# US 2023/0410622 A1 · Life safety device with machine learning based analytics

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

- Number: US 2023/0410622 A1
- Document ID: US20230410622A1
- Status: Published application
- Kind: A1
- Application number: 17/845,743
- Filed: 2022-06-21
- Published: 2023-12-21
- Earliest claimed: 2022-06-21
- Assignees: The ADT Security Corporation
- Applicants: The ADT Security Corporation
- Inventors: Ryan Nathanial SANDLER
- Web page: https://patentgrep.com/patent/US20230410622A1

## CPC

- G08B 17/117: by using a detection device for specific gases, e.g. combustion products, produced by the fire (G08B17/103, G08B17/11 take precedence)
- G08B 29/186: {Fuzzy logic; neural networks}
- G08B: SIGNALLING SYSTEMS, e.g. PERSONAL CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS

## Abstract

A detector is provided. The detector includes a sensor configured to detect airborne particulates at a premises and processing circuitry. The processing circuitry is configured to determine a characteristic associated with the detected airborne particulates, compare the characteristic associated with the detected airborne particles to data associated with predefined characteristics of burned materials, detect, based at least on the comparison, presence of a fire; and determine, if the presence of the fire is detected, a characteristic of the fire based on the characteristic associated with the detected airborne particles and the comparison.

## Claims

### Claim 1

A detector, comprising:
at least one sensor configured to detect airborne particulates at a premises;
processing circuitry in communication with the at least one sensor, the processing circuitry configured to:
determine at least one characteristic associated with the detected airborne particulates;
compare the at least one characteristic associated with the detected airborne particles to data associated with predefined characteristics of burned materials;
detect, based at least on the comparison, a presence of a fire; and
determine, if the presence of the fire is detected, at least one characteristic of the fire based on the at least one characteristic associated with the detected airborne particles and the comparison.

### Claim 2 (depends on 1)

The detector of claim 1, wherein the processing circuitry is further configured to determine, based on the at least one characteristic of the fire, at least one fire response characteristic.

### Claim 3 (depends on 2)

The detector of claim 2, wherein the fire response characteristic comprises at least one of an egress point and an ingress point.

### Claim 4 (depends on 1)

The detector of claim 1, wherein the processing circuitry is further configured to determine, based on the at least one characteristic associated with the detected airborne particulates, a visibility condition.

### Claim 5 (depends on 1)

The detector of claim 1, wherein the at least one sensor is configured to detect gas at the premises, and the processing circuitry is further configured to determine at least one characteristic associated with the detected gas.

### Claim 6 (depends on 1)

The detector of claim 1, wherein the at least one characteristic of the fire comprises at least one of burn rate, spread pattern, and an identity of a substance that is burning.

### Claim 7 (depends on 1)

The detector of claim 1, wherein the processing circuitry is further configured to transmit an alarm signal, the alarm signal comprising an indication of a fire response characteristic comprising one of an egress point and an ingress point.

### Claim 8 (depends on 1)

The detector of claim 1, wherein the predefined characteristics of burned materials comprises at least one of smoke yield, smoke composition, soot yield, and soot composition.

### Claim 9 (depends on 1)

The detector of claim 1, wherein the processing circuitry is further configured to request data specific to the detected airborne particulates from a database to retrieve the predefined characteristics of burned materials.

### Claim 10 (depends on 1)

The detector of claim 1, wherein the processing circuitry is further configured to train a model to determine the characteristic of the fire using machine learning.

### Claim 11

A method implemented by a detector, the method comprising:
detecting airborne particulates at a premises;
determining at least one characteristic associated with the detected airborne particulates;
comparing the at least one characteristic associated with the detected airborne particulates to data associated with predefined characteristics of burned materials;
detecting, based at least on the comparison, a presence of a fire; and
determining, if the presence of the fire is detected, at least one characteristic of the fire based on the at least one characteristic associated with the detected airborne particles and the comparison.

### Claim 12 (depends on 11)

The method of claim 11, further comprising determining, based on the at least one characteristic of the fire, at least one fire response characteristic.

### Claim 13 (depends on 12)

The method of claim 12, wherein the fire response characteristic comprises at least one of an egress point and an ingress point.

### Claim 14 (depends on 11)

The method of claim 11, further comprising determining, based on the at least one characteristic associated with the detected airborne particulates, a visibility condition.

### Claim 15 (depends on 11)

The method of claim 11, further comprising receiving, from the at least one sensor, information pertaining to a detected gas at the premises, and determining at least one characteristic associated with the detected gas.

### Claim 16 (depends on 11)

The method of claim 11, wherein the at least one characteristic of the fire comprises at least one of burn rate, spread pattern, and an identity of a substance that is burning.

### Claim 17 (depends on 11)

The method of claim 11, further comprising transmitting an alarm signal, the alarm signal comprising an indication of a fire response characteristic comprising one of an egress point and an ingress point.

### Claim 18 (depends on 11)

The method of claim 11, wherein the predefined characteristics of burned materials comprises at least one of smoke yield, smoke composition, soot yield, and soot composition.

### Claim 19 (depends on 11)

The method of claim 11, further comprising requesting data specific to the detected airborne particulates from a database to retrieve the predefined characteristics of burned materials.

### Claim 20 (depends on 11)

The method of claim 11, further comprising training a model to determine the characteristic of the fire using machine learning.

### Claim 21

A detector, comprising:
at least one sensor configured to detect airborne particulates at a premises;
processing circuitry configured to:
determine at least one characteristic associated with the detected airborne particulates;
request data specific to the detected airborne particulates from a database to retrieve predefined characteristics of burned materials;
compare the at least one characteristic associated with the detected airborne particles to the predefined characteristics of burned materials;
detect, based at least on the comparison, a presence of a fire;
determine, if the presence of the fire is detected, at least one characteristic of the fire based on the at least one characteristic associated with the detected airborne particles and the comparison; and
train a model to determine the characteristic of the fire using machine learning.
