# US 2023/0297060 A1 · Machine learning optimization of control code in an industrial automation environments

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

- Number: US 2023/0297060 A1
- Document ID: US20230297060A1
- Status: Published application
- Kind: A1
- Application number: 17/698,448
- Filed: 2022-03-18
- Published: 2023-09-21
- Earliest claimed: 2022-03-18
- Assignees: Rockwell Automation Technologies, Inc.
- Applicants: Rockwell Automation Technologies, Inc.
- Inventors: Jordan C. Reynolds; John J. Hagerbaumer; Troy W. Mahr; Thomas K. Jacobsen; Giancarlo Scaturchio
- Web page: https://patentgrep.com/patent/US20230297060A1

## CPC

- G05B 19/056: {Programming the PLC}
- G06N 20/00: Machine learning
- G05B 2219/13115: Optimize ladder diagram block by rearrangement of serial and parallel
- G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

Various embodiments of the present technology generally relate to industrial automation environments. More specifically, embodiments include systems and methods for applying machine learning techniques to industrial control code to detect errors, perform optimizations, and generate predictions. In some embodiments, a design application in an industrial automation environment generates a functional block diagram configured for implementation by a programmable logic controller. The design application generates feature vectors that represent the functional block diagram configured for ingestion by a machine learning model. The design application supplies the feature vectors to the machine learning model. The design application receives a machine learning output that comprises the recommendation feedback generated by the machine learning model and responsively modifies the functional block diagram based on the recommendation feedback.

## Claims

### Claim 1

A system to optimize a functional block diagram that corresponds to a control program in an industrial automation environment, the system comprising:
a memory that stores executable components; and
a processor, operatively coupled to the memory, that executes the executable components, the executable components comprising:
a design application component configured to generate a functional block diagram configured for implementation by a programmable logic controller to control an automated process;
a machine learning interface component configured to generate feature vectors based on the functional block diagram that are configured for ingestion by a machine learning model and supply the feature vectors to the machine learning model to obtain recommendation feedback on the functional block diagram from the machine learning model; and
the design application component configured to receive a machine learning output that comprises the recommendation feedback generated by the machine learning model component and responsively modify the functional block diagram based on the recommendation feedback.

### Claim 2 (depends on 1)

The system of claim 1 further comprising:
the machine learning model configured to ingest the feature vectors that represent the functional block diagram, process the ingested feature vectors, compare the ingested feature vectors to historical functional block diagram data using machine learning algorithms, generate the machine learning output based on the comparison, and transfer the machine learning output.

### Claim 3 (depends on 1)

The system of claim 1 further comprising:
a training component configured to receive historical functional block diagrams for the machine learning model component, generate additional feature vectors that represent the historical functional block diagrams for ingestion by the machine learning model component; and
the machine learning model configured to ingest the additional feature vectors and train itself using the additional feature vectors.

### Claim 4 (depends on 1)

The system of claim 1 wherein:
the design application component is configured to generate the functional block diagram comprises the design application component configured to generate an image that depicts the functional block diagram; and
the machine learning interface component is configured to generate the feature vectors that represent the functional block diagram comprises the machine learning interface component configured to generate the feature vectors based on the image.

### Claim 5 (depends on 1)

The system of claim 1 wherein:
the design application component is configured to generate the functional block diagram comprises the design application component configured to generate ladder logic that depicts the functional block diagram; and
the machine learning interface component is configured to generate the feature vectors that represent the functional block diagram comprises the machine learning interface component configured to generate the feature vectors based on the ladder logic.

### Claim 6 (depends on 1)

The system of claim 1 wherein the recommendation feedback indicates an error in the functional block diagram.

### Claim 7 (depends on 1)

The system of claim 1 wherein the recommendation feedback indicates a block recommendation for the functional block diagram.

### Claim 8

A method to optimize a functional block diagram that corresponds to a control program in an industrial automation environment, the method comprising:
generating a functional block diagram configured for implementation by a programmable logic controller to control an automated process;
generating feature vectors based on the functional block diagram that are configured for ingestion by a machine learning model;
supplying the feature vectors to the machine learning model to obtain recommendation feedback on the functional block diagram from the machine learning model; and
receiving a machine learning output that comprises the recommendation feedback generated by the machine learning model and responsively modifying the functional block diagram based on the recommendation feedback.

### Claim 9 (depends on 8)

The method of claim 8 further comprising:
ingesting the feature vectors that represent the functional block diagram;
processing the ingested feature vectors, comparing the ingested feature vectors to historical functional block diagram data using machine learning algorithms, and responsively generating the machine learning output based on the comparison; and
transferring the machine learning output.

### Claim 10 (depends on 8)

The method of claim 8 further comprising:
receiving historical functional block diagrams for the machine learning model;
generating additional feature vectors that represent the historical functional block diagrams for ingestion by the machine learning model;
and training the machine learning model based on the additional feature vectors.

### Claim 11 (depends on 8)

The method of claim 8 wherein:
generating the functional block diagram comprises generating an image that depicts the functional block diagram; and
generating the feature vectors that represent the functional block diagram comprises generating the feature vectors based on the image.

### Claim 12 (depends on 8)

The method of claim 8 wherein:
generating the functional block diagram comprises generating ladder logic that represents the functional block diagram; and
generating the feature vectors that represent the functional block diagram comprises generating the feature vectors based on the ladder logic.

### Claim 13 (depends on 8)

The method of claim 8 wherein the recommendation feedback indicates an error in the functional block diagram.

### Claim 14 (depends on 8)

The method of claim 8 wherein the recommendation feedback indicates a block recommendation for the functional block diagram.

### Claim 15

A non-transitory computer-readable medium having stored thereon instructions to optimize a functional block diagram that corresponds to a control program in an industrial automation environment, wherein the instructions, in response to execution, cause a system comprising a processor to perform operations, the operations comprising:
generating a functional block diagram configured for implementation by a programmable logic controller to control an automated process;
generating feature vectors based on the functional block diagram that are configured for ingestion by a machine learning model;
supplying the feature vectors to the machine learning model to obtain recommendation feedback on the functional block diagram from the machine learning model; and
receiving a machine learning output that comprises the recommendation feedback generated by the machine learning model and responsively modifying the functional block diagram based on the recommendation feedback.

### Claim 16 (depends on 15)

The non-transitory computer-readable medium of claim 15, the operations further comprising:
ingesting the feature vectors that represent the functional block diagram;
processing the ingested feature vectors, comparing the ingested feature vectors to historical functional block diagram data using machine learning algorithms, and generating the machine learning output based on the comparison; and
transferring the machine learning output.

### Claim 17 (depends on 15)

The non-transitory computer-readable medium of claim 15, the operations further comprising:
receiving historical functional block diagrams for the machine learning model;
generating additional feature vectors that represent the historical functional block diagrams for ingestion by the machine learning model; and
training the machine learning model using the additional feature vectors.

### Claim 18 (depends on 15)

The non-transitory computer-readable medium of claim 15 wherein:
generating the functional block diagram comprises generating an image that depicts the functional block diagram; and
generating the feature vectors that represent the functional block diagram comprises generating the feature vectors based on the image.

### Claim 19 (depends on 15)

The non-transitory computer-readable medium of claim 15 wherein:
generating the functional block diagram comprises generating ladder logic that represent the functional block diagram; and
generating the feature vectors that represent the functional block diagram comprises generating the feature vectors based on the ladder logic.

### Claim 20 (depends on 15)

The non-transitory computer-readable medium of claim 15 wherein the recommendation feedback indicates at least one of an error in the functional block diagram or a block recommendation for the functional block diagram.
