# US 2026/0093856 A1 · Factory floor modeling

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

- Number: US 2026/0093856 A1
- Document ID: US20260093856A1
- Status: Published application
- Kind: A1
- Application number: 18/899,410
- Filed: 2024-09-27
- Published: 2026-04-02
- Earliest claimed: 2024-09-27
- Assignees: Intel Corporation
- Applicants: Intel Corporation
- Inventors: Sameer Sheorey; Michael Paulitsch
- Web page: https://patentgrep.com/patent/US20260093856A1

## CPC

- G06F 30/13: Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
- G06F 30/27: using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
- G06N 3/045: Combinations of networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/0475: Generative networks
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

System and techniques for facility floor modeling are described herein. A scan set that includes depth data and correlated image data from various positions on a facility floor is obtained. This data includes representations of equipment in the factory. A trained model is invoked, receiving the scan set as input and producing, as output, a computer model of the equipment along with associated information. The computer model of the equipment is integrated into a larger computer model of the factory and the information regarding the equipment is stored in a data structure that corresponds to the equipment's representation within the computer model of the factory.

## Claims

### Claim 1

An apparatus for facility floor modeling, the apparatus comprising:
a memory including instructions; and
processing circuitry, configured by the instructions when in operation, to obtain a scan set, the scan set including depth data and correlated image data collected from a set of positions on a floor of a facility, a field-of-view (FOV) represented in the scan set including a portion of equipment installed on the floor;
invoke a trained model with an input including the scan set, and an output including a computer model of the portion of equipment and information of the portion of equipment;
place the computer model of the portion of equipment into a computer model of the floor; and
write the information of the portion of equipment into a data structure that corresponds to the computer model of the portion of equipment as placed into the computer model of the floor.

### Claim 2 (depends on 1)

The apparatus of claim 1, wherein the data structure is part of a Building Information Modeling (BIM) database.

### Claim 3 (depends on 1)

The apparatus of claim 1, wherein the computer model of the portion of computer equipment is one of a set of available computer models corresponding to portions of equipment observable on the floor.

### Claim 4 (depends on 1)

The apparatus of claim 1, wherein the information of the portion of equipment includes read data obtained from image data via optical character recognition performed by the trained model.

### Claim 5 (depends on 1)

The apparatus of claim 1, wherein the information of the portion of equipment includes an operating parameter based on a connection to another portion of equipment, the connection present in the scan set.

### Claim 6 (depends on 1)

The apparatus of claim 1, wherein the processing circuitry is configured by the instructions to train the trained model including:
providing a user interface to present the output of the trained model;
receiving a modification to the output of the trained model via the user interface to create a new output; and
modifying the trained model to conform to the new output when invoked on the input.

### Claim 7

At least one non-transitory machine readable medium including instructions for facility floor modeling, the instructions, when executed by processing circuitry, cause the processing circuitry to perform operations comprising:
obtaining a scan set, the scan set including depth data and correlated image data collected from a set of positions on a floor of a facility, a field-of-view (FOV) represented in the scan set including a portion of equipment installed on the floor;
invoking a trained model with an input including the scan set and an output including a computer model of the portion of equipment and information of the portion of equipment;
placing the computer model of the portion of equipment into a computer model of the floor; and
writing the information of the portion of equipment into a data structure that corresponds to the computer model of the portion of equipment as placed into the computer model of the floor.

### Claim 8 (depends on 7)

The at least one non-transitory machine readable medium of claim 7, wherein the computer model of the portion of equipment is a Boundary Representation (BRep) model that defines volume limits for the portion of equipment.

### Claim 9 (depends on 7)

The at least one non-transitory machine readable medium of claim 7, wherein the data structure is part of a Building Information Modeling (BIM) database.

### Claim 10 (depends on 9)

The at least one non-transitory machine readable medium of claim 9, wherein input to the trained model includes the BIM, the BIM being partially complete.

### Claim 11 (depends on 7)

The at least one non-transitory machine readable medium of claim 7, wherein the computer model of the portion of computer equipment is one of a set of available computer models corresponding to portions of equipment observable on the floor.

### Claim 12 (depends on 11)

The at least one non-transitory machine readable medium of claim 11, wherein the output of the trained model specifies a key for the computer model of the portion of equipment, the key corresponding to a pre-constructed version of the computer model for the portion of equipment.

### Claim 13 (depends on 7)

The at least one non-transitory machine readable medium of claim 7, wherein the information of the portion of equipment includes read data obtained from image data via optical character recognition performed by the trained model.

### Claim 14 (depends on 13)

The at least one non-transitory machine readable medium of claim 13, wherein the read data includes an equipment identifier (ID), an equipment make, or an equipment model.

### Claim 15 (depends on 7)

The at least one non-transitory machine readable medium of claim 7, wherein the information of the portion of equipment includes an operating parameter based on a connection to another portion of equipment, the connection present in the scan set.

### Claim 16 (depends on 15)

The at least one non-transitory machine readable medium of claim 15, wherein the operating parameter is one of a voltage, a current, a liquid, a gas, a pressure, or a flow rate.

### Claim 17 (depends on 7)

The at least one non-transitory machine readable medium of claim 7, wherein the portion of equipment is a machine, a pipe, a conduit, a wire, or a receptacle.

### Claim 18 (depends on 7)

The at least one non-transitory machine readable medium of claim 7, wherein the operations comprise training the trained model including:
providing a user interface to present the output of the trained model;
receiving a modification to the output of the trained model via the user interface to create a new output; and
modifying the trained model to conform to the new output when invoked on the input.

### Claim 19 (depends on 7)

The at least one non-transitory machine readable medium of claim 7, wherein the operations comprise creating the trained model including:
obtaining an initial model;
obtaining a set of depth data for a second floor from a first sensor at different positions on the second floor;
obtaining a set of image data for the second floor from a second sensor at different positions on the second floor;
aligning the set of depth data with the set of image data based on positions of the first sensor and the second sensor to create an aligned set of scan data;
obtaining a complete Building Information Modeling (BIM) database for the second floor;
obtaining a complete computer model of the second floor; and
iteratively adjusting the initial model by:
providing the aligned set of scan data as input and producing a training output of BIM data and computer models of equipment;
measuring an error between the training output and the complete BIM database and the complete computer model of the second floor;
adjusting an internal structure of the initial model to cause the error to be reduced; and
repeating with different inputs until the error is below a threshold.

### Claim 20 (depends on 7)

The at least one non-transitory machine readable medium of claim 7, wherein the trained model is a transformer model.
