Evaluation and computer-implemented method for presence of anomalies in laboratory instruments

Abstract

Aspects of the invention relate to a computer-implemented method comprising the steps of: obtaining log data from or by a first computing device, the log data comprising a first plurality of log data items; and obtaining exception data, the exception data comprising information indicative of the presence of an exception in the laboratory instrument, where the first computing device is configured to access a log file of the laboratory instrument, the log file comprising a first plurality of log file entries, each log file entry of the first plurality of log file entries comprising a plurality of fields, and the first computing device is configured to construct log data by, for each log data item of the first plurality of log data items: selecting first field data of a first field of a respective log file entry, the first field of the respective log file entry specifies a respective event that has occurred in the laboratory instrument; selecting second field data for a second field of the respective log file entry, the second field of the respective log file entry further describing a respective event occurring in the laboratory instrument, where the selection of the respective second field is based on the first field data for the respective first field of the respective log file entry; and constructing each log data item by using the first field data of the first field of the respective log file entry and the second field data of the second field of the respective log file entry. Additional aspects of the invention relate to other computer-implemented methods, data processing systems, laboratory instruments, automated laboratory systems, computer program products, and computer-readable media.

Description

Technical Field

[0001] The present invention relates to analyzing and / or testing biological samples. Background Art

[0002] The reliability of the results of studying biological samples (also referred to herein as "samples") in an automated laboratory system (also referred to herein as an "ALS") depends on whether the laboratory instruments included by the ALS correctly and completely perform the tasks associated therewith. In the event of an error in the laboratory instruments, this may reduce the reliability of the results. Due to the large number of samples studied by the ALS and the number and complexity of the laboratory instruments, it is difficult to detect such errors or anomalies after the study of the biological samples has been completed. Summary of the invention

[0003] Therefore, it is an object of the present invention to make the analysis and / or testing of biological samples more reliable. In particular, an improved computer-implemented method is provided.

[0004] This object is solved by the attached independent claims. Detailed description of the invention follows from the respective dependent claims.

[0005] A first aspect of the present invention relates to a computer-implemented method comprising the following steps:

[0006] - obtaining log data from or by the first computing device, the log data comprising a first plurality of log data items; and

[0007] - obtaining anomaly data comprising information indicative of the presence of an anomaly in the laboratory instrument,

[0008] The first computing device is configured to access a log file of a laboratory instrument, the log file comprising a first plurality of log file entries, each log file entry in the first plurality of log file entries comprising a plurality of fields, and the first computing device is configured to construct log data by performing the following operations for each log data item in the first plurality of log data items:

[0009] ○Select the first field data of the first field of the corresponding log file entry, the corresponding log file entry

[0010] The purpose first field specifies the corresponding event that has occurred in the laboratory instrument;

[0011] ○Select the second field data of the second field of the corresponding log file entry, the corresponding log file entry

[0012] The second field of the purpose further describes the corresponding event that has occurred in the laboratory instrument, where

[0013] The selection of the corresponding second field is based on the first

[0014] Field data; and

[0015] ○ By using the first field data of the first field of the corresponding log file entry and the corresponding log

[0016] Each log data item is constructed by using the second field data of the second field of the file entry.

[0017] In particular, the presence of anomalies is evaluated by using the log data.Herein, the respective event may also be referred to as an event associated with the respective log data item and / or as an event associated with the respective log data entry.

[0018] Advantageously, according to the aforementioned first aspect of the invention, the selection of the first field and the second field allows to simplify the data by discarding irrelevant information, and allows to make correlations between log file entries indicating the log file. In other words, by constructing log data items by log file entries or mapping log data items to log file entries, it allows better data correlation. Therefore, the method according to the first aspect of the invention allows better, faster and / or easier data processing. Alternatively or additionally, the log data is used to detect anomalies, and the improved data correlation between the log data items causes better, faster and / or easier anomaly detection. Therefore, the reliability of the results of the sample analysis provided by the laboratory instrument of the automated laboratory system, for example, is increased.

[0019] According to the present invention, a computing device, such as the above-mentioned first computing device, may include or be a cloud computing system or platform. The computing device may include or be a distributed computer system. The computing device may include or be a network of connected desktop PCs and / or workstations and / or laptops and / or tablet PCs, etc. The network may be a LAN and / or a wireless LAN. In particular, according to the present invention, the computing device may be a desktop PC, a workstation, a laptop or a tablet PC, etc. As described herein, a computing device (e.g., a first computing device) may include one or more processors (e.g., a CPU, a GPU, a tensor processing unit (TPU), etc.) and a memory.

[0020] The first computing device may be included in the laboratory instrument. Alternatively, the first computing device may not be included in the laboratory instrument or in an ALS comprising the laboratory instrument. In this case, in particular, the first computing device may be in data communication with the laboratory instrument. The first computing device may in particular be a server or a cloud computing system that is in data communication with the laboratory instrument (e.g., with its computing device).

[0021] In particular, obtaining data (e.g., log data) by the first computing device may include constructing the data based on one or more inputs, i.e., creating the data. For example, obtaining log data by the first computing device may include constructing the log data by the first computing device using a first plurality of log files. In yet another example, obtaining data (e.g., log data) by the first computing device may include accessing a first portion of the data, and generating a second portion of the data, for example, based on the first portion of the data.

[0022] In some embodiments, the step of obtaining data (e.g., log data) from a first computing device may include, for example, accessing the data by a second computing device. In particular, the data may be created and / or provided to the first computing device by the second computing device. For example, in the embodiments, the first computing device is included in a laboratory instrument or in an ALS that includes a laboratory instrument. In addition, in the embodiments, the second computing device may be a server or cloud computing system configured to: (i) evaluate the presence of anomalies in a laboratory instrument by using log data, and (ii) generate anomaly data. In particular, the second computing device is different from the first computing device.

[0023] In the present disclosure, "accessing data by a computing device (e.g., a second computing device)" may include, for example, retrieving data from at least one memory of the computing device, from a memory of another computing device (e.g., a first computing device), or from another remote data storage (database, secondary memory, cloud storage, etc.). Thus, in some cases, retrieving data may include downloading data. Exemplarily, the computing device may retrieve data in response to a user command and / or in response to a notification sent by another computing device (e.g., a first computing device). Additionally or alternatively, "accessing data by a computing device" may include, for example, receiving data from a user or another computing device. These options are not mutually exclusive. For example, accessing data by a computing device may include receiving data, storing the data in a memory of the computing device, and retrieving data by accessing the memory.

[0024] According to a first aspect, a log file is a file comprising a plurality of log file entries. A log file may be created by a laboratory instrument or an automated laboratory system during its operation, such as before, after and / or during a laboratory instrument or ALS analysis or processing of one or more samples. A log file may include one or more log file entries, each of which includes information specifying a corresponding event occurring in a laboratory instrument. The corresponding event may be related to a corresponding sample tube or a corresponding component of a laboratory instrument. For example, a log file may include a log file entry describing an event involving a sample tube and a log file entry describing an event involving another sample tube. Alternatively or additionally, a log file may include a log file entry describing an event involving a component of a laboratory instrument and a log file entry describing an event involving another component of a laboratory instrument. Alternatively, a log file may include a log file entry describing an event related to a single component of a laboratory instrument or to a single sample tube or may be composed of a log file entry describing an event related to a single component of a laboratory instrument or to a single sample tube. Alternatively or additionally, a log file may be organized in a database.

[0025] In some embodiments, the log file entries may be included in a JSON string, and / or the log file may be organized using a markup language, such as HTML, latex, etc. In this case, each log file entry may be labeled with a log file entry tag and each entry. Alternatively or additionally, the fields of the log file entry may be divided into categories and labeled in the log file with category-related field tags.

[0026] The log file may include a list of log file entries. The log file entries in the list may be sorted chronologically according to the time of creation of the log file entries, such as from the entry most recently generated to the entry most recently generated, or from the entry most recently generated to the entry most recently generated. Alternatively or additionally, the log file entries in the list may be sorted chronologically according to the time of occurrence of events described by the entries, such as from the entry describing the least recently occurring event to the entry describing the most recently occurring event, or from the entry describing the most recently occurring event to the entry describing the least recently occurring event.

[0027] For example, the event specified by the log entry can be a change in the state of a sample processed by a laboratory instrument. For example, the event specified by other log entries of the log file can be an event that a sample tube has arrived at or left a component of a laboratory instrument. In particular, the event specified by another log entry of the log file can be an event that a component of a laboratory instrument has started or completed the execution of an instruction such as for performing one or more processing steps (e.g., analysis of a sample). The event specified by other log entries of the log file can be the approval or rejection of a route plan for a sample.

[0028] In particular, one or more log file entries of a log file include a plurality of fields. For example, a log file entry may include a field containing data of an event that has occurred in a specified laboratory instrument. In addition, a log file entry may include one or more fields containing data that further describe an event that has occurred in a laboratory instrument. In addition, a log file entry may include a field containing data of a time point for specifying the creation of a log file entry, and / or a field containing data of a time point for specifying an event that has occurred in a laboratory instrument. In addition, a log file entry may include a field containing data of a sample to which the specified log file entry relates. For example, if a log file entry specifies a change in the state of a sample processed by a laboratory instrument, the log file entry includes a field containing data of a sample that uniquely determines the state of the sample.

[0029] In particular, each log file entry of the log file may be a list including one or more alphanumeric character strings, each alphanumeric character string constituting field data for a corresponding field of the log file entry. For example, the character strings of the log file entry may be separated by delimiters, such as characters such as commas, semicolons, slashes, etc.

[0030] The first field data may include a title of the corresponding event or one or more identifiers of the corresponding event. The second field includes second field data describing the corresponding event. The second field data may include a location of a component of the laboratory instrument where the corresponding event has occurred or an identifier or descriptor of a processing step performed by the laboratory instrument.

