US 12,361,389 B2Granted

Vehicle sharing service optimization

AssigneeVolvo Car Corporation

InventorsJoão Pedro Cavaco Pinto Coelho, Sébastien Gérard Bernard Geffraye, Frida Birgitta Schedin, Saroosh Shabbir, Per Jesper Daniel Fredriksson

Life of this patent

Estimating the term…

patentgrep

Loading patent…

Abstract

Vehicle sharing service optimization (e.g., using a computerized tool) is enabled. For example, a system can comprise a memory that stores computer executable components, and a processor that executes the computer executable components stored in the memory, wherein the computer executable components comprise: an intent grouping component that, using a defined intent grouping algorithm, groups intents of a user into an intent group, wherein the intents comprise respective requests to schedule an appointment to use a vehicle, a demand component that, based on the intents of the intent group and using a defined demand algorithm, determines a demand of the user, wherein the demand comprises one or more parameters representative of a need, of the user, for the vehicle, and an output component that generates an output representative of the demand.

Claims20

Dependent claims

A system, comprising:

a memory that stores computer executable components;
and
a processor that executes at least one of the computer executable components that:
trains, using a historical database of past intents for shared vehicles, a machine learning model to determine demands of users for the shared vehicles, wherein the historical database comprises the past intents, past demands associated with the past intents, and past characteristics of the past intents, and wherein the training comprises:
for each past user:
grouping past intents of a past user into groups, wherein the grouping comprises:
based on ordering the past intents of the past user according to respective start times, determine respective differences in start times between the past intents of the past user, and determine respective differences in durations between the past intents of the past user;
and
for each pair of past intents:
for a pair of past intents satisfying a first criterion comprising the pair of past intents having a respective difference in start times being more than a first defined threshold or a respective difference in durations being more than a second defined threshold, and the pair of past intents not overlapping, group the pair of past intents with other pairs of past intents that satisfy the first criterion, and
for the pair of past intents satisfying a second criterion comprising the pair of past intents having the respective difference in start times not being more than the first defined threshold or the respective difference in durations being more than the second defined threshold, the pair of past intents not overlapping, and a ratio of durations of the pair of past intents being outside a defined range, group the pair of past intents with other pairs of past intents that satisfy the second criterion;
and
for each group of past intents, determine a respective demand for the group based on the past characteristics of the past intents in the group;
groups, using the machine learning model, intents of a user into an intent group, wherein the intents comprise respective requests associated with scheduling an appointment to use a shared vehicle;
determines, using the machine learning model, based on the intents of the intent group, a demand of the user, wherein the demand comprises one or more parameters representative of a need, of the user, for the shared vehicle;
remotely controls an autonomous vehicle of the shared vehicles to drive to a defined location based on the demand;
and
generates an output representative of the demand.

15 more elements

A non-transitory machine-readable medium, comprising executable instructions that, when executed by a processor, facilitate performance of operations, comprising:

training, using a historical database of past intents for shared vehicles, a machine learning model to determine demands of users for the shared vehicles, wherein the historical database comprises the past intents, past demands associated with the past intents, and past characteristics of the past intents, and wherein the training comprises:
for each past user:
grouping past intents of a past user into groups, wherein the grouping comprises:
based on ordering the past intents of the past user according to respective start times, determining respective differences in start times between the past intents of the past user, and determining respective differences in durations between the past intents of the past user;
and
for each pair of past intents of the past user:
for a pair of past intents satisfying a first criterion comprising the pair of past intents having a respective difference in start times being more than a first defined threshold or a respective difference in durations being more than a second defined threshold, and the pair of past intents not overlapping, grouping the pair of past intents with other pairs of past intents that satisfy the first criterion, and
for the pair of past intents satisfying a second criterion comprising the pair of past intents having the respective difference in start times not being more than the first defined threshold or the respective difference in durations being more than the second defined threshold, the pair of past intents not overlapping, and a ratio of durations of the pair of past intents being outside a defined range, grouping the pair of past intents with other pairs of past intents that satisfy the second criterion;
and
for each group of past intents, determining a respective demand for the group based on the past characteristics of the past intents in the group;
grouping, using the machine learning model, intents of a user into an intent group, wherein the intents comprise respective requests associated with scheduling an appointment to use a shared vehicle;
determining, using the machine learning model, based on the intents of the intent group, a demand of the user, wherein the demand comprises one or more parameters representative of a need, of the user, for the shared vehicle;
remotely controlling an autonomous vehicle of the shared vehicles to drive to defined location based on the demand;
and
generating an output representative of the demand.

12 more elements

A method, comprising:

training, by a system comprising a processor, using a historical database of past intents for shared vehicles, a machine learning model to determine demands of users for the shared vehicles, wherein the historical database comprises the past intents, past demands associated with the past intents, and past characteristics of the past intents, and wherein the training comprises:
for each past user:
grouping past intents of a past user into groups, wherein the grouping comprises:
based on ordering the past intents of the past user according to respective start times, determining respective differences in start times between the past intents of the past user, and determining respective differences in durations between the past intents of the past user;
and
for each pair of past intents of the past user:
for a pair of past intents satisfying a first criterion comprising the pair of past intents having a respective difference in start times being more than a first defined threshold or a respective difference in durations being more than a second defined threshold, and the pair of past intents not overlapping, grouping the pair of past intents with other pairs of past intents that satisfy the first criterion, and
for the pair of past intents satisfying a second criterion comprising the pair of past intents having the respective difference in start times not being more than the first defined threshold or the respective difference in durations being more than the second defined threshold, the pair of past intents not overlapping, and a ratio of durations of the pair of past intents being outside a defined range, grouping the pair of past intents with other pairs of past intents that satisfy the second criterion;
and
for each group of past intents, determining a respective demand for the group based on the past characteristics of the past intents in the group;
grouping, by the system, using the machine learning model, intents of a user into an intent group, wherein the intents comprise respective requests associated with scheduling an appointment to use a shared vehicle;
determining, by the system, using the machine learning model, based on the intents of the intent group a demand of the user, wherein the demand comprises one or more parameters representative of a need, of the user, for the shared vehicle;
remotely controlling, by the system, an autonomous vehicle of the shared vehicles to drive to defined location based on the demand;
and
generating, by the system, an output representative of the demand.

12 more elements

Dependencies are inferred from the claim text.

Facts

Application: 18/049,948Filed 26 Oct 2022
Published: US 2024/0144195 A12 May 2024
Granted: 15 Jul 2025
Term ends: Not estimatedThe source records don’t support an estimate.

Classification8 codes

G06Q 10/1093
Time management, e.g. calendars, reminders, meetings or time accounting: calendar-based scheduling for persons or groups
G06Q 30/0645
Buying, selling or leasing transactions: rental transactions; Leasing transactions
G06Q 10/022
Reservations, e.g. for tickets, services or events: reservation systems with overbooking processing
G06Q 10/06315
Resource planning, allocation, distributing or scheduling for enterprises or organisations: needs-based resource requirements planning or analysis