# US 2026/0017560 A1 · A Computer Implemented Method and Computer Program Product for Analyzing Data Originating from at Least One Device

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0017560 A1
- Document ID: US20260017560A1
- Status: Published application
- Kind: A1
- Application number: 18/881,350
- Filed: 2023-07-05
- Published: 2026-01-15
- Earliest claimed: 2022-07-07
- Assignees: Siemens Aktiengesellschaft
- Applicants: Siemens Aktiengesellschaft
- Inventors: Ralf GROSS; Ingo THON
- Web page: https://patentgrep.com/patent/US20260017560A1

## CPC

- G06N 20/00: Machine learning
- G06N 3/042: Knowledge-based neural networks; Logical representations of neural networks
- G06N 3/045: Combinations of networks
- G06N 3/084: Backpropagation, e.g. using gradient descent
- G06N 3/09: Supervised learning
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A method and computer program product for analyzing data originating from at least one device, wherein the computer program product includes an ML-artifact forming the functionality of a ML-model and a monitoring-artifact, where the ML-model is trained and/or set up or can be trained and/or set up with a machine learning method using training-data, where the ML-artifact produces an ML-result as a response of inputting input-data, based on data originating from the device, into the ML-artifact, where the monitoring-artifact produces an ML-result-reliability-information regarding the ML-result as a response to inputting the input-data into the monitoring-artifact, where the training-data or part of the training-data is used to train the monitoring artifact, and where the computer program product outputs a qualified-ML-result based on the ML-result and the ML-result-reliability-information when being executed by a computer.

## Claims

### Claim at position 1

-13. (canceled)

### Claim 14

A non-transitory computer-readable medium encoded with a computer program which, when executed by a computer, causes analysis of data originating from at least one device, the computer program comprising:
program code for providing a ML-artifact comprising a functionality of a ML-model; and
program code for providing a monitoring-artifact;
wherein the ML-model is at least one of trained and trainable with a machine learning method utilizing training-data;
wherein the ML-artifact is configured to produce an ML-result as a response to inputting input-data, based on data originating from at least one device, into the ML-artifact;
wherein the monitoring-artifact is configured to produce ML-result-reliability-information regarding the ML-result as a response to inputting the input-data into the monitoring-artifact;
wherein the training-data or part of the training-data is utilized to train the monitoring artifact; and
wherein the computer program outputs a qualified-ML-result based on the ML-result and the ML-result-reliability-information when being executed by the computer.

### Claim 15 (depends on 14)

The non-transitory computer-readable medium according to claim 14, wherein the ML-artifact is one of configured for training the ML-model based on training data and being able to train the ML-model based on the training data; and wherein the monitoring-artifact is one of configured to train itself based on training data and is able to train itself based on the training data.

### Claim 16 (depends on 15)

The non-transitory computer-readable according to claim 15, wherein after training or setting-up the ML-artifact and the monitoring-artifact, based on the training-data, the trained ML-artifact and the trained or amended monitoring-artifact are combined into the computer program.

### Claim 17 (depends on 14)

The non-transitory computer-readable medium according to claim 14, wherein the ML-artifact and the monitoring-artifact are executed in a single program run or cycle, if the computer program is executed on the computer.

### Claim 18 (depends on 14)

The non-transitory computer-readable medium according to claim 14, wherein the computer program comprises a single artifact.

### Claim 19 (depends on 14)

The non-transitory computer-readable medium according to claim 14, wherein one of (i) the computer program comprises a computational graph which comprises the ML-artifact and the monitoring-artifact and (ii) one single computational graph comprising the ML-artifact and the monitoring-artifact is realized in which the single computational graph forms part of the computer program.

### Claim 20 (depends on 14)

The non-transitory computer-readable medium according to claim 14, wherein at least one of (i) the computer program further comprises a preprocessing-artifact for converting the data from the at least one device into the input-data and (ii) the computer program further comprises a postprocessing-artifact for generating the qualified-ML-result.

### Claim 21 (depends on 19)

The non-transitory computer-readable medium according to claim 19, wherein the single computational graph comprises at least one of the preprocessing-artifact and the postprocessing-artifact.

### Claim 22 (depends on 20)

The non-transitory computer-readable medium according to claim 20, wherein the single computational graph comprises at least one of the preprocessing-artifact and the postprocessing-artifact.

### Claim 23

A computer implemented method for producing a qualified output of an ML-model, based on data originating from at least one device, the method comprising:
inputting input-data, based on the data origination from the at least one device, into an ML-artifact, comprising the functionality of the ML-model, to produce an ML-result;
inputting the input-data in a monitoring-artifact to produce ML-result-reliability-information; and
outputting a qualified-ML-result based on the ML-result and the ML-result-reliability-information;
wherein said inputting input-data into the ML-artifact, inputting the input-data in the monitoring-artifact and outputting a qualified-ML-result based on the ML-result and the ML-result-reliability-information are performed in a single run of a computer program product when executed on a computer.

### Claim 24

A computer implemented method for producing at least one of a trained ML-model and an ML-artifact and a trained or amended monitoring artifact, based on training-data, comprising:
inputting training-data into an ML-artifact, comprising a functionality of an ML-model, to produce a trained ML-artifact; and
inputting the training-data into a monitoring-artifact to produce the trained or amended monitoring-artifact;
wherein said inputting the training-data into the ML-artifact, inputting the training-data in the monitoring-artifact, and producing the trained ML-artifact and the trained or amended monitoring-artifact are performed in a single run of a computer program when executed on a computer.

### Claim 25 (depends on 24)

The computer implemented method according to claim 24, wherein after at least one of training and amending the ML-artifact and the monitoring-artifact based on the training-data, the trained ML-artifact and the trained or amended monitoring-artifact are combined into the computer program.

### Claim 26 (depends on 23)

The computer implemented method according to claim 23, wherein one of (i) the computer program comprises the ML-artifact and the monitoring-artifact as one single computational graph and (ii) the computer program is realized as a single artifact.

### Claim 27 (depends on 24)

The computer implemented method according to one claim 24, wherein one of (i) the computer program comprises the ML-artifact and the monitoring-artifact as one single computational graph and (ii) the computer program is realized as a single artifact.

### Claim 28 (depends on 25)

The computer implemented method according to claim 25, wherein one of (i) the computer program comprises the ML-artifact and the monitoring-artifact as one single computational graph and (ii) the computer program is realized as a single artifact.

### Claim 29 (depends on 23)

The computer implemented method according to one claim 23, wherein the computer program comprises:
program code for providing the ML-artifact comprising the functionality of the ML-model; and
program code for providing the monitoring-artifact;
wherein the ML-model is at least one of trained and trainable with a machine learning method utilizing training-data;
wherein the ML-artifact is configured to produce the ML-result as a response to inputting the input-data, based on the input-data originating from at least one device, into the ML-artifact;
wherein the monitoring-artifact is configured to produce the ML-result-reliability-information regarding the ML-result as a response to inputting the input-data into the monitoring-artifact;
wherein the training-data or part of the training-data is utilized to train the monitoring artifact; and
wherein the computer program outputs the qualified-ML-result based on the ML-result and the ML-result-reliability-information when being executed by the computer.