[0031] For example, if the corresponding event is a change in the state of the sample, the first field data may include or consist of an alphanumeric string specifying a change in the state of the sample that has occurred (e.g., the string "SampleStateChanged"), or a string specifying a change in the state of the sample that has occurred (e.g., the string "SampleStateChanged"). In this case, the second field data may include or consist of an alphanumeric string specifying the current state of the sample. For example, if the sample has become ready for scheduling, the alphanumeric string of the second field data may be or include the string "ReadyForSchedulingState". Alternatively, if processing of the sample has been completed, the alphanumeric string of the second field data may be or include the string "CompletedState".

[0032] For example, if the corresponding event is an event that the sample tube has reached a component of a laboratory instrument, the first field data may include an alphanumeric string specifying that the sample tube has reached a component of the laboratory instrument (e.g., the string "SiteReached") or consist of an alphanumeric string specifying that the sample tube has reached a component of the laboratory instrument (e.g., the string "SiteReached"). In this case, the second field data may include an alphanumeric string specifying which component has been reached or consist of an alphanumeric string specifying which component has been reached. For example, if the sample has reached storage, the alphanumeric string of the second field data may be the string "Storage" or include the string "Storage". Conversely, if the sample has reached analyzer AU5800, the alphanumeric string of the second field data may be the alphanumeric string "AU5800" or include the alphanumeric string "AU5800".

[0033] In particular, the step of obtaining abnormal data may be performed by the first computing device.

[0034] Alternatively, the step of obtaining the abnormal data may be performed by the second computing device. In particular, in this case, the abnormal data may be obtained from the first computing device.

[0035] In some embodiments, the step of obtaining the abnormal data, for example by the second computing device, may include generating the data, for example based on the log data. For example, obtaining the log data by the first computing device may include constructing the log data by the first computing device using the first plurality of log files. In particular, in this case, the method may include providing the abnormal data to the first computing device. In particular, in these embodiments, the step of obtaining the log data may include accessing the data and is performed by the second computing device.

[0036] In yet another example, the step of obtaining the anomaly data may include accessing a first portion of the anomaly data, and generating a second portion of the anomaly data, for example based on the first portion of the data.

[0037] In some embodiments, the step of obtaining the abnormal data may include accessing the data. In particular, in these embodiments, the step may be performed by a first computing device that obtains the abnormal data from a second computing device. In addition, in these embodiments, the second computing device may be configured to generate the abnormal data by using log data.

[0038] The selection of data for a field of a log file entry may include selecting or determining data according to a particular pattern or rule. In particular, the selection of data for a field of a log file entry may include selecting a field of a log file entry, and accessing and optionally storing data included in the field.

[0039] For example, for at least one log data item in the first plurality of log data items, the selection of the first field data of the first field of the corresponding log file entry includes selecting the field at a given position in the log file entry as the first field. For example, the log file entry of the log file can be composed of a list of fields, and can be configured so that the field at the k-th position of the list includes data of an event that has occurred in the specified laboratory instrument. In this case, for at least one log data item in the first plurality of log data items, the selection of the first field data of the first field of the corresponding log file entry includes selecting the field at the k-th position of the list as the first field, and accessing the data contained in the first field, the data contained in the first field being the first field data.

[0040] For example, for at least one log data item among the first plurality of log data items, selection of second field data of a second field of a corresponding log file entry includes determining a corresponding position in the corresponding log file entry by using the first field data, selecting the field at the corresponding position in the corresponding log file entry as the second field, and accessing data contained in the second field, the latter data being the second field data.

[0041] In particular, the corresponding position can be determined by using a first lookup table that associates possible values ​​of the first field data with positions on the log file entry. For example, if the first field data of the first field of the corresponding log entry consists of the alphanumeric string "SampleStateChanged", the corresponding position can be determined as the m1-th position in the corresponding log entry by using the first lookup table. In this case, the field at the m1-th position in the corresponding log file entry is selected as the second field. The data contained in the second field is the second field data, and can be, for example, the alphanumeric string "ReadyForSchedulingState" or the string "CompletedState". Conversely, if the first field data of the first field of the corresponding log entry consists of the alphanumeric string "SiteReached", the corresponding position can be determined as the m2-th position in the corresponding log entry by using the first lookup table. In this case, the field at the m2-th position in the corresponding log file entry is selected as the second field. In this case, the data contained in the second field is the second field data, and can be, for example, the string "Storage" or the alphanumeric string "AU5800".

[0042] Alternatively, if the log file is organized by using a markup language, the selection of the second entry may be performed by selecting an appropriate tag among tags marking entries of the log file based on the first entry.

[0043] According to a first aspect of the present invention, the log data includes a plurality of log data items L1, L2, . . . , L N-1 In particular, for each i=1, 2, ..., N-1, the log data item L i Refers to the corresponding event E that occurs in the laboratory instrument i For example, the corresponding event E i is used to construct the log data item L i The log data may include N-1 character strings, where N may be an integer greater than or equal to 1. N may be equal to 1, 2, 3, 4, 5, 6, 7, 8, etc. Each character string may correspond to a log data item included by the log data. The log data may be used to detect anomalies in laboratory instruments.

[0044] For example, the log data can be a list including a plurality of log data items. In particular, each log data item can be an alphanumeric string constructed by using the first field data of the first entry of the corresponding log entry and the second field data of the second entry of the corresponding log entry. The log data can be stored in one or more files. Additionally or alternatively, the log data can be stored in a memory of the first computing device and / or in a memory of the second computing device. The log data items included in the log data can relate to the same components of a single sample tube or a laboratory instrument.

[0045] The log data items in the log data may be sorted in the log data according to the order in which the corresponding log file entries are sorted in the log file. In particular, the log data items in the log data are sorted in chronological order such that if an event associated with a previous log data item entry occurs before an event associated with a subsequent log data item, the log data item precedes another log data item in the log data, or such that if an event associated with a previous log data item entry occurs after an event associated with the subsequent log data item, the log data item precedes another log data item in the log data. Alternatively or additionally, the log data items in the list may be sorted in chronological order such that if a corresponding log file entry associated with the previous data item has been generated before a corresponding log file entry associated with the subsequent data item, the log data item precedes another log data item in the log data, or such that if a corresponding log file entry associated with the previous data item has been generated after a corresponding log file entry associated with the subsequent data item, the log data item precedes another log data item in the log data.

[0046] The log data item may be constructed using N-1 corresponding log file entries, each consisting of a first log file entry and first N-2 log file entries, the first log file entry relating to a corresponding sample tube or a corresponding laboratory instrument component, the first N-2 log file entries following the first log file entry in the log file and relating to the corresponding sample tube or the corresponding laboratory instrument component, respectively.

[0047] The construction of the log data may include associating log file entries with log data items and / or mapping a plurality of log file entries onto a plurality of log data items according to a specific mapping.

[0048] In particular, the log data item may be constructed by using a second lookup table that associates possible values ​​of the log data item with possible values ​​of first field data and possible values ​​of second field data included in the corresponding log file entry. For example, if the first field data of the first field and the second field data of the second field of the corresponding log entry consist of the alphanumeric strings "SiteReached" and "Storage", respectively, then the log data item may be the string "StorageReached".

[0049] In some implementations, a log data item may be constructed by concatenating first field data of a first field of a corresponding log entry and second field data of a second field of the corresponding log entry. The concatenation may be performed by a delimiter, such as a dot, a comma, a slash, etc. For example, if the first field data of the first field and the second field data of the second field of the corresponding log entry are composed of the alphanumeric strings "SiteReached" and "Storage", respectively, then the log data item may be the string "SiteReached.Storage".

[0050] In particular, the possible values ​​of the log data item, such as the possible alphanumeric strings that can be constructed from the log file, constitute a log data vocabulary with a finite number of entries, each entry of the log data vocabulary is exactly one of the possible values ​​of the log data item. For example, the log data vocabulary includes the entries "SiteReached.ALI5800", "SiteReached.Storage", "SampleStateChanged.ReadyForSchedulingState", and "SampleStateChanged.CompletedState".

[0051] According to a first aspect of the invention, the anomaly data comprises information indicating the presence of an anomaly in the laboratory instrument, i.e. whether an anomaly has occurred in the laboratory instrument. Using the anomaly data, an anomaly in the laboratory instrument can be detected. The anomaly data can be generated according to the log data or based on the log data.

[0052] For example, the anomaly data may include or consist of an alphanumeric string specifying that an anomaly has occurred (e.g., an alphanumeric string "Anomalyoccurred") that specifies that an anomaly has occurred. Additionally or alternatively, the anomaly data may include or consist of a numeric value between 0 and 1 that provides an estimate of the probability that the anomaly has occurred. Additionally or alternatively, the anomaly data may include or consist of an alphanumeric string that uniquely specifies the type of anomaly that has occurred. The anomaly data may also include information specifying how the anomaly can be resolved.

[0053] The exception data may be displayed or sent to a user as a text and / or email message, and / or printed by a user, particularly to provide guidance on how to resolve an anomaly in a laboratory instrument.

[0054] An anomaly may be a fault and / or malfunction of at least one component of a laboratory instrument that may prevent the laboratory instrument from correctly performing its task and / or analysis of a sample. Such a fault and / or malfunction may be caused by a defective part of the laboratory instrument. An anomaly may include, for example, a situation where the motor of a centrifuge may not be working properly, such that a sample may not be fully centrifuged. In addition, an anomaly may be caused by an insufficient amount of a reagent loaded in a component of the laboratory instrument.

[0055] Alternatively or additionally, the anomaly may be caused by a defect of the sample tube, such as a misplaced label, the presence of contaminants in the sample, an insufficient amount of sample in the tube, etc. The anomaly may also be a degradation of the performance of a component of the laboratory instrument, such as caused by an overload of the component and / or the instrument. For example, the degradation may be an increase in the average processing time taken by the component to perform the processing step it is configured to perform. The anomaly may also be an interruption of processing or a sample, such as caused by an unexpected behavior of a component of the instrument. For example, the unexpected behavior may be an unexpected movement of a robotic arm of the instrument.

[0056] A laboratory instrument may include or consist of one or more separate instruments.

[0057] The laboratory instrument may also include or consist of an automated laboratory system. The laboratory instrument may include robotic assistance, such as a robotic arm, and / or a conveyor for transferring samples and / or transferring samples from an instrument to another instrument, such as from a stand-alone instrument to an automated laboratory system.

