# US 2023/0419105 A1 · Ensemble of narrow ai agents for vehicles

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

- Number: US 2023/0419105 A1
- Document ID: US20230419105A1
- Status: Published application
- Kind: A1
- Application number: 18/036,150
- Filed: 2021-04-17
- Published: 2023-12-28
- Earliest claimed: 2019-11-07
- Assignees: Autobrains Technologies Ltd
- Applicants: Cartica Ai Ltd.
- Inventors: Igal Raichelgauz; Karina Odinaev
- Web page: https://patentgrep.com/patent/US20230419105A1

## CPC

- G06N 20/20: Ensemble learning
- B25J 9/161: {Hardware, e.g. neural networks, fuzzy logic, interfaces, processor}
- B60W 60/001: {Planning or execution of driving tasks}
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G06N 5/043: Distributed expert systems; Blackboards
- B25J 9/163: {learning, adaptive, model based, rule based expert control}
- G05B 19/0426: {Programming the control sequence}
- G06N 3/045: Combinations of networks
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- B25J: MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES ({manipulators specially adapted for use in surgery A61B34/70; manipulators used in cleaning hollow articles B08B9/04}; manipulators associated with rolling mills B21B39/20; manipulators associated with forging machines B21J13/10; {manipulators associated with picking-up and placing mechanisms B23P19/007}; means for holding wheels or parts thereof B60B30/00; {vehicles with ground-engaging propulsion means, e.g. walking members B62D57/02, B62D57/032; devices for picking-up and depositing articles or materials between conveyors B65G47/90, B65G47/91; manipulators with gripping or holding means for transferring packages B65H67/065}; cranes B66C; {manipulators used in the protection or supervision of pipe-line installations F17D5/00; walking equipment adapted for nuclear steam-generators F22B37/006}; manipulators specially adapted for, or associated with, nuclear reactors G21C; {apparatus used for handling wafers during manufacture or treatment of semiconductor H10P72/50})
- B60W: CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- 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)

## Abstract

A method for operating an ensemble of narrow AI agents related to a vehicle, the method may include obtaining one or more sensed information units; determining, by a perception unit and based on the one or more sensed information units, one or more relevant narrow AI agents of the ensemble that are relevant to a processing of the one or more sensed information units; wherein the ensemble is relevant to a first plurality of scenarios; processing the one or more sensed information units, by the one or more relevant narrow AI agents, to provide one or more narrow AI agent driving decisions; wherein each narrow AI agent is relevant to a respective fraction of the first plurality of scenarios; and processing, by a driving decision unit, the one or more narrow AI agent driving decisions to provide an output driving decision.

## Claims

### Claim 1

A method for operating an ensemble of narrow AI agents related to a vehicle, the method comprises:
obtaining one or more sensed information units;
determining, by a perception unit and based on the one or more sensed information units, one or more relevant narrow AI agents of the ensemble that are relevant to a processing of the one or more sensed information units; wherein the ensemble is relevant to a first plurality of scenarios;
processing the one or more sensed information units, by the one or more relevant narrow AI agents, to provide one or more narrow AI agent driving decisions; wherein each narrow AI agent is relevant to a respective fraction of the first plurality of scenarios; and
processing, by a driving decision unit, the one or more narrow AI agent driving decisions to provide an output driving decision.

### Claim 2 (depends on 1)

The method according to claim 1 wherein each relevant narrow AI agent is relevant to a dedicated class.

### Claim 3 (depends on 2)

The method according to claim 2 wherein each class is defined by at least a part of one or more scenarios, wherein the at least part of the one or more scenarios are a fraction of the first plurality of scenarios.

### Claim 4 (depends on 2)

The method according to claim 2 wherein each class is defined by an anchor.

### Claim 5 (depends on 4)

The method according to claim 4 wherein an anchor is a contextual cue.

### Claim 6 (depends on 1)

The method according to claim 1 wherein the determining, by the perception unit, of one or more relevant narrow AI agents of the ensemble, is executed without detection of objects that are below a predefined number of pixels.

### Claim 7 (depends on 6)

The method according to claim 6 wherein the predefined number of pixels does not exceed few tens of pixels.

### Claim 8 (depends on 1)

The method according to claim 1 wherein the perception unit is trained to classify sensed information units to classes, wherein each class is at least a part of one or more scenarios, the one or more scenarios are a fraction of the first plurality of scenarios.

### Claim 9 (depends on 1)

The method according to claim 1 comprising training the perception unit to classify sensed information units to classes, wherein each class is at least a part of one or more scenarios, the one or more scenarios are a fraction of the first plurality of scenarios.

### Claim 10 (depends on 9)

The method according to claim 9 comprising receiving, by the perception unit a definition of at least some of the classes before training.

