# US 2024/0202404 A1 · Apparatuses, methods, and computer program products for predicting fugitive leaks

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

- Number: US 2024/0202404 A1
- Document ID: US20240202404A1
- Status: Published application
- Kind: A1
- Application number: 18/176,780
- Filed: 2023-03-01
- Published: 2024-06-20
- Earliest claimed: 2022-12-20
- Assignees: Honeywell International Inc.
- Applicants: Honeywell International Inc.
- Inventors: Shahrul Azman Bin Zainal Abidin; Madhukar Madhavamurthy Gundappa; Ramesh Murugan Natarajan
- Web page: https://patentgrep.com/patent/US20240202404A1

## CPC

- G06F 30/27: using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Methods, apparatuses, and computer program products for predicting fugitive leaks are provided. For example, a computer-implemented method may include receiving current operating conditions data associated with current operation of one or more operational systems and generating, using a fugitive leak prediction model, fugitive leak predictions corresponding to the current operation of the one or more operational systems. The fugitive leak prediction model may be a machine learning model trained based at least in part on historical operating conditions data associated with past operation of the one or more operational systems and historical fugitive emissions data associated with the past operation of the one or more operational systems, and the fugitive leak prediction model may be configured to generate the fugitive leak predictions based at least in part on the current operating conditions data

## Claims

### Claim 1

An apparatus comprising at least one processor and at least one non-transitory memory comprising program code stored thereon, wherein the at least one non-transitory memory and the program code are configured to, with the at least one processor, cause the apparatus to at least:
receive current operating conditions data associated with current operation of one or more operational systems;
generate, using a fugitive leak prediction model, at least one fugitive leak prediction corresponding to the current operation of the one or more operational systems,
wherein the fugitive leak prediction model comprises a machine learning model trained based at least in part on historical operating conditions data associated with past operation of the one or more operational systems and historical fugitive emissions data associated with the past operation of the one or more operational systems,
wherein the fugitive leak prediction model is configured to generate the fugitive leak predictions based at least in part on the current operating conditions data; and
output the at least one fugitive leak prediction.

### Claim 2 (depends on 1)

The apparatus of claim 1, wherein the at least one non-transitory memory and the program code are configured to, with the at least one processor, further cause the apparatus to at least:
generate at least one repair alert for one or more components of the one or more operational systems based at least in part on the fugitive leak predictions, wherein to output the at least one fugitive leak prediction the apparatus is caused to at least output the at least one repair alert.

### Claim 3 (depends on 1)

The apparatus of claim 1, wherein each of the fugitive leak predictions comprises one or more predicted fugitive emissions values representing predicted fugitive emissions associated with the one or more operational systems.

### Claim 4 (depends on 1)

The apparatus of claim 1, wherein the fugitive leak predictions are generated based at least in part on correlations between one or more sensed operating conditions values of the historical operating conditions data and one or more sensed fugitive emissions values of the historical fugitive emissions data, wherein each sensed operating conditions value of the one or more sensed operating conditions values represents an operating condition of the one or more operational systems at an instance of time during the past operation of the one or more operational systems, wherein each sensed fugitive emissions value of the one or more sensed fugitive emissions values represents fugitive emissions at an instance of time during the past operation of the one or more operational systems, and wherein the fugitive leak prediction model is trained based at least in part on the correlations.

### Claim 5 (depends on 1)

The apparatus of claim 1, wherein the at least one non-transitory memory and the program code are configured to, with the at least one processor, further cause the apparatus to at least:
train the fugitive leak prediction model based at least in part on the historical operating conditions data and the historical fugitive emissions data.

### Claim 6 (depends on 1)

The apparatus of claim 1, wherein the fugitive leak prediction model is trained based at least in part on simulated fugitive emissions data associated with simulated operation of the one or more operational systems, and the fugitive leak predictions are generated based at least in part on correlations between one or more simulated operating conditions values of the simulated fugitive emissions data and one or more estimated fugitive emissions values of the simulated fugitive emissions data associated with the one or more simulated operating conditions values, wherein the fugitive leak prediction model is trained based at least in part on the correlations, and
wherein each simulated operating conditions value of the one or more simulated operating conditions values represents an operating condition of the one or more operational systems during the simulated operation of the one or more operational systems, and wherein each estimated fugitive emissions value of the one or more estimated fugitive emissions values represents estimated fugitive emissions resulting from the one or more simulated operating conditions during the simulated operation of the one or more operational systems.

### Claim 7 (depends on 6)

The apparatus of claim 6, wherein the at least one non-transitory memory and the program code are configured to, with the at least one processor, further cause the apparatus to at least:
generate the simulated fugitive emissions data by generating the one or more estimated fugitive emissions values based at least in part on the one or more simulated operating conditions values.

### Claim 8 (depends on 1)

The apparatus of claim 1, wherein the fugitive leak prediction model is trained based at least in part on historical fugitive leak data associated with the one or more operational systems, wherein the historical fugitive leak data identifies one or more fugitive leaks detected within the one or more operational systems at one or more instances of time during the past operation of the one or more operational systems.