[0058] In an embodiment of the present invention, the laboratory instrument may include or consist of a pre-analytical instrument and / or a post-analytical instrument and / or an analytical instrument. The laboratory instrument may be, for example, a centrifuge.

[0059] Generally, an automated laboratory system (hereinafter also referred to as an "ALS") is an assembly including a plurality of components and a computing device, wherein the computing device is operably connected to the components and configured to control each component. The components may be analytical instruments, pre-analytical instruments, post-analytical instruments, input / output modules, transport components (e.g., tracks, belts, tube carriers) configured to move samples, and the like.

[0060] An analytical instrument (hereinafter also referred to as an "analyzer") is an instrument configured to perform one or more analytical steps, such as measuring one or more characteristics of a sample, such as the concentration of a given analyte. A pre-analytical instrument is an instrument configured to perform one or more pre-analytical steps on a biological sample to prepare the sample for the analytical instrument. For example, a centrifuge, an aliquoter, and a decapper are pre-analytical instruments. A post-analytical instrument is an instrument configured to perform one or more post-analytical steps on a biological sample after the sample has been processed by one or more analytical instruments. For example, a recapper and a sample storage unit are post-analytical instruments.

[0061] The laboratory instrument may be comprised by an automated laboratory system. The laboratory instrument may be part of or comprised by a laboratory or a laboratory complex or a laboratory facility. The laboratory instrument may be programmable.

[0062] In an embodiment, the step of obtaining anomaly data may include evaluating the presence of anomalies in the laboratory instrument by using the log data and generating anomaly data.

[0063] The anomaly data may be generated based on the log data. The generation may include removing or discarding log data items that are not relevant to anomaly detection, in particular log data items that do not contain information about the anomaly. The assessment of the presence of an anomaly using the log data may be accomplished by inspecting and / or assessing the log data to obtain information indicating the presence of an anomaly.

[0064] In an embodiment, the log data may be obtained from a first computing device, and the step of obtaining the anomaly data may be performed by a second computing device.

[0065] In particular, in this embodiment, the step of obtaining abnormal data may include evaluating the presence of anomalies in a laboratory instrument by using log data and generating abnormal data. In addition, in this embodiment, log data is obtained from a first computing device, for example, by a second computing device. In particular, this embodiment of the method according to the first aspect is performed by a second computing device. In this case, the first computing device may be included in the laboratory data, and the second computing device is a server or cloud computing system that performs data communication with the first computing device. For example, the second computing device can obtain (e.g., receive) log data from the first computing device, and obtain abnormal data by evaluating the presence of anomalies in a laboratory instrument using the log data.

[0066] In an embodiment, the evaluation of the presence of anomalies in a laboratory instrument can be performed by using log data as an input to at least one of a machine learning algorithm and / or a decision tree. For example, the anomaly data can be an output of a machine learning algorithm. In particular, the machine learning algorithm is a trained machine learning algorithm.

[0067] Advantageously, the application of machine learning algorithms or decision trees for identifying anomalies in laboratory instruments makes anomaly detection more accurate and therefore allows for more reliable results regarding testing and / or analyzing samples.

[0068] The machine learning algorithm can be configured to initiate resolution of the anomaly by generating resolution data based on the anomaly data, and by providing the instructions to the laboratory instrument, such as providing the instructions to the first computing device (if the first computing device is included in the laboratory instrument). In particular, the resolution data includes information indicating one or more actions for resolving the anomaly present in the laboratory instrument. For example, the resolution data may include instructions for reconfiguring and / or reprogramming the laboratory instrument and / or for rerouting one or more samples so that the anomaly is resolved or avoided.

[0069] In embodiments, the machine learning algorithm may include or consist of a neural network. The use of a neural network improves the reliability of anomaly detection because it is trained on non-anomalous data and therefore also allows detection of unforeseen and / or rare abnormal behaviors that were unknown and / or observed during the training of the machine learning algorithm.

[0070] In particular, the neural network may include at least one of a convolutional neural network, a recurrent neural network, a recurrent neural network, a long short-term memory recurrent neural network, a sequence-to-sequence model, or a shallow neural network.

[0071] These types of neural networks make anomaly detection particularly accurate. Thus, the reliability of the results regarding sample testing and / or analysis can be further improved.

[0072] In particular, the neural network may include at least one layer comprising a plurality of long short-term memory units. Alternatively or additionally, the neural network may include a dense layer as an output layer. The dense layer may include a plurality of nodes, each node including an activation function, such as a sigmoid function, a hyperbolic tangent function, a ReLu function, etc.

[0073] In an embodiment, the method may include the following steps:

[0074] - obtaining a baseline log data item L by or from the first computing device using a corresponding baseline log file entry of the log file N ,

[0075] Therein, the assessment of the presence of anomalies in the laboratory instrument is based on baseline log data items.

[0076] In particular, the baseline log data item may be such that an event associated with the baseline log data item occurs after an event associated with the log data item of the log data. Alternatively or additionally, the baseline log data item may be such that the baseline log file entry has been generated after a corresponding log file entry used to construct the log data item of the log data. For example, the baseline log data item may be compared to the output of the neural network to assess the presence of an anomaly.

[0077] In an embodiment, the step of obtaining, by the first computing device, a reference log data item may include:

[0078] - selecting first reference field data of a first field of a corresponding reference log file entry, the first field of the reference log file entry specifying a corresponding event that has occurred in the laboratory instrument;

[0079] - selecting second reference field data of a second field of a corresponding reference log file entry, the second field of the reference log file entry further describing the corresponding event that has occurred in the laboratory instrument, wherein the selection of the second field of the reference log file entry is based on the first reference field data of the first field of the reference log file entry; and

[0080] - constructing the reference log data item by using first reference field data of a first field of a corresponding reference log file entry and second reference field data of a second field of the corresponding reference log file entry.

[0081] Base fields and base field data may be specified for fields and field data of log file entries of a log file as described above.

[0082] For example, a corresponding baseline log file entry is the first log file entry of the log file that follows N-1 corresponding log file entries used to construct the log data item of the log data and relates to the sample tube or laboratory instrument component to which the log data item of the log data relates. In particular, the baseline log data item relates to the most distant recent event that (i) follows the event related to the log data item of the log data and (ii) relates to the same sample tube or laboratory instrument component to which the log data item of the log data relates. Therefore, in the absence of an anomaly in the instrument, the possible values ​​of the baseline log data item are similar to those of the log data items L1, L2, ..., L N-1 Related.

[0083] Exemplarily, the step of obtaining, by the first computing device, a baseline log data item may include selecting a baseline log file entry. In particular, selecting the baseline log file entry includes determining a first log file entry of the log file that follows N-1 corresponding log file entries used to construct the log data item of the log data and relates to the sample tube or laboratory instrument component to which the log data item of the log data relates.

[0084] In an embodiment, assessing the presence of an anomaly in a laboratory instrument using log data may include:

[0085] - generating output data by using the log data, the output data comprising information indicative of one or more expected reference log data items, and

[0086] -Evaluate the presence of anomalies in laboratory instruments by using output data and baseline log data items.

[0087] In some embodiments, the output data is the output of a neural network.

[0088] For example, each of the one or more expected baseline log data items relates to an event that, in the absence of an abnormality, (i) relates to the same sample tube or laboratory instrument component that the log data item of the log data relates to, and, in the absence of an abnormality, (ii) may immediately follow an event related to the log data item of the log data. In this case, each of the one or more expected baseline log data items is one of the log data items that is expected to immediately follow the log data item of the log data in the absence of an abnormality.

[0089] According to the present invention, if in event E A Before and events E1, E2, ..., E N-1 Each event in is not followed by events E1, E2, ..., E N-1Other events involving the same sample tube or laboratory instrument component, then the event E involving the same sample tube or laboratory instrument A Following events E1, E2, ..., E N-1 In addition, if event E F Following events E1, E2, ..., E N-1 , then with event E F Related to and involving log data items L1, L2, ..., L N-1 Log data item L for the same sample tube or laboratory instrument component involved F Followed by log data items L1, L2, ..., L N-1 .

[0090] In particular, one or more expected reference data constitute a subset of the log data vocabulary, such as a proper subset. In the following, the subset may also be referred to as a "reference subset".

[0091] For example, the reference subset may include S entries of the log data vocabulary, which are most likely to follow the log data items L1, L2, ..., L N-1 In this case, in particular, the evaluation of the presence of an anomaly comprises determining the reference log data item L N Is it included in the benchmark subset? If the benchmark subset does not include the log data item L N , then it is considered that an anomaly exists and abnormal data is generated.

[0092] In other words, according to the above-mentioned embodiment, the log data items L1, L2, . . . , L N-1 To generate an output ("output data"), which can be based on L1, L2, ..., L N-1 Specify L N One or more predictions of the string. The output data can be compared to L using, for example N to detect anomalies. For example, it may be possible to check the output data and log data items L N Are they compatible with each other.

[0093] In an embodiment, the output data may include information indicating a probability that the baseline log data item is equal to one or more expected baseline log data items.

[0094] In particular, the output data may include information specifying the reference subset and specifying for each entry in the reference subset that, in the absence of anomalies, each entry will immediately follow the log data items L1, L2, . . . , L N-1 In this case, in particular, the evaluation of the presence of an anomaly comprises determining the reference log data item L N Whether it is included in the benchmark subset and / or related to LN Whether the associated probability is greater than or equal to a given threshold or is among the T terms with the highest probability.

[0095] In other words, according to this embodiment, the output data may represent the probability that the Nth string of the log data is equal to the prediction of the Nth string based on the N-1 strings of the log data. For example, if the probability is below a predetermined threshold, an abnormality may be detected.

[0096] In an embodiment, if an anomaly has been assessed to exist in the laboratory instrument, the method may further comprise the step of obtaining resolution data comprising information indicative of one or more actions for resolving the anomaly present in the laboratory instrument.

