# US 2021/0276457 A1 · Systems and methods for occupant classification

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

- Number: US 2021/0276457 A1
- Document ID: US20210276457A1
- Status: Published application
- Kind: A1
- Application number: 17/261,210
- Filed: 2019-07-17
- Published: 2021-09-09
- Earliest claimed: 2018-07-19
- Assignees: Magna Seating Inc.
- Applicants: Magna Seating Inc.
- Inventors: Bruno D Carraro
- Web page: https://patentgrep.com/patent/US20210276457A1

## CPC

- G01G 19/44: for weighing persons
- B60N 2/0025: {by using weight measurement}
- B60N 2/0027: {for detecting the position of the occupant or of occupant's body part}
- B60N 2/0033: {mounted on or in the foam cushion}
- B60N 2/64: Back-rests {or cushions (B60N2/2222 takes precedence)}
- B60R 21/01516: {using force or pressure sensing means}
- G01G 19/4142: {for controlling activation of safety devices, e.g. airbag systems (electrical circuits for triggering safety arrangements in case of vehicle accidents B60R21/015)}
- G01G 19/52: Weighing apparatus combined with other objects, e.g. furniture (with walking sticks A45B3/08)
- G06N 3/0499: Feedforward networks
- G06N 3/09: Supervised learning
- B60N 2220/30: using a neural network
- G06N 3/08: Learning methods
- G01G: WEIGHING (sorting by weighing B07C5/16)
- B60N: SEATS SPECIALLY ADAPTED FOR VEHICLES; VEHICLE PASSENGER ACCOMMODATION NOT OTHERWISE PROVIDED FOR
- B60R: VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

An occupant classification system for a seat assembly (20) includes a plurality of sensors (32), a posture classifier and a weight classification system. The seat assembly includes a seat cushion (22) and a seat back (24). Each of the plurality of sensors (32) measures a force applied to the seat cushion (22) by an occupant of the seat assembly. The posture classifier identifies a posture of the occupant based on the distribution of forces applied to each of the plurality of sensors (32). The weight classification system identifies a weight class of the occupant based on the posture and the magnitude of forces applied to each of the plurality of sensors (32).

## Claims

### Claim 1

An occupant classification system for a seat assembly wherein the seat assembly includes a seat cushion and a seat back, the system comprising:
a plurality of sensors wherein each of the plurality of sensors measures a force applied to the seat cushion by an occupant of the seat assembly;
a posture classifier for identifying a posture of the occupant based on the distribution of forces applied to each of the plurality of sensors; and
a weight classification system for identifying a weight class of the occupant based on the posture and the magnitude of forces applied to each of the plurality of sensors.

### Claim 2 (depends on 1)

The occupant classification system of claim 1 wherein the posture classifier comprises a neural network, a support vector machine, a logistic regression, a decision tree, a Naïve-Bayes or nearest neighbors.

### Claim 3 (depends on 1)

The occupant classification system of claim 1 wherein the weight classification system comprises a deterministic component and a probabilistic component.

### Claim 4 (depends on 3)

The occupant classification system of claim 3 wherein the deterministic component comprises the sum of the forces applied to each of the plurality of sensors.

### Claim 5 (depends on 3)

The occupant classification system of claim 3 wherein the probabilistic component comprises a neural network, a support vector machine, a logistic regression, a decision tree, a Naïve-Bayes, nearest neighbors, regression-based models or a radial basis network.

### Claim 6 (depends on 1)

The occupant classification system of claim 1 further comprising a second plurality of sensors wherein each of the second plurality of sensors measures a force applied to the seat back, wherein:
the posture classifier identifies the posture of the occupant based on the distribution of forces applied to each of the plurality of sensors and the distribution of forces applied to each of the second plurality of sensors; and
the weight classification system identifies the weight class of the occupant based on the posture, the magnitude of forces applied to each of the plurality of sensors and the magnitude of forces applied to each of the second plurality of sensors.

### Claim 7

A method associated with classifying an occupant of a seat assembly, wherein the seat assembly includes a seat cushion and a seat back, the method comprising the steps of:
measuring a plurality of forces applied by the occupant to the seat cushion;
using the plurality of forces to identify a posture of the occupant; and
using the posture and the plurality of forces to identify a weight class of the occupant.

### Claim 8 (depends on 7)

The method of claim 7 wherein a probabilistic method is used to identify the posture.

### Claim 9 (depends on 8)

The method of claim 8 wherein the probabilistic method comprises a neural network, a support vector machine, a logistic regression, a decision tree, a Naïve-Bayes or nearest neighbors.

### Claim 10 (depends on 7)

The method of claim 7 wherein a deterministic method is used to identify the weight class.

### Claim 11 (depends on 10)

The method of claim 10 wherein a probabilistic method is used to identify the weight class if the deterministic method does not identify a single weight class.

### Claim 12 (depends on 11)

The method of claim 11 wherein the deterministic method comprises a sum of the forces applied to each of the plurality of sensors.

### Claim 13 (depends on 12)

The method of claim 12 wherein the probabilistic method comprises a neural network, a support vector machine, a logistic regression, a decision tree, a Naïve-Bayes, nearest neighbors, regression-based models or a radial basis network.

### Claim 14 (depends on 7)

The method of claim 7 further comprising the steps of:
measuring a second plurality of forces applied by the occupant to the seat back;
using the plurality of forces and the second plurality of forces to identify the posture of the occupant; and
using the posture, the plurality of forces and the second plurality of forces to identify the weight class of the occupant.

### Claim 15

A method for deriving an occupant classification system for a seat assembly, wherein the seat assembly includes a seat cushion and a seat back, the method comprising the steps of:
using a probabilistic method to train a posture classifier to differentiate between a plurality of postures;
for each of the plurality of postures, using a deterministic method to derive a weight classification system for identifying one of a plurality of weight classes; and
when the weight classification system is unable to identify the one of the plurality of weight classes, using a second probabilistic method to train the weight classification system to identify the one of the plurality of weight classes.

### Claim 16 (depends on 15)

The method of claim 15 wherein the probabilistic method comprises a neural network, a support vector machine, a logistic regression, a decision tree, a Naïve-Bayes, nearest neighbors, regression-based models or a radial basis network.

### Claim 17 (depends on 15)

The method of claim 15 wherein the second probabilistic method comprises a neural network, a support vector machine, a logistic regression, a decision tree, a Naïve-Bayes, nearest neighbors, regression-based models or a radial basis network.
