# US 2025/0094677 A1 · Self-service artificial intelligence platform leveraging data-based and physics-based models for providing real-time controls and recommendations

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

- Number: US 2025/0094677 A1
- Document ID: US20250094677A1
- Status: Published application
- Kind: A1
- Application number: 18/969,187
- Filed: 2024-12-04
- Published: 2025-03-20
- Earliest claimed: 2019-03-29
- Assignees: Bert Labs Private Limited
- Applicants: Bert Labs Private Limited
- Inventors: Rohit KOCHAR; Jayesh JAIN; Yasin KHAN
- Web page: https://patentgrep.com/patent/US20250094677A1

## CPC

- G06F 30/27: using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
- G05B 13/0265: {the criterion being a learning criterion}
- G05B 13/04: involving the use of models or simulators
- G06N 20/00: Machine learning
- G05B 2219/2614: HVAC, heating, ventillation, climate control
- G09B 23/06: for physics
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- 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
- G09B: EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS (devices for psychotechnics or for testing reaction times A61B5/16; games, sports, amusements A63; projectors, projector screens G03B)

## Abstract

The present disclosure relates to development of a self-service artificial intelligence platform by integrating data-based model with physics-based model and vice-versa to generate real-time recommendations and control actions. Further, the present disclosure provides the system and method for at least one of data collection and preparation, developing a hybrid system/control model, and developing a physics-based model driven by data-based model and vice versa to generate real-time recommendations and control actions.

## Claims

### Claim 1

A method of optimizing an enterprise application, the method comprising:
configuring a physics-based model using simulated data of the enterprise, wherein the physics-based model includes models of virtual sensors;
generating independent variable values for a data-based model using the physics-based model when data is unavailable in a database, thereby enabling the data-based model to learn;
building a baseline data-based model offline using the independent variable values generated by the physics-based model, wherein the baseline data-based model is trained without real-time field data;
deploying the baseline data-based model at a client site, wherein the data-based model starts receiving real-time data from the enterprise application; and
enabling the data-based model to self-learn once deployed at the client site, using the real-time data, thereby improving control readiness and optimizing enterprise outcomes.

### Claim 2 (depends on 1)

The method as claimed in claim 1, wherein the real-time data received from the enterprise application is used to continuously update and refine the data-based model after deployment.

### Claim 3 (depends on 1)

The method as claimed in claim 1, wherein the physics-based model generates independent variable values in the form of virtual sensor data, which is provided to the data-based model for training.

### Claim 4 (depends on 1)

The method as claimed in claim 1, wherein the physics-based model is configured to receive real-time feedback from the data-based model, wherein the feedback includes real-time data from the data-based model, the real-time data being used to simulate at least one sub-component of the enterprise application and being used to calibrate the physics-based model in real time to improve the accuracy of the physics-based model.

### Claim 5 (depends on 1)

The method as claimed in claim 1, wherein the data-based model and the physics-based model operate in a cyclic manner, with feedback from one model refining the predictions and calculations of the other model.

### Claim 6 (depends on 1)

The method as claimed in claim 1, wherein the independent variable values generated by the physics-based model include operational parameters and environmental conditions that influence the enterprise application.

### Claim 7 (depends on 1)

The method as claimed in claim 1, wherein the independent variable values generated by the physics-based model include sensor data representing physical variables such as temperature, humidity, pressure, and energy consumption.

### Claim 8 (depends on 1)

The method as claimed in claim 1, wherein the self-learning process of the data-based model involves adjusting model parameters to minimize prediction errors based on the real-time data received from the enterprise application.

### Claim 9 (depends on 1)

The method as claimed in claim 1, wherein the baseline data-based model is trained using simulated sensor data from the physics-based model in the absence of real-time field data, and is capable of providing reliable predictions once deployed.

### Claim 10 (depends on 1)

The method as claimed in claim 1, wherein the physics-based model simulates the behaviour of the enterprise application and predicts outcomes based on a range of operational and environmental conditions.

### Claim 11 (depends on 1)

The method as claimed in claim 1, wherein the baseline data-based model is configured to adaptively improve its predictive accuracy over time through continuous learning from the real-time field data received from the enterprise application.

### Claim 12 (depends on 1)

The method as claimed in claim 1, wherein the baseline data-based model is designed to work with various types of enterprise applications, including HVAC systems, manufacturing processes, or energy management systems.

### Claim 13 (depends on 1)

The method as claimed in claim 1, wherein the real-time data received from the enterprise application is used to adjust control actions in real time, providing dynamic optimization based on actual conditions.

### Claim 14 (depends on 1)

The method as claimed in claim 1, wherein the data-based model is trained using a combination of historical data and synthetic data generated by the physics-based model, allowing it to handle scenarios where real-time field data is incomplete or unavailable.

### Claim 15 (depends on 1)

The method as claimed in claim 1, wherein the feedback loop between the data-based model and the physics-based model enables the correction of drift or inaccuracies in the data-based model's predictions over time.

### Claim 16 (depends on 1)

The method as claimed in claim 1, wherein the baseline data-based model is deployed at the client site in an environment that includes a monitoring system for tracking the performance and prediction accuracy of the data-based model.

### Claim 17 (depends on 1)

The method as claimed in claim 1, wherein the self-learning process involves periodic recalibration of the data-based model using updated real-time field data to ensure its predictions remain accurate and relevant to changing operational conditions.

### Claim 18 (depends on 1)

The method as claimed in claim 1, wherein the deployment of the baseline data-based model includes configuring the model to integrate with existing enterprise software systems to enable seamless data exchange and control actions.

### Claim 19 (depends on 1)

The method as claimed in claim 1, wherein the data-based model improves the physics-based model by providing real-time feedback that enables the physics-based model to adapt and correct inaccuracies in its predictions, wherein the feedback from the data-based model is used to update model parameters, reduce model biases, and improve the overall accuracy of the physics-based model.