[0097] The resolution data may include one or more suggestions on how to resolve the anomaly. For example, the detected anomaly may be that the robot arm moves the sample too fast and / or pushes the sample too hard. Thus, the resolution data may, for example, instruct an operator to check whether the robot arm is working as expected. In particular, it may include suggestions to readjust, for example, a joint of the robot arm. In particular, the anomaly data may include, for example consist of, the resolution data.

[0098] The step of obtaining the anomaly data may be performed by the second computing device. In particular, in this case, the resolution data may be obtained from the first computing device, for example by receiving or retrieving said data.

[0099] In some embodiments, the step of obtaining the solution data may include accessing the data. In particular, in these embodiments, the step may be performed by a first computing device that obtains the solution data from a second computing device. Additionally, in these embodiments, the second computing device may be configured to generate the solution data using log data and / or exception data.

[0100] The resolution data can be printed and / or displayed to the user, thereby advising the user how to resolve the anomaly, particularly what actions to take to resolve the anomaly. In this way, the user is guided in the correct operation and / or correct maintenance of the laboratory instrument. The resolution data can also be used by the machine learning algorithm to resolve the anomaly detected by the machine learning algorithm, for example, by modifying the route plan of one or more sample tubes to bypass the laboratory instrument in which the anomaly is detected. In this way, the automated laboratory system automatically tests and / or analyzes samples and detects and / or corrects anomalies, such as errors in one or more laboratory instruments thereof.

[0101] In an embodiment, the step of obtaining solution data may include generating solution data by using the log data, and optionally providing the solution data to the first computing device. In particular, in this case, the step of obtaining solution data may be performed by the second computing device.

[0102] The solution data may be generated by applying a machine learning algorithm to the log data, in particular a neural network or a decision tree as specified above. The solution data may be printed and / or displayed via the first computing device.

[0103] In another embodiment, the output of the machine learning algorithm may include or be an error code. In particular, the training of the machine learning algorithm may be performed using anomaly log data items and non-anomaly log data items, and the output of the trained machine learning algorithm may be explicitly associated with an error code that specifies the presence or absence of an anomaly and, if the anomaly is present, the type of anomaly. Diagnostics for resolving the anomaly being detected may be performed through a lookup table that associates an error code with a possible solution.

[0104] Herein, providing data to a computing device may be performed by making the data available to the computing device, such as by storing the data in a memory included in or accessible by the computing device, and / or by granting the computing device access to the data so that the computing device can retrieve, copy and / or access the data. For example, providing data to a computing device may include sending the data to the computing device, such as via a network. Providing data to a computing device may include sending a link to the data to the computing device, such as via a network.

[0105] In an embodiment, the step of obtaining log data by the first computing device may include:

[0106] - Access log files of laboratory instruments; and

[0107] - construct log data,

[0108] The step of constructing log data includes performing the following operations on each log data item in the first plurality of log data items:

[0109] ○ Select the first field data of the first field of the corresponding log file entry;

[0110] o selecting second field data of a second field of the corresponding log file entry, wherein the selection of the second field is based on first field data of a first field of the corresponding log file entry;

[0111] as well as

[0112] o Each log data item is constructed by using first field data of a first field of a corresponding log file entry and second field data of a second field of the corresponding log file entry.

[0113] The log file may be accessed at a shared site, such as a network storage, in particular a cloud storage. In particular, the log file may be accessed at a memory included in the laboratory instrument. For example, if the first computing device is included in the laboratory instrument, the log file may be stored in the memory of the first computing device and accessed at the memory of the first computing device.

[0114] In an embodiment, the method according to the first aspect may further include the following steps:

[0115] - providing the log data by the first computing device to a third computing device, the third computing device being configured to assess the presence of an anomaly in the laboratory instrument by using the log data,

[0116] And wherein, the step of obtaining log data by the first computing device may include:

[0117] - Access log files of laboratory instruments; and

[0118] - constructing log data, wherein the step of constructing log data comprises, for each log data item in the first plurality of log data items, performing the following operations:

[0119] ○ Select the first field data of the first field of the corresponding log file entry;

[0120] o selecting second field data of a second field of the corresponding log file entry, wherein the selection of the second field is based on first field data of a first field of the corresponding log file entry;

[0121] as well as

[0122] o Each log data item is constructed by using first field data of a first field of a corresponding log file entry and second field data of a second field of the corresponding log file entry.

[0123] According to this embodiment, in particular, the first computing device constructs log data by selecting relevant information so that correlations between log file entries of the log file are indicated. In addition, by distributing the method of the first aspect on the first computing device and the third computing device, which may be different from each other, anomaly detection can be performed in a computer resource-saving manner, in particular a memory-efficient manner. Therefore, the reliability of the sample test and / or analysis results can be increased in an efficient manner with respect to the use of computer hardware.

[0124] In particular, the third computing device may be the second computing device described herein. In particular, in this embodiment, the step of obtaining the abnormal data may be performed by the first computing device. For example, the abnormal data may be obtained from a third computing device, such as the second computing device. Additionally, the step of obtaining the abnormal data by the first computing device may include, for example, receiving the data from the third computing device. The step of obtaining the abnormal data by the first computing device may include retrieving the data from a memory of the third computing device (e.g., the second computing device) and / or accessing the data at a memory of the third computing device (e.g., the second computing device).

[0125] In particular, in this embodiment, the step of obtaining anomaly data may include evaluating the presence of anomalies in the laboratory instrument by using log data and generating anomaly data. In addition, in this embodiment, log data is obtained from the first computing device, for example by the second computing device.

[0126] In particular, this embodiment of the method according to the first aspect is performed by a first computing device. In this case, the first computing device can be included in a laboratory instrument, and the second computing device is a server or cloud computing system that communicates data with the first computing device. In particular, the first computing device can construct log data and provide the data to the second computing device, and the second computing device is configured to evaluate the presence of anomalies. If anomalies exist, the first computing device obtains, for example, receives and / or retrieves the abnormal data from the second computing device that has generated the data.

[0127] In an embodiment, constructing the log data may include, for each log data item in the first plurality of log data items, performing the following operations:

[0128] - selecting time data of a respective third field of a respective log file entry, the time data comprising information indicating a time when the respective log file entry has been created and / or a time when a respective event described by the first field and / or the second field of the respective log file entry has occurred;

[0129] The construction of each log data item may be performed by using time data of a corresponding third field of a corresponding log file entry.

[0130] Advantageously, including time data when constructing a log data item for evaluating anomalies in a laboratory instrument allows tracking of the duration of a processing step to be kept, and allows detection of anomalies that result in a delay in the completion of one or more processing steps. In this way, anomalies may be detected more accurately and / or a greater range of anomalies may be detected.

[0131] The time data may include at least one of a timestamp, a system time of a computing device of a laboratory analyzer and / or a computing device collecting log file entries, a point in time in hours and / or minutes and / or seconds, a time interval, a day, a month, a year, or a date.

[0132] The third field and corresponding field data may be specified as described above with respect to the first and second fields and their corresponding field data.

[0133] In this embodiment, the method also takes into account the time of the log file entries. In particular, the log data, which may be data to be fed to the machine learning algorithm, may include N-1 strings and N-1 corresponding times, such that each string and each corresponding time may be obtained from a corresponding log file entry of the log file. In this case, the string entry and the corresponding time may be together in the same input.

[0134] In particular, by using relative time data, wherein the relative time data can be generated by using time data and can include information indicating the relative time at which the corresponding log file entry has been created and / or the time at which the corresponding event described by the corresponding log file entry has occurred. The relative time can relate to a given time, for example, with respect to a log file entry having an earliest time entry.

[0135] In particular, in this embodiment, the log data includes a plurality of entries D1, D2, . . . , D N-1 or a list with multiple entries D1, D2, ..., D N-1 For each i=1, 2, ..., N-1, the entry D i It also includes the log data item L i and the relative time value ΔT i In particular, for each i=1, 2, ..., N-1, ΔT i =T i -T R , where T i It is event E i The time point at which the E i The time point of the associated log entry. Time point T R is a reference time point, which may be, for example, time points T1, T2, ..., T N-1 One of the , for example the smallest one.

[0136] Similarly, other "parameters" (ie, values) of the log file entry may be considered. For example, if the laboratory instrument includes a robot, the value may be the current of the robot which may be correlated with other parameters present in the log file entry.

[0137] In an embodiment, the log data may include a second plurality of log data items, and constructing the log data may include, for each log data item in the second plurality of log data items, performing the following operations:

[0138] - selecting time data of a respective third field of a respective log file entry, the time data comprising information indicating a time when the respective log file entry has been created and / or a time when a respective event described by the first field and / or the second field of the respective log file entry has occurred;

[0139] - constructing a respective log data item of the second plurality of log data items by using the time data of the respective third field of the respective log file entry.

[0140] Similar to the previous embodiments, this embodiment allows for more accurate anomaly detection and / or allows for detection of a wider range of anomalies, such as anomalies that cause a delay in the completion of one or more processing steps.

[0141] In the currently described embodiment, the method also takes into account the time of the log entries. In particular, the log data that can be fed to the machine learning algorithm (particularly the neural network) can include N-1 strings and N-1 corresponding times, so that each string and each corresponding time can be obtained from the corresponding log file entry of the log file. In this case, the string entry and the corresponding time can be together in the same input.

[0142] In particular, by using relative time data, wherein the relative time data can be generated by using time data and includes information indicating the relative time at which the corresponding log entry has been created and / or the time at which the corresponding event described by the corresponding log file entry has occurred. The relative time can be related to a given time, for example, to the log file entry with the earliest time entry.

[0143] In particular, in this embodiment, the log data may include log data items L1, L2, . . . , L N-1 The first list and the plurality of values ​​ΔT1, ΔT2, ..., ΔT N-1 For example, the first list can be processed by a machine learning algorithm (such as a long short-term memory recurrent neural network as described above), which processes the first list to generate output data.

[0144] The second list can be processed by other machine learning algorithms, such as a feed-forward neural network. The latter algorithm can be a binary classifier that is trained to classify its input (e.g., the second list) as anomalous or non-anomalous. In particular, if the relative time values ​​ΔT1, ΔT2, ..., ΔT N-1If one or more of is outside the corresponding range, other machine learning algorithms can be trained to classify its input as abnormal.

