# US 2025/0148141 A1 · Method for Tracking Construction Site Progress

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

- Number: US 2025/0148141 A1
- Document ID: US20250148141A1
- Status: Published application
- Kind: A1
- Application number: 18/501,738
- Filed: 2023-11-03
- Published: 2025-05-08
- Earliest claimed: 2023-11-03
- Assignees: Dalux Aps
- Applicants: DALUX ApS
- Inventors: Anders Rong
- Web page: https://patentgrep.com/patent/US20250148141A1

## CPC

- G06F 30/13: Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
- G06T 7/001: {using an image reference approach}
- G06Q 10/06: Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q 10/103: {Workflow collaboration or project management}
- G06Q 50/08: Construction
- G06T 7/0002: {Inspection of images, e.g. flaw detection}
- G06T 7/0008: {checking presence/absence}
- G06T 17/00: Three-dimensional [3D] modelling for computer graphics
- G06V 20/647: {by matching two-dimensional images to three-dimensional objects}
- G06T 2200/24: involving graphical user interfaces [GUIs]
- G06T 2207/10016: Video; Image sequence
- G06T 2207/10032: Satellite or aerial image; Remote sensing
- G06T 2207/20081: Training; Learning
- G06T 2207/20084: Artificial neural networks [ANN]
- G06T 2207/30184: Infrastructure
- G06T 2207/30232: Surveillance
- G06T 2210/04: Architectural design, interior design
- G06T 2210/62: Semi-transparency
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

Methods for tracking construction site progress, such as for a site that includes a building based on 3D digital representation. Building information modeling (BIM) may provide a digital representation of the physical and functional characteristics of a place, such as a site with a building, the building optionally based on 3D digital representation.

## Claims

### Claim 1

A method for tracking construction site progress comprising:
a. providing a digital representation of a 3D model of at least part of the site;
b. capturing one or more images at the site with a camera;
c. using a machine learning model to identify objects of the 3D model in said captured images, said machine learning model having been trained on objects of at least one 3D model using information about a visual appearance, coordinates, and surface normals of said objects in said at least one 3D model; and
d. comparing the identified objects against the 3D model to conclude which objects of the 3D model are missing at the site.

### Claim 2 (depends on 1)

The method according to claim 1, further comprising:
e. classifying any identified objects of the 3D model into at least one of groups:
i. superfluous objects;
ii. misplaced objects;
iii. wrong objects; and
iv. wrongly installed objects.

### Claim 3 (depends on 1)

The method according to claim 1, further comprising identifying any deform, poorly or faulty constructed objects.

### Claim 4 (depends on 1)

The method according to claim 1, further comprising identifying features of the 3D model.

### Claim 5 (depends on 1)

The method according to claim 1, wherein images from said camera are collected and stored together with input from sensors detecting motion of said camera.

### Claim 6 (depends on 1)

The method according to claim 1, wherein said camera is carried and moved by a drone or a person.

### Claim 7 (depends on 1)

The method according to claim 1, comprising a step of generating output which is readable on a mobile device to show missing, superfluous, misplaced objects, wrong and/or wrongly installed objects of the 3D model.

### Claim 8 (depends on 1)

The method according to claim 1, comprising a step of generating output which is readable on a browser running on a computer to provide a report showing missing, superfluous, misplaced objects, wrong and/or wrongly installed objects of the 3D model.

### Claim 9 (depends on 1)

The method according to claim 1, comprising applying a filter to generate output showing missing, superfluous, partly finished and/or misplaced objects and/or poorly or faulty constructed objects of the 3D model selectable from one or more categories.

### Claim 10 (depends on 9)

The method according to claim 9, wherein the categories are selected among walls, windows, doors, ventilation tubes, ventilation vents, water tubes, electrical installations, and fire alert systems.

### Claim 11 (depends on 1)

The method according to claim 1, wherein a user applies a filter to the 3D model to provide a simplified model of the 3D model.

### Claim 12 (depends on 1)

The method according to claim 1, wherein tracking site progress comprises monitoring progress of construction of infrastructure.