### Claim 11 (depends on 9)

The method according to claim 9 comprising defining, by the perception unit at least some of the classes.

### Claim 12 (depends on 11)

The method according to claim 11 wherein the defining comprises performing an unsupervised training.

### Claim 13 (depends on 9)

The method according to claim 9 wherein the at least part of one or more scenarios is at least one out of (a) one or more factors of a scenario, (b) one or more element of a scenario, (c) one or more parameters of a scenario, and (d) one or more variables of a scenario.

### Claim 14 (depends on 9)

The method according to claim 9 wherein each narrow AI agent is associated with a dedicated class and the method comprises training each narrow AI agent to output a narrow AI agent driving decision associated with the dedicated class.

### Claim 15 (depends on 14)

The method according to claim 14 wherein the training comprises training each narrow AI agent using sensed information units of the dedicated class.

### Claim 16 (depends on 14)

The method according to claim 14 comprising training the perception unit to associate a driving decision within each class.

### Claim 17 (depends on 1)

The method according to claim 1 wherein the narrow AI agents are end-to-end narrow AI agents.

### Claim 18 (depends on 17)

The method according to claim 17 comprising training the driving decision unit to generated output driving decisions based on at least one narrow AI agent driving decision.

### Claim 19 (depends on 1)

The method according to claim 1 wherein for at least some of the narrow AI agents the respective fraction is smaller than one percent of the first plurality of scenarios.

### Claim 20 (depends on 1)

The method according to claim 1 wherein a number of narrow AI agents relevant to one of the first plurality of scenarios differs from a number of narrow AI agents relevant to another of the first plurality of scenarios.

### Claim 21 (depends on 1)

The method according to claim 1 wherein a number of narrow AI agents exceeds one thousand.

### Claim 22 (depends on 1)

The method according to claim 1 wherein a number of narrow AI agents is exceeds ninety nine thousand.

### Claim 23 (depends on 1)

The method according to claim 1 wherein at least some of the narrow AI agents comprise at least a portion of a neural network.

### Claim 24 (depends on 1)

The method according to claim 1 comprising feeding, by the perception unit the one or more sensed information units to each one of the one or more relevant narrow AI agents.

### Claim 25 (depends on 1)

The method according to claim 1 comprising feeding, by the perception unit the one or more sensed information units to each one of the one or more relevant narrow AI agents and maintaining at least one irrelevant narrow AI agent in a low power mode in which a power consumption of the at least one irrelevant narrow AI agent is lower than a power consumption of a relevant narrow AI agent.

### Claim 26 (depends on 1)

The method according to claim 1 comprising determining which part of the one or more sensed information units to send to each relevant narrow AI agent.

### Claim 27 (depends on 1)

The method according to claim 1 wherein a narrow AI agent driving decision is a command for autonomously controlling a vehicle.

### Claim 28 (depends on 1)

The method according to claim 1 wherein a narrow AI agent driving decision is an advanced driver-assistance systems (ADAS) command.

### Claim 29 (depends on 1)

The method according to claim 1 comprising averaging, by the driving decision unit, the one or more narrow AI agent driving decisions.

### Claim 30 (depends on 1)

The method according to claim 1 wherein each narrow AI agent driving decision of the one or more narrow AI agent driving decisions is associated with a time period.

### Claim 31 (depends on 1)

The method according to claim 1 wherein the processing, by the driving decision unit, of the one or more narrow AI agent driving decisions comprises applying risk reduction optimization.

### Claim 32

A non-transitory computer readable medium that stores instructions for operating an ensemble of narrow AI agents, the operating comprises:
obtaining one or more sensed information units;
determining, by a perception unit and based on the one or more sensed information units, one or more relevant narrow AI agents of the ensemble that are relevant to a processing of the one or more sensed information units; wherein the ensemble is relevant to a first plurality of scenarios;
processing the one or more sensed information units, by the one or more relevant narrow AI agents, to provide one or more narrow AI agent driving decisions; wherein each narrow AI agent is relevant to a respective fraction of the first plurality of scenarios; and
processing, by a driving decision unit, the one or more narrow AI agent driving decisions to provide an output driving decision.

### Claim 33

A system for driving decision determination, the system comprises:
an ensemble of AI narrow agents;
a perception unit that is configured to:
obtain one or more sensed information units; and
determine, based on the one or more sensed information units, one or more relevant narrow AI agents of the ensemble that are relevant to a processing of the one or more sensed information units; wherein the ensemble is relevant to a first plurality of scenarios;
wherein the one or more relevant narrow AI agents are configured to process the one or more sensed information units, to provide one or more narrow AI agent driving decisions; wherein each narrow AI agent is relevant to a respective fraction of the first plurality of scenarios; and
a driving decision unit that is configured to process the one or more narrow AI agent driving decisions to provide an output driving decision.