[0145] In some embodiments, the first list and the second list can be processed by the same neural network. For example, the two lists can be combined in a single input of the neural network, or projected into two different layers.

[0146] In an embodiment, each log data item in the first plurality of log data items may be associated with a single log file entry in the plurality of log file entries, and each log data item in the second plurality of log data items may be associated with a single log file entry in the plurality of log file entries.

[0147] Log data items may be associated with log file entries by specific patterns and / or rules and / or mappings.

[0148] In an embodiment, each log file entry in the first plurality of log file entries may relate to a corresponding sample tube. Additionally or alternatively, for each log data item in the first plurality of log data items, the corresponding log file entry may relate to a given sample tube.

[0149] Herein, the sample tube may be a container for containing a biological sample such as body fluid including blood, saliva, urine, and the like.

[0150] The biological sample can be a sample of a subject's body fluid. For example, the body fluid can be a physiological fluid, such as blood, saliva, urine, sweat, amniotic fluid, cerebrospinal fluid, ascites, etc. The biological sample can include one or more components, which can potentially include an analyte of interest for performing a clinical test on the sample. The clinical test can include one or more procedures that, when performed on the biological sample or its components, allow the value of a parameter, such as a clinical parameter, to be estimated. In particular, the clinical test can include physical, biological, optical, mechanical, immunological and / or chemical procedures.

[0151] The sample tube can be a cylindrical or cubic container containing one or more of the above-mentioned substances. The sample tube can be a small glass tube. The sample tube can be closable and / or sealable. The sample tube can include a lid to close and / or seal the container containing one or more of the above-mentioned substances. The sample tube can be configured to be positioned at or held by a laboratory instrument, such as, for example, a centrifuge. The sample tube can be placed in a sample tube rack, which is configured to hold a plurality of sample tubes.

[0152] Herein, the term "corresponding sample tube" indicates that each log file entry may relate to a corresponding sample tube. Typically, different log file entries may relate to different sample tubes. Alternatively or in addition, the term "given sample tube" indicates that each log file entry may relate to the same sample tube, i.e., different log file entries may relate to the same sample tube.

[0153] In this embodiment, the method allows parallelization, e.g., multithreading, by recording and evaluating anomalies by reviewing events related to several tubes in the same log file. In particular, the log data that can be fed to the machine learning algorithm can relate to the same tube. Tube parallelization simplifies and / or improves the assessment of the presence of anomalies, e.g., it allows distinguishing tube anomalies from instrument anomalies and / or allows detection of problematic tubes.

[0154] In an embodiment, constructing the log data may include, for each log data item in the first plurality of log data items, evaluating whether the corresponding log file entry relates to a given sample tube. For example, if the corresponding log file entry relates to a given sample tube, constructing the log data includes selecting the corresponding log file entry.

[0155] According to this embodiment, when constructing log data, the first computing device may evaluate whether log file entries used to construct the log data items relate to the same sample tube.

[0156] In an embodiment, each log file entry in the first plurality of log file entries may include a fourth field including sample tube data, wherein the sample tube data may include information indicating the respective sample tube to which each log file entry relates. Additionally, for each log data item in the first plurality of log data items, evaluating whether the respective log file entry relates to a given sample tube may be performed using the sample tube data included in the fourth field of the respective log file entry.

[0157] The fourth field may be specified similarly to that described above for the first field, the second field, and the third field.

[0158] The sample tube data may include data identifying the patient and / or sample contained in the sample tube and / or the test and / or analysis to be performed on the sample tube. For example, the tube data includes, for example, consists of, an alphanumeric string that uniquely identifies the sample tube to which the log entry relates.

[0159] In this embodiment, each log file entry of the log file may include a tube field ("fourth field") having a barcode of the tube to which it relates. In addition, when constructing log data, the first computing device may use the tube field of the log file entry to evaluate whether the log file entries used to construct the log data item relate to the same tube.

[0160] In an embodiment, for at least a first log data item of the first plurality of log data items, a first field of the corresponding log file entry may specify that the sample tube has arrived at a location of the laboratory instrument, and a second field of the corresponding log file entry may include information specifying the location at which the sample tube has arrived.

[0161] For example, the first field may consist of the string "SiteReached". If this is the case, the second field may consist of one of the strings "TrackConnection", "LeftCentrifuge", "AnalyserAU5800", "InputRegion", "OutputRegion", "Storage", "Decapper", "Recapper", etc., depending on the location reached.

[0162] Additionally or alternatively, for at least a second log data item of the first plurality of log data items, a first field of the corresponding log file entry may specify a location from which the sample tube has left the laboratory instrument, and a second field of the corresponding log file entry may include information specifying the location from which the sample tube has left.

[0163] For example, the first field may consist of the string "SiteExited". If this is the case, the second field may consist of one of the strings "TrackConnection", "LeftCentrifuge", "AnalyserAU5800", "InputRegion", "OutputRegion", "Storage", "Decapper", "Recapper", etc., depending on the location the sample tube has left.

[0164] Additionally or alternatively, for at least a third log data item of the first plurality of log data items, the first field of the corresponding log file entry may specify that a status of the sample tube has changed, and the second field of the corresponding log file entry may include information specifying the current status of the sample tube.

[0165] For example, the first field may consist of the string "SampleStateChanged". If this is the case, the second field may consist of one of the strings "ReadyForSchedulingState", "WaitForNewRoutePlanState", "Runningstate", "CompletedState", "SampleAtTimeoutSiteState", etc., depending on the current state of the sample tube.

[0166] Additionally or alternatively, for at least a fourth log data item of the first plurality of log data items, the first field of the corresponding log file entry may specify a task start of the laboratory, and the second field of the corresponding log file entry may include information specifying which task started.

[0167] For example, the first field may consist of the string "InstructionStarted". If this is the case, the second field may consist of one of the strings "DecapInstruction", "RecapInstruction", "CentrifugeInstruction", "AnalyserAU5800Instruction", etc., depending on the task started.

[0168] Additionally or alternatively, for at least a fifth log data item in the first plurality of log data items, the first field of the corresponding log file entry may specify task completion of the laboratory, and the second field of the corresponding log file entry may include information specifying which task was completed.

[0169] For example, the first field may consist of the string "InstructionFinished". If this is the case, the second field may consist of one of the strings "DecapInstruction", "RecapInstruction", "CentrifugeInstruction", "AnalyserAU5800Instruction", etc., depending on the task started.

[0170] Additionally or alternatively, for at least a sixth log data item of the first plurality of log data items, a first field of the corresponding log file entry may specify that a route plan for the sample tube has been evaluated and a second field of the corresponding log file entry may include information specifying whether the route plan has been accepted or rejected.

[0171] For example, the first field might consist of the string "ProcessRoutePlanAcceptance". If this is the case, the second field might consist of the string "True" or "False", depending on whether the route plan has been accepted or rejected.

[0172] In particular, a location of a laboratory may refer to an area and / or a component of a laboratory device. For example, a location, ie an area or a component of a laboratory device, may be a track of an analyzer, a pre-processing instrument, a post-processing instrument or a laboratory instrument.

[0173] The term "status" may refer to the status of a sample with respect to testing, processing, and / or analysis of the sample, ie, for example, whether the sample is awaiting testing or whether a particular test has already been performed.

[0174] According to the invention, a task may be the execution of a processing step by a laboratory instrument, such as a test step and / or an analysis step for testing and / or analyzing a sample tube, respectively. For example, it may refer to the execution of a specific test (e.g., PCR test, etc.) and / or analysis (e.g., spectral analysis, etc.). For example, a task may be the execution of a processing step, such as a step of centrifuging, decapping, recapping, storing, etc.

[0175] Typically, a route plan of a laboratory instrument (eg, an ALS) includes information indicating how the laboratory instrument (eg, an ALS) is to process a biological sample. In other words, the route plan includes one or more processing steps, such as their time sequence.

[0176] In an embodiment, the method may further include the step of printing and / or displaying the log data and / or the exception data to a user.

[0177] Thereby, guidance may be provided to the user on how to resolve an anomaly detected by the method according to the first aspect of the present invention or any of the corresponding embodiments.

[0178] In an embodiment, constructing the log data may include, for each log data item in the first plurality of log data items, performing the following operations:

[0179] ○ Select the fifth field data of the fifth field of the corresponding log file entry, the corresponding log file entry

[0180] The fifth field of the purpose further describes the corresponding event that has occurred in the laboratory instrument, where:

[0181] The selection of the fifth field is based on the first field of the corresponding first field of the corresponding log file entry

[0182] data and / or second field data of a corresponding second field; and

[0183] The construction of each log data item is performed by using fifth field data of a fifth field of a corresponding log file entry.

[0184] In particular, for at least one log data item among the first plurality of log data items, selection of fifth field data of the fifth field of the corresponding log file entry includes determining a corresponding position in the corresponding log file entry by using the first field data and / or the second field data, thereby selecting the field at the corresponding position in the corresponding log file entry as the fifth field, and accessing the data contained in the fifth field, the latter data being the fifth field data.

[0185] The fifth field data may, for example, include, for example, consist of, an alphanumeric string. For example, if the first field data consists of the string "SiteReached", and the second field data is the string "Recapper", then the fifth field may be the string "RecapStop". For example, in this case, the resulting log data item may be constructed by concatenating the above strings separated by dots, for example, the log file data item may consist of the string "SiteReached.Recapper.RecapStop".

[0186] In particular, in some embodiments, for at least a first log data item in the first plurality of log data items, a first field of the corresponding log file entry specifies that the sample tube has reached the location of the laboratory instrument, and a second field of the corresponding log file entry includes information specifying the location at which the sample tube has reached. In the embodiment, for at least a second log data item in the first plurality of log data items, a fifth field of the corresponding log file entry specifies the status of the task performed at the location specified in the second field. For example, if the first field and the second field consist of the strings "SiteReached" and "Decapper", accordingly, depending on the status of the centrifugation task, the fifth field can consist of one of the strings "Decap", "DecapStop", etc.