### Claim 13 (depends on 1)

The method according to claim 1, wherein quantities of objects are summarized, allowing correlation with budgets, supply chain and/or detection of waste or theft at the site.

### Claim 14

A method for tracking construction site progress comprising:
a. providing a digital representation of a 3D model, optionally comprising metadata;
b. capturing one or more images comprising pixels at the site with a camera;
c. providing a location and a direction of said image;
d. rendering the 3D model into a virtual projection of a part comprising virtual pixels; wherein said virtual projection comprises for each virtual pixel:
i. coordinates of the rendered virtual projection of the 3D model;
ii. a surface normal of the rendered virtual projection of the 3D model; and
iii. any Building Information Modeling (BIM) identification marker of the rendered virtual projection of the 3D model;
e. feeding said rendered virtual projection into a digital processor, thereby allowing displaying or comparing a relevant part of the 3D model and the image in the same format;
f. using a machine learning model to identify objects of the 3D model in said captured images, said machine learning model having been trained on objects of at least one 3D model using information about a visual appearance, coordinates, and surface normals of said objects in said at least one 3D model; and
g. comparing the identified objects against the 3D model to conclude which objects of the 3D model are missing at the site.

### Claim 15 (depends on 14)

The method according to claim 14, wherein said virtual projection further comprises for each virtual pixel a distance from the rendered part of the 3D Model to the location.

### Claim 16 (depends on 14)

The method according to claim 14, wherein said virtual projection further comprises for each virtual pixel binary information indicating if the rendered part of the 3D model faces the camera.

### Claim 17 (depends on 14)

The method according to claim 14, comprising visually presenting (i) the 3D model, (ii) the one or more images and (iii) any identified missing objects by at least one of the following:
I. visual representations of (i) the 3D model with the corresponding (ii) image side by side;
II. a generated composite image of (i) and (ii), wherein the transparency of (i) or (ii) may be toggled by a user;
III. the images (ii) with any missing objects (iii) of the 3D model;
IV. a 2D plan providing information about any missing objects; and/or
V. a summary report comprising information about any missing object.

### Claim 18 (depends on 1)

A system comprising a processor and memory, the system configured to perform steps a., c., and d. of claim 1, and to receive input of images captured by a camera in step b. of claim 1.

### Claim 19 (depends on 4)

The method according to claim 4, wherein said features are selected among openings, voids, and holes.

### Claim 20 (depends on 7)

The method according to claim 7, wherein said mobile device is selected among a mobile phone and a tablet computer.

### Claim 21 (depends on 11)

The method according to claim 11, wherein said simplified model is selected among an architectural model, a structural model, a mechanical model, an electrical model, a plumbing model, a ventilation model and a fire protection model.

### Claim 22 (depends on 12)

The method according to claim 12, wherein said infrastructure is selected among roads, tunnels, bridges, railways, stations, and electrical grids.

### Claim 23 (depends on 14)

The method according to claim 14, wherein said virtual projection is an equirectangular projection.

### Claim 24 (depends on 23)

The method according to claim 23, wherein said equirectangular projection is a cube map with six sides.

### Claim 25 (depends on 14)

The method according to claim 14, wherein said digital processor is selected among a CPU and a GPU.

### Claim 26 (depends on 1)

A non-transitory computer-readable storage medium storing executable instructions that, when executed by a hardware processor, cause the hardware processor to perform steps a., c., and d. of claim 1 using image data from step b. of claim 1.

### Claim 27 (depends on 14)

A system comprising a processor and memory, the system configured to perform steps a., d., e., f., and g., of claim 14, and to receive input of image data of steps b. and c. of claim 14.

### Claim 28 (depends on 14)

A non-transitory computer-readable storage medium storing executable instructions that, when executed by a hardware processor, cause the hardware processor to perform steps a., d., e., f., and g., of claim 14 using image data from steps b. and c. of claim 14.

### Claim 29 (depends on 1)

The method according to claim 1, wherein said training of said machine learning model is done using the 3D model of step a., and/or one or more different 3D models.

### Claim 30 (depends on 14)

The method according to claim 14, wherein said training of said machine learning model is done using the 3D model of step a., and/or one or more different 3D models.