### Claim 9 (depends on 8)

The apparatus of claim 8, wherein the fugitive leak predictions are generated based at least in part on correlations between one or more sensed operating conditions values of the historical operating conditions data corresponding to the one or more instances of time at which the one or more fugitive leaks are detected and estimated fugitive emissions values determined with respect to the one or more sensed operating conditions values, and wherein the fugitive leak prediction model is trained based at least in part on the correlations.

### Claim 10 (depends on 1)

The apparatus of claim 1, wherein the historical fugitive emissions data identifies one or more sensed fugitive emissions values representing fugitive emissions sensed by one or more fugitive emissions sensors during the past operation of the one or more operational systems.

### Claim 11

A computer-implemented method comprising:
receiving current operating conditions data associated with current operation of one or more operational systems;
generating, using a fugitive leak prediction model, at least one fugitive leak prediction corresponding to the current operation of the one or more operational systems,
wherein the fugitive leak prediction model comprises a machine learning model trained based at least in part on historical operating conditions data associated with past operation of the one or more operational systems and historical fugitive emissions data associated with the past operation of the one or more operational systems,
wherein the fugitive leak prediction model is configured to generate the fugitive leak predictions based at least in part on the current operating conditions data; and
outputting the at least one fugitive leak prediction.

### Claim 12 (depends on 11)

The method of claim 11, further comprising generating at least one repair alert for one or more components of the one or more operational systems based at least in part on the fugitive leak predictions and outputting the at least one fugitive leak prediction by at least outputting the at least one repair alert.

### Claim 13 (depends on 11)

The method of claim 11, wherein each of the fugitive leak predictions comprises one or more predicted fugitive emissions values representing predicted fugitive emissions associated with the one or more operational systems.

### Claim 14 (depends on 11)

The method of claim 11, wherein the fugitive leak predictions are generated based at least in part on correlations between one or more sensed operating conditions values of the historical operating conditions data and one or more sensed fugitive emissions values of the historical fugitive emissions data, wherein each sensed operating conditions value of the one or more sensed operating conditions values represents an operating condition of the one or more operational systems at an instance of time during the past operation of the one or more operational systems, wherein each sensed fugitive emissions value of the one or more sensed fugitive emissions values represents fugitive emissions at an instance of time during the past operation of the one or more operational systems, and wherein the fugitive leak prediction model is trained based at least in part on the correlations.

### Claim 15 (depends on 11)

The method of claim 11, wherein the at least one non-transitory memory and the program code are configured to, with the at least one processor, further cause the apparatus to at least:
train the fugitive leak prediction model based at least in part on the historical operating conditions data and the historical fugitive emissions data.

### Claim 16 (depends on 11)

The method of claim 11, wherein the fugitive leak prediction model is trained based at least in part on simulated fugitive emissions data associated with simulated operation of the one or more operational systems, and the fugitive leak predictions are generated based at least in part on correlations between one or more simulated operating conditions values of the simulated fugitive emissions data and one or more estimated fugitive emissions values of the simulated fugitive emissions data associated with the one or more simulated operating conditions values, wherein the fugitive leak prediction model is trained based at least in part on the correlations, and
wherein each simulated operating conditions value of the one or more simulated operating conditions values represents an operating condition of the one or more operational systems during the simulated operation of the one or more operational systems, and wherein each estimated fugitive emissions value of the one or more estimated fugitive emissions values represents estimated fugitive emissions resulting from the one or more simulated operating conditions during the simulated operation of the one or more operational systems.

### Claim 17 (depends on 16)

The method of claim 16, further comprising generating the simulated fugitive emissions data by generating the one or more estimated fugitive emissions values based at least in part on the one or more simulated operating conditions values.

### Claim 18 (depends on 11)

The method of claim 11, wherein the fugitive leak prediction model is trained based at least in part on historical fugitive leak data associated with the one or more operational systems, wherein the historical fugitive leak data identifies one or more fugitive leaks detected within the one or more operational systems at one or more instances of time during the past operation of the one or more operational systems.

### Claim 19 (depends on 18)

The method of claim 18, wherein the fugitive leak predictions are generated based at least in part on correlations between one or more sensed operating conditions values of the historical operating conditions data corresponding to the one or more instances of time at which the one or more fugitive leaks are detected and estimated fugitive emissions values determined with respect to the one or more sensed operating conditions values, and wherein the fugitive leak prediction model is trained based at least in part on the correlations.

### Claim 20

A computer program product comprising at least one non-transitory computer-readable storage medium having computer-readable program code portions stored therein, the computer-readable program code portions comprising an executable portion configured to:
receive current operating conditions data associated with current operation of one or more operational systems;
generate, using a fugitive leak prediction model, at least one fugitive leak prediction corresponding to the current operation of the one or more operational systems,
wherein the fugitive leak prediction model comprises a machine learning model trained based at least in part on historical operating conditions data associated with past operation of the one or more operational systems and historical fugitive emissions data associated with the past operation of the one or more operational systems,
wherein the fugitive leak prediction model is configured to generate the fugitive leak predictions based at least in part on the current operating conditions data; and
output the at least one fugitive leak prediction.