[0187] A second aspect of the present invention relates to a computer-implemented method comprising the following steps:

[0188] - providing a log file of the laboratory instrument to a computing device, the log file comprising a first plurality of log file entries, each log file entry in the first plurality of log file entries comprising a plurality of fields;

[0189] - obtaining anomaly data comprising information indicative of the presence of an anomaly in the laboratory instrument,

[0190] The computing device is configured to construct log data by performing the following operations for each log data item in the first plurality of log data items, the log data comprising the first plurality of log data items:

[0191] ○Select the first field data of the first field of the corresponding log file entry, the corresponding log file entry

[0192] The purpose first field specifies the corresponding event that has occurred in the laboratory instrument;

[0193] ○Select the second field data of the second field of the corresponding log file entry, the corresponding log file entry

[0194] The second field of the purpose further describes the corresponding event that has occurred in the laboratory instrument, where

[0195] The selection of the corresponding second field is based on the first

[0196] Field data; and

[0197] ○ By using the first field data of the first field of the corresponding log file entry and the corresponding log

[0198] The computing device is configured to evaluate the presence of an anomaly in the laboratory instrument using the log data.

[0199] All advantages, specifications, descriptions, additional features and embodiments of the previous aspects also apply to the second aspect of the invention. In particular, the method according to the second aspect of the invention allows better, faster and / or easier processing of data and / or better, faster and / or easier detection of anomalies. Thus, the reliability of the results of the sample analysis provided by the laboratory instrument of, for example, an automated laboratory system is increased.

[0200] In particular, the method according to the first aspect of the present invention can be performed by a computing device included in a laboratory instrument. According to a second aspect, for example, the computing device of the laboratory instrument provides a log file to a server that performs the method according to the first aspect of the present invention. Additionally or alternatively, any of the features of the second computing device according to the first aspect of the present invention can be applied to the computing device according to the second aspect of the present invention. In some embodiments, the step of obtaining abnormal data may include accessing the data. In particular, the data can be created and / or provided by a computing device.

[0201] All methods according to the first and / or second aspects of the invention may be performed on-the-fly while samples are being processed in a laboratory comprising one or more laboratory instruments.

[0202] A third aspect of the present invention relates to a computer-implemented method comprising the following steps:

[0203] - accessing a log file of a laboratory instrument, the log file comprising a first plurality of log file entries, each log file entry in the first plurality of log file entries comprising a plurality of fields;

[0204] - constructing labeled training log data, the labeled training log data comprising a plurality of training log data items and labeled data; and

[0205] - initiating training of the machine learning algorithm, wherein the training of the machine learning algorithm is performed by using labeled training log data,

[0206] The step of constructing the labeled training log data includes performing the following operations on each of the plurality of training log data items:

[0207] ○Select the first field data of the first field of the corresponding log file entry, the corresponding log file entry

[0208] The purpose first field specifies the corresponding event that has occurred in the laboratory instrument;

[0209] ○Select the second field data of the second field of the corresponding log file entry, the corresponding log file entry

[0210] The second field of the purpose further describes the corresponding event that has occurred in the laboratory instrument, where

[0211] The selection of the second field is based on first field data of the first field of the corresponding log file entry;

[0212] as well as

[0213] ○ By using the first field data of the first field of the corresponding log file entry and the corresponding log

[0214] Each training log data item is constructed using the second field data of the second field of the file entry.

[0215] According to an embodiment, the method comprises the following steps: applying a trained machine learning algorithm, in particular by using log data as input to the machine learning algorithm to identify anomalies in the laboratory instrument.

[0216] According to an embodiment, log data is constructed from log file entries by associating log file entries with log data items and / or by mapping a plurality of log file entries onto a plurality of log data items according to a specific mapping.

[0217] All advantages, specifications, descriptions, additional features and embodiments of the previous aspects also apply to the second aspect of the invention. In particular, the method according to the second aspect of the invention allows better, faster and / or easier processing of data and / or better, faster and / or easier detection of anomalies. Thus, the reliability of the results of the sample analysis provided by the laboratory instrument of, for example, an automated laboratory system is increased.

[0218] For example, the additional features relating to the log data items according to the first aspect of the invention may also be applied to the training log data items according to the third aspect of the invention. In particular, the additional features relating to the construction of the log data items according to the first aspect of the invention may also be applied to the construction of the training log data items according to the third aspect of the invention.

[0219] The labeled data may include information indicating a true value (ground truth) associated with a plurality of training log data items. In particular, the ground truth specifies the expected output that the training machine learning algorithm needs to generate by processing a plurality of log data items. For example, the labeled data includes information indicating one or more expected baseline log data items. In particular, any of the various features of one or more expected baseline data items of the first aspect of the present invention may be included in the features of one or more expected baseline log data items of the third aspect of the present invention, or combined with the features of one or more expected baseline log data items of the third aspect of the present invention.

[0220] For example, for each of the one or more expected baseline log data items, the marker data may include the probability that the corresponding event associated with each of the expected baseline log data items immediately follows the event related to the training log data item in the absence of anomalies. In this case, each of the one or more expected baseline log data items is one of the log data items that is expected to immediately follow the training log data item in the absence of anomalies. In some embodiments, the marker data specifies the log data item that immediately follows the training log data item. For example, the marker data may consist of the log data item that immediately follows the training log data item.

[0221] In particular, the information included in the labeled data may be the same as the information included in the output data of the first aspect of the invention. For example, during training, the information included in the labeled data is compared with the information included in the output of the machine learning algorithm. Typically, the training is performed by using a plurality of labeled log data.

[0222] In some embodiments, a plurality of labeled training log data are constructed by obtaining a non-abnormal log file and by grouping the log file entries into a plurality of training groups each containing N log file entries. In particular, for each training group, the N log file entries of the group are composed of a first log file entry and the first N-1 log file entries, wherein the first log file entry relates to a corresponding sample tube or a corresponding laboratory instrument component, and the first N-1 log file entries are after the first log file entry in the log file and relate to the corresponding sample tube or the corresponding laboratory instrument component. In particular, each labeled training log data in the plurality of labeled training log data is obtained by a corresponding training group in the plurality of training groups. In particular, training log data items are constructed according to the first N-1 log file entries of the corresponding training group. The expected benchmark log data items constituting the labeled data are constructed by using the Nth entry of the corresponding training group. Then, the machine learning algorithm is trained using the plurality of labeled training log data.

[0223] All advantages, specifications, descriptions, additional features and embodiments of the previous aspects also apply to the third aspect of the invention.

[0224] The initiation of training of the machine learning algorithm may include instructing a computing device, such as a client device and / or a server, to perform some actions to start training. The initiation may include providing training data to the computing device.

[0225] In an embodiment of the third aspect, the step of initiating training of the machine learning algorithm can be performed by providing labeled training log data to a fourth computing device, and the fourth computing device is configured to train the machine learning algorithm by using the labeled training log data.

[0226] In particular, the fourth computing device may be the second computing device described herein.

[0227] In an alternative embodiment, the step of initiating training of the machine learning algorithm may be performed by training the machine learning algorithm using labeled training log data.

[0228] All methods according to the first aspect and / or the second aspect and / or the third aspect of the present invention may be performed on-the-fly while a sample is being processed in a laboratory comprising one or more laboratory instruments.

[0229] A fourth aspect of the present invention relates to a data processing system comprising a processor configured to execute the method according to one of the preceding aspects or corresponding embodiments.

[0230] The system may comprise at least one of a first computing device, a second computing device, a third computing device or a fourth computing device. In particular, at least two of the second computing device, the third computing device and the fourth computing device may be identical. All computing devices may be different from each other. The computing devices may in particular be located in different laboratories or laboratory locations. In particular, the first computing device may be comprised by or be part of a laboratory instrument.

[0231] A fifth aspect of the invention relates to a laboratory instrument comprising a data processing system according to the fourth aspect.

[0232] A sixth aspect of the invention relates to an automated laboratory system comprising a data processing system according to the fifth aspect.

[0233] A seventh aspect of the present invention relates to a computer program product comprising instructions, which, when executed by a computer system, causes the computer system to perform a method according to any of the first aspect, the second aspect, the third aspect, or the corresponding embodiments.

[0234] An eighth aspect of the present invention relates to a computer-readable medium comprising instructions, which, when executed by a computer system, causes the computer system to perform a method according to any of the first aspect, the second aspect, the third aspect, or the corresponding embodiments.

[0235] All advantages, specifications, descriptions, additional features and embodiments of the first, second and third aspects also apply to the fourth, fifth, sixth, seventh and eighth aspects of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0236] In the following, the drawings are described. The drawings depict one or more embodiments of the present invention. The various features depicted in the drawings may be combined into other embodiments. The drawings show:

[0237] Figure 1 : schematic representation of an exemplary first computing device and an exemplary second computing device according to aspects of the present invention;

[0238] Figure 2 : a first swim lane diagram including exemplary steps performed by a first computing device and a second computing device;

[0239] Figure 3 : a second swim lane diagram including exemplary steps performed by the first computing device and the second computing device;

[0240] Figure 4 : a schematic representation of an excerpt from an exemplary log file; and

[0241] Figure 5 : Schematic representation of an excerpt of exemplary log data. DETAILED DESCRIPTION

[0242] In the following, method steps denoted by the same reference numerals are performed in the same manner.

[0243] Figure 1 ( Figure 1 ) shows a schematic representation of an exemplary first computing device 110 and an exemplary second computing device 120 according to aspects of the present invention.

[0244] exist Figure 1 In the illustrated embodiment, the first computing device 110 is a cluster computer located at a laboratory 100, which includes a set of laboratory instruments 120 for processing biological samples. The first computing device 110 may also be a desktop PC, a server, a distributed computing system, or a cloud computing system.

[0245] The first computing device 110 includes a first processor 111, a first memory 112 or storage unit, a first I / O interface 113, and a first interface controller 114 (NIC). The first processor 111 is connected to the first memory 112 and the first NIC 114. The first NIC 114 is connected to the first I / O interface 113. The first I / O interface 113 is configured to send output of the first computing device 110 to at least one second computing device 210 or receive input from at least one second computing device 210 by means of a protocol suite 12 for exchanging data. Figure 1 , only one second computing device 210 is shown.

