlled method and which: records the speech of a user using a microphone device connected to the computer device, wherein the speech of the user contains information for controlling at least one function of the liquid handling device; recognizes at least one function command from the user's speech; and outputs control data depending on the at least one function command, wherein the computer device is programmed to transmit the control data to a liquid handling device via a wireless connection.to control at least one function of the liquid handling device depending on the control data. Method (300) for performing voice control of at least one liquid handling device (1), in particular a hand pipetting device (1) in a laboratory, using a remote computer device (100) connected to a microphone device (80), wherein the at least one liquid handling device (1) comprises an actuating device (50) for electrically detecting a manual user action, a wireless network adapter (18) for exchanging data via a wireless connection, and an electronic control device (17),* wherein the method (300) comprises the computer-implemented steps performed by the electronic control device (17) to detect an actuating device (50) caused by a user action;(301)- to activate a voice control input mode in response to the detection of the actuation, during which a speech recognition procedure is activated on a remote computer device to detect a voice input for controlling at least one function of the fluid handling device, and during which the control device is able to receive control data from the remote computer device; (302)- to provide a start signal suitable for initiating the speech recognition process in response to the detection of the actuation; (303)- to transmit the start signal to the computer device via the wireless link in order to initiate the speech recognition process on the computer device; (304)* wherein the method comprises the computer-implemented steps performed by the computer device (100) to* receive the start signal via the wireless link;(305)* is triggered by the start signal to initiate the speech recognition process (306), which is a program-controlled process that: - records the user's speech using the microphone device, wherein the user's speech contains information for controlling at least one function of the liquid handling device (307); - recognizes at least one function command from the user's speech (308); and - depending on the at least one function command, outputs control data (309),* transmitting the control data to the liquid handling device via a wireless connection; (310)* wherein the method comprises the computer-implemented steps performed by the control device (17) to:* receive the control data (311); and* control the at least one function of the liquid handling device depending on the control data (312). Method for controlling at least one function of a liquid handling device, in particular a hand pipette, wherein the liquid handling device comprises an actuating device for electrically detecting a manual user action, a wireless network adapter for exchanging data via a wireless connection, and an electronic control device,* wherein the method comprises the computer-implemented steps performed by the control device to: - detect an actuating of the actuating device caused by a user action; - activate a voice control input mode in response to the detection of the actuating, during which a speech recognition method is activated on a remote computer device to detect a voice input for controlling at least one function of the liquid handling device, and during which the control device is able to:to receive control data from the remote computer device, - to provide a start signal suitable for initiating the speech recognition process in response to the detection of the actuation, - to send the start signal via the wireless connection to the remote computer device, which is connected to a microphone device, in order to initiate the speech recognition process on the remote computer device, - preferably: to wait for a period of time which is the maximum period of time during which the voice control input mode is active before it is deactivated at the end of the period of time in order to receive control data, - in response to the sending of the start signal and in particular during the aforementioned period of time, to receive the control data from the remote computer device via a wireless connection, and - to control the at least one function of the fluid handling device depending on the control data. Method for controlling at least one function of a liquid handling device, in particular a hand pipette, using a speech recognition method executed by a computer device, wherein the method comprises the following computer-implemented steps executed by the computer device to: * receive a start signal via the wireless connection, * be triggered by the start signal to initiate a speech recognition process, which is a program-controlled process that: - records the user's speech via a microphone device connected to the computer device, wherein the user's speech contains information for controlling at least one function of the liquid handling device, - recognize at least one function command from the user's speech, and - output control data depending on the at least one function command.* to send the control data to the liquid handling device via a wireless connection in order to control at least one function of the liquid handling device depending on the control data. Program code for implementing the method according to claim 14 or claim 15 or claim 16.

Images