[0246] Figure 1 The second computing device 210 shown is located at a location different from the location of the laboratory 100. The second computing device 210 is located at the laboratory 100. The second computing device 210 can be a desktop PC, a workstation, a cluster computer or a server. The second computing device 210 can be included by one or more laboratory instruments 220 in the laboratory 100, or a computing unit of one or more laboratory instruments 220 in the laboratory 100. The second computing device 210 includes a second processor 211, a second memory 212 or storage unit, a second I / O interface 213 and a second NIC 214. The second NIC 214 is connected to the second processor 211 and the second I / O interface 213. The second processor 211 is also connected to the second memory 212.

[0247] The first computing device 110 can output instructions to or receive data from a set of laboratory instruments 120 located at a location in the laboratory 100. Input and output are transmitted between the first computing device 110 and the set of laboratory instruments 120 by means of a protocol suite 11 for exchanging data. In some embodiments, the first computing device 110 can be included in an automated laboratory system (not shown) that includes the set of laboratory instruments 120.

[0248] The first computing device 110 may send data and / or instructions to the second computing device 210, and vice versa, via the protocol suite 12. The first computing device 110 may receive the instructions and initiate the performance of one or more clinical tests on the sample by instructing one or more of the set of laboratory instruments 120 at the laboratory 100. The set of laboratory instruments 120 may, for example, be included in an automated laboratory system that includes the first computing device 110 or is in data communication with the first computing device 110.

[0249] The first computing device 110 and the second computing device 210 may include more than one processor 111, 211 for task parallelization. The first memory 112 and the second memory 212 may include or be RAM and / or a magnetic hard drive and / or solid state storage and / or cloud storage.

[0250] Figure 2 ( Figure 2 ) shows a first swim lane diagram including exemplary steps performed by a first computing device and a second computing device. In this embodiment, the second computing device 210 performs an embodiment of the method according to the first aspect of the present invention. In particular, the second memory 212 stores a computer program including instructions, which, when executed by the processor 211, causes the second computing device 210 to perform the method according to the first aspect of the present invention.

[0251] In step 310, particularly in the first step, as Figure 1 The first computing device 110 is shown accessing a log file of a set of laboratory instruments 120. In particular, the first computing device can generate a log file by using log information provided by the set of laboratory instruments 120.

[0252] Figure 4 ( Figure 4 ) shows a schematic representation of an excerpt 500 of an exemplary log file. In particular, the log file is a text file and includes a plurality of log file entries 510, 520, 530, 540, each log file entry being a row of the log file. The log file entries 510, 520, 530, 540 are arranged in chronological order in the log file according to the creation time of the log file entries, for example from the entry farthest from the most recently generated to the entry most recently generated. In particular, the entries of the log file can be arranged according to their timestamps. In some cases, for example, if the log file entries are generated by different instruments in a set of laboratory instruments 120, two or more log file entries can have the same timestamp. In particular, in this case, the two or more log file entries are randomly arranged relative to each other and are arranged in chronological order relative to other log file entries.

[0253] Each entry of the log file entry consists of a plurality of fields 511, 512, 521, 522, and each field of the log entries 510, 520, 530, 540 contains an alphanumeric string including the corresponding field data. The fields of each log entry are separated from each other by one or more spaces. For example, log entry 530 includes the fields "12:56:50.164.UTC", "SiteExited", "Pc|Mc|SiteReached Exited", "Decapper.DecapStop", and "Decapper.Decap".

[0254] Reference Figure 2 At step 311, the first computing device 110 constructs log data and a reference log data item L N The log data includes a first plurality of log data items, L1, L2, ..., L N-1 . Figure 5 ( Figure 5 ) shows a schematic representation of an excerpt of exemplary log data 600. In particular, the log data is collected in a text file including a plurality of log file entries 610, 620, each log data item including an alphanumeric string. The log data items 610, 620 are arranged according to the order of the corresponding log entries 610, 620. In particular, the construction of the log data includes, for each log data item 610, 620 of the first plurality of log data items, performing the following operations:

[0255] ○ Select the first field number of the first field 511, 521 of the corresponding log file entry 510, 520

[0256] The first fields 511, 521 of the corresponding log file entries 510, 520 specify the

[0257] The corresponding events that have occurred in the laboratory instrument 120;

[0258] o Select the second field number of the second field 512, 522 of the corresponding log file entry 510, 520

[0259] According to the second fields 512, 522 of the corresponding log file entries 510, 520 are further described in

[0260] corresponding events occurring in the set of laboratory instruments 120; and

[0261] o By using the first word of the first field 511, 521 of the corresponding log file entry 510, 520

[0262] Each log data item 610 , 620 is constructed by combining the segment data and the second field data of the second field 512 , 522 of the corresponding log file entry 510 , 520 .

[0263] According to the present embodiment, for each log data item 610, 620 of the first plurality of log data items, the selection of the first field data of the first field 511, 521 of the corresponding log file entry 510, 520 includes selecting the field at the second position in the log file entry as the first field, the field specifying an event that has occurred at or in the set of laboratory instruments 120. For example, the field at the second position of the log entry 510 is a string "Site Reached" 511, which specifies that the sample tube has reached the location of the set of laboratory instruments 120. The field at the second position of the log entry 520 is a string "InstructionStarted" 521, which specifies that a task of the laboratory has started.

[0264] According to this embodiment, for each log data item 610, 620 in the first plurality of log data items, selection of the second field data of the second field 512, 522 of the corresponding log file entry 510, 520 is performed by determining the corresponding position in the corresponding log file entry 510, 520 using the first field data, thereby selecting the field at the corresponding position in the corresponding log file entry 510, 520 as the second field 512, 522.

[0265] The corresponding position can be determined by using a first lookup table (not shown) that associates possible values ​​of the first field data with positions on the log file entry. For example, according to the first lookup table, if the first field data is the string "Site Reached" 511, the corresponding position in the log file entry is the sixth position, which is the string "Decapper.DecapStop" in the log file entry 510, which specifies that the sample has reached the decapper. In particular, the latter string specifies the position at which the sample tube has reached. In addition, for example, according to the first lookup table, if the first field data is the string "InstructionStarted" 521, the corresponding position in the log file entry is the fifth position, which is the string "DecapInstruction" in the log file entry 520. In particular, the latter string specifies that a task has started, in this case, the processing step of decapping.

[0266] Alternatively, in some implementations, if the log file is organized by using a markup language, the selection of the second entry may be performed by selecting an appropriate tag among tags marking entries of the log file based on the first entry.

[0267] Each log data item 610, 620 of the first plurality of log data items is constructed by concatenating a first field 511, 521 and a second field 512, 522 of the corresponding log file entry 510, 520 with a delimiter, in this case a dot. For example, the log data item "SiteReached.Decapper.DecapStop" 610 is constructed by concatenating the string "SiteReached" 511 with the string "Decapper.DecapStop" 512. Similarly, the log data item "InstructionStarted.DecapInstruction" 620 is constructed by concatenating the string "InstructionStarted" 521 with the string "DecapInstruction" 522.

[0268] The construction of the reference log data items is performed in the same manner as the construction of the log data items 610 , 620 described above.

[0269] Reference Figure 2 In a next step 410, the second computing device 220 obtains the log data from the first computing device 110 by receiving the data sent by the first computing device 110. Alternatively, the second computing device 220 obtains the log data from the first computing device 110 by accessing the log data at a memory device storing the log data, and the first computing device 110 makes the log data available to the second computing device 210.

[0270] In the next step 411, the second computing device 210 obtains the anomaly data. In this embodiment, obtaining the anomaly data includes using the log data items L1, L2, ..., L N-1 and the base log data item L N To evaluate the presence of anomalies. In particular, log data items L1, L2, ..., L N-1 is used as input to a trained machine learning algorithm, such as a trained long short-term memory recurrent neural network. The trained machine learning algorithm generates output data. The output data includes the following information, which specifies, for each entry of the log data vocabulary, that each entry will immediately follow the log data items L1, L2, ..., L2 in the absence of anomalies. N-1 In this case, by determining the probability of NThe presence of an anomaly is evaluated by determining whether the associated probability is greater than or equal to a given threshold, such as 0.05. In particular, if the probability is below the threshold, an anomaly is detected, and if an anomaly is detected, anomaly data is generated. In this case, in particular, an anomaly is detected, and therefore anomaly data is generated. The anomaly data specifies that an anomaly has been detected. In addition, the anomaly data may include resolution data, which includes information indicating one or more actions for resolving an anomaly present in the group of laboratory instruments 120.

[0271] At step 412, second computing device 412 provides the exception data to first computing device 110 by sending the data to first computing device 110. Second computing device 210 may provide, transmit, send, or output the exception data to first computing device 110 over the network, and in particular protocol suite 12.

[0272] At step 312, the first computing device 110 receives the anomaly data from the second computing device 210. At step 313, the first computing device 110 provides the anomaly data to the user, for example, by displaying the anomaly data on a screen. In particular, the first computing device 110 may provide one or more actions for resolving the detected anomaly. Specifically, the first computing device 110 may output and / or print and / or display to the user one or more actions for resolving the detected anomaly.

[0273] For example, the detected anomaly may involve incomplete testing performed on the sample, such as incomplete centrifugation and / or incomplete growth of biological material, and / or downtime of one or more laboratory instruments and / or unavailability of one or more laboratory instruments in the laboratory instruments. The first computing device 110 can provide suggestions to the user on how to resolve these anomalies. For example, the first computing device 110 can instruct the user to check and / or investigate whether the centrifuge is working properly. The first computing device 110 can also print and / or display error messages and / or notifications to the user, wherein the error messages and / or notifications are related to certain anomalies that can be detected by the machine learning algorithm.

[0274] In the case where the first computing device 110 is included in an automated laboratory system, the second computing device 210 may also detect an anomaly of a laboratory instrument in the set of laboratory instruments 120 and instruct the first computing device 110 to resolve the detected anomaly based on the anomaly and resolution data. For example, resolution of the anomaly may be accomplished by rerouting one or more sample tubes to bypass one or more instruments in the set of laboratory instruments 120 where the anomaly was detected. Thus, the automated laboratory system may be able to autonomously detect and correct anomalies.

[0275] In this case, the first computing device 110 performs an embodiment of the method according to the first aspect of the present invention. In particular, in this embodiment, the first computing device 110 performs the step of obtaining log data by constructing the log data (the above step 311) and the step of obtaining abnormal data by receiving abnormal data from the second computing device 210 (such as the above step 312).

[0276] Figure 3 ( Figure 3 ) shows a second swim lane diagram including exemplary steps performed by a first computing device, a second computing device, and a third computing device. In this embodiment, the first computing device 110 performs an embodiment of the method according to the second aspect of the present invention. In particular, the first memory 112 stores a computer program including instructions, which, when executed by the processor 111, causes the first computing device 110 to perform the method according to the second aspect of the present invention.

[0277] At step 320 , the first computing device 110 provides (eg, sends) the log file to the second computing device 210 .

[0278] At step 420, the second computing device constructs log data and reference log data items. Figure 2 At step 311, the second computing device 210 constructs the abnormal data. Figure 2 The construction of the abnormal data and the baseline log data items is performed at step 411. At step 412, the second computing device 210 provides (eg, sends) the abnormal data to the first computing device 110.

[0279] At step 321, the first computing device 110 obtains abnormal data from the second computing device 210, and displays the abnormal data to the user at step 313. Figure 2 In step 312 , the first computing device 110 obtains the exception data by receiving the exception data from the second computing device 210 .

[0280] In this case, the second computing device 210 performs an embodiment of the method according to the first aspect of the present invention. In particular, in this embodiment, the second computing device 110 performs the step of obtaining log data by constructing the log data (the above step 420) and the step of obtaining abnormal data by constructing abnormal data (such as the above step 421).

Claims

1. A computer-implemented method comprising the following steps: - obtaining (410) log data (600) from or by the first computing device (110), the log data comprising a first plurality of log data items (610, 620); and -Get (411) abnormal data, wherein the first computing device (110) is configured to access (310) a log file of a laboratory instrument (120), the log file (500) comprising a first plurality of log file entries (510, 520, 530, 540), each log file entry of the first plurality of log file entries comprising a plurality of fields (511, 512, 521, 522), and the first computing device (110) is configured to construct (311) the log data (600) by performing the following operation for each log data item (610, 620) of the first plurality of log data items: o Selecting the first field data of the first field (511, 521) of the corresponding log file entry (510, 520), wherein the first field (511, 521) of the corresponding log file entry (510, 520) 521) specifying a corresponding event that has occurred in said laboratory instrument (120); o Select the second field data of the second field (512, 522) of the corresponding log file entry (510, 520), the second field (512, 522) of the corresponding log file entry (510, 520) 522) further describing the corresponding event that has occurred in the laboratory instrument (120), wherein the selection of the corresponding second field is based on first field data of the corresponding first field of the corresponding log file entry; and o constructing each log data item by using first field data of a first field of the corresponding log file entry and second field data of a second field of the corresponding log file entry; The abnormality data includes information indicating the presence of an abnormality in the laboratory instrument (120), and the presence of the abnormality is evaluated by using the log data (600).

2. The method according to claim 1, wherein: The step of obtaining the anomaly data includes evaluating the presence of an anomaly in the laboratory instrument (120) by using the log data and generating the anomaly data.

3. The method according to claim 2, wherein: The log data is obtained from the first computing device (110), and the step of obtaining the anomaly data is performed by a second computing device (210), the second computing device (210) being different from the first computing device (110).

4. The method according to claim 2 or claim 3, wherein: An assessment of the presence of an anomaly in the laboratory instrument (120) is performed by using the log data as input to at least one of a machine learning algorithm and / or a decision tree.

5. The method according to any one of claims 2 to 4, further comprising the following steps: - obtaining a baseline log data item by or from said first computing device (110) using a corresponding baseline log file entry of said log file, Therein, the assessment of the presence of an anomaly in the laboratory instrument is based on the baseline log data items.

6. The method according to any one of claims 2 to 5, wherein: If an abnormality has been assessed in the laboratory instrument (120), the method further comprises the following steps: - obtaining resolution data, the resolution data comprising information indicative of one or more actions for resolving an anomaly present in the laboratory instrument (120).

7. The method according to any one of claims 2 to 6, wherein: The step of obtaining log data by the first computing device (110) includes: - accessing a log file of the laboratory instrument (120); and constructing the log data, wherein the step of constructing the log data comprises, for each log data item of the first plurality of log data items, performing the following operations: ○ Select the first field data of the first field of the corresponding log file entry; o selecting second field data of a second field of the corresponding log file entry, wherein selection of the second field is based on first field data of a first field of the corresponding log file entry; and o Each log data item is constructed by using first field data of a first field of the corresponding log file entry and second field data of a second field of the corresponding log file entry.

8. The method according to claim 1, further comprising the steps of: - providing, by the first computing device (110), the log data to a third computing device, the third computing device being configured to assess the presence of an anomaly in the laboratory instrument (120) by using the log data, And wherein the step of obtaining log data by the first computing device (110) comprises: - accessing log files of said laboratory instrument (120); and - constructing the log data, wherein the step of constructing the log data comprises, for each log data item in the first plurality of log data items, performing the following operations: ○ Select the first field data of the first field of the corresponding log file entry; o selecting second field data of a second field of the corresponding log file entry, wherein selection of the second field is based on first field data of a first field of the corresponding log file entry; and o Each log data item is constructed by using first field data of a first field of the corresponding log file entry and second field data of a second field of the corresponding log file entry.

9. A method according to any one of the preceding claims, wherein: Each of the first plurality of log file entries relates to a corresponding sample tube, and / or wherein, For each log data item in the first plurality of log data items, the corresponding log file entry relates to a given sample tube.

10. A method according to any one of the preceding claims, wherein: i. for at least a first log data item of the first plurality of log data items, a first field of the corresponding log file entry specifies that a sample tube has arrived at a location of the laboratory instrument (120), and a second field of the corresponding log file entry includes information specifying the location at which the sample tube has arrived, and / or ii. for at least a second log data item of the first plurality of log data items, a first field of the corresponding log file entry specifies a location from which the sample tube has left the laboratory instrument (120), and a second field of the corresponding log file entry includes information specifying the location from which the sample tube has left, and / or iii. for at least a third log data item of the first plurality of log data items, a first field of the corresponding log file entry specifies that a state of the sample tube has changed, and a second field of the corresponding log file entry includes information specifying a current state of the sample tube, and / or iv. for at least a fourth log data item in the first plurality of log data items, the first field of the corresponding log file entry specifies the start of a task of the laboratory, and the second field of the corresponding log file entry includes information specifying which task started, and / or v. for at least a fifth log data item in the first plurality of log data items, a first field of the corresponding log file entry specifies that a task of the laboratory was completed, and a second field of the corresponding log file entry includes information specifying which task was completed, and / or ⅵ. For at least a sixth log data item of the first plurality of log data items, the first field of the corresponding log file entry specifies that a route plan for a sample tube has been evaluated, and the second field of the corresponding log file entry includes information specifying whether the route plan has been accepted or rejected.

11. A computer-implemented method comprising the steps of: - providing (320) a log file (500) of a laboratory instrument (120) to a computing device (210), the log file (500) comprising a first plurality of log file entries (510, 520, 530, 540), each log file entry of the first plurality of log file entries comprising a plurality of fields; - obtaining (321) anomaly data, the anomaly data comprising information indicating the presence of an anomaly in the laboratory instrument (120), wherein the computing device (210) is configured to construct log data (600) comprising a first plurality of log data items (610, 620) by, for each log data item in the first plurality of log data items, performing the following operations: o selecting first field data of a first field (511, 521) of a corresponding log file entry (510, 520), the first field of the corresponding log file entry specifying a corresponding event that has occurred in the laboratory instrument (120); o selecting second field data of a second field (512, 522) of the corresponding log file entry (510, 520), the second field of the corresponding log file entry further describing the corresponding event that has occurred in the laboratory instrument (120), wherein The selection of the corresponding second field is based on first field data of the corresponding first field of the corresponding log file entry; and o constructing each log data item by using first field data of a first field of the corresponding log file entry and second field data of a second field of the corresponding log file entry; And wherein the computing device (210) is configured to evaluate the presence of an anomaly in the laboratory instrument (120) by using the log data.

12. A computer-implemented method comprising the steps of: - accessing a log file of a laboratory instrument, the log file comprising a first plurality of log file entries, each log file entry of the first plurality of log file entries comprising a plurality of fields; - constructing labeled training log data, the labeled training log data comprising a plurality of training log data items and labeled data; and - starting training of a machine learning algorithm, wherein the training of the machine learning algorithm is performed by using the labeled training log data, wherein the step of constructing the labeled training log data comprises, for each of the plurality of training log data items, performing the following operations: o selecting first field data of a first field of a corresponding log file entry, the first field of the corresponding log file entry specifying a corresponding event that has occurred in the laboratory instrument (120); o selecting second field data of a second field of the corresponding log file entry, the second field of the corresponding log file entry further describing the corresponding event that has occurred in the laboratory instrument (120), wherein the selection of the second field is based on first field data of a first field of the corresponding log file entry; and o Each of the training log data items is constructed by using first field data of the first field of the corresponding log file entry and second field data of the second field of the corresponding log file entry.

13. The method according to claim 12, comprising the steps of: The trained machine learning algorithm is applied to identify anomalies in the laboratory instrument, in particular by using the log data as input to the machine learning algorithm.

14. The method according to claim 13, wherein: The log data is constructed from the log file entries by associating the log file entries with log data items and / or by mapping the plurality of log file entries onto a plurality of log data items according to a specific mapping.

15. A data processing system comprising a processor configured to perform the method according to any one of the preceding claims.

16. A computer program product comprising instructions which, when the program is executed by a computer system, cause the computer system to perform the method according to any one of claims 1 to 14.

17. A computer readable medium comprising instructions which, when executed by a computer system, cause the computer system to perform the method according to any one of claims 1 to 14.

Images