Granted Patent B2

US 12,617,419 · App. 18/542,554 · Granted May 5, 2026

Inventors: Michelle Brachman (Quincy, MA); Zahra Ashktorab (Brooklyn, NY); Michael Desmond (White Plains, NY); Hyo Jin Do (Cambridge, MA); Casey Dugan (Cambridge, MA); James Johnson (Somerville, MA); Qian Pan (Canton, MA); Raj Sanjay Shah (Atlanta, GA)

Assignee: International Business Machines Corporation

B60W50/14B60K35/80B60K2360/11B60W2050/146

Abstract

Claims (90)

1 . A method, comprising:

receiving a plurality of best practice rules pertaining to best practices for interacting with a system, wherein the system is a smart vehicle system in a vehicle;

receiving a first user interaction from a user, wherein the first user interaction is a user control of a steering system changing a steering angle, and wherein the first user interaction is detected using a steering sensor;

determining, by a best practice determiner, whether the first user interaction fits at least one of the best practice rules;

responsive to determining, by the best practice determiner, that the first user interaction does not fit the at least one of the best practice rules, generating, using a processor, a modified user interaction by modifying the first user interaction based, at least in part, on the received plurality of best practice rules; and

presenting to the user the modified user interaction.

2 . The method of claim 1 , further comprising:

receiving a second user interaction from the user;

determining, by the best practice determiner, whether the second user interaction fits at least one of the best practice rules; and

responsive to determining, by the best practice determiner, that the second user interaction fits the at least one of the best practice rules, storing the second user interaction as a sample interaction comprising training data.

3 . The method of claim 1 , further comprising:

identifying system documentation for at least one other system;

determining best practices for the at least one other system by the best practice determiner analyzing the system documentation for the at least one other system;

determining whether the best practices for the at least one other system fit the first user interaction by comparing the first user interaction to the best practices for the at least one other system;

responsive to determining that the best practices for the at least one other system fit the first user interaction, generating a rationale explaining why the best practices for the at least one other system fit the first user interaction; and

presenting to the user the rationale explaining why the best practices for the at least one other system fit the first user interaction.

4 . The method of claim 1 , further comprising:

determining differences between the first user interaction and the best practice rules; and

outputting the determined differences to best practice descriptions of best practices for the system.

5 . The method of claim 1 , further comprising:

responsive to the steering sensor detecting the first user interaction, determining, by the best practice determiner, whether an adaptive cruise control for the vehicle is enabled while the first user interaction is detected; and

responsive to determining, by the best practice determiner, that the adaptive cruise control for the vehicle is enabled while the first user interaction is detected, determining, in real time based on the best practice rules, that best practice during use of the adaptive cruise control is for the user to keep control of steering system and for the vehicle not to drift into other lanes,

wherein the modified user interaction comprises a suggestion pertaining to safe operation of the vehicle.

6 . The method of claim 5 , further comprising:

accessing, using a transceiver, best practices for at least one other smart vehicle system;

determining whether the best practices for the at least one other smart vehicle system fit the first user interaction by comparing the first user interaction to the best practices for at least one other smart vehicle system;

responsive to determining that the best practices for the at least one other smart vehicle system fit the first user interaction, generating a rationale explaining why the best practices for the at least one other smart vehicle system fit the first user interaction; and

presenting to the user the rationale explaining why the best practices for the at least one other smart vehicle system fit the first user interaction.

7 . The method of claim 1 , wherein the best practice determiner comprises a large language model.

8 . A system, comprising:

a processor programmed to initiate executable operations comprising:

receiving a plurality of best practice rules pertaining to best practices for interacting with the system;

receiving a first user interaction from a user;

determining, by a best practice determiner, whether the first user interaction fits at least one of the best practice rules;

responsive to determining, by the best practice determiner, that the first user interaction does not fit the at least one of the best practice rules, generating a modified user interaction by modifying the first user interaction based, at least in part, on the received plurality of best practice rules; presenting to the user the modified user interaction;

identifying system documentation for at least one other system;

determining best practices for the at least one other system by the best practice determiner analyzing the system documentation for the at least one other system;

determining whether the best practices for the at least one other system fit the first user interaction by comparing the first user interaction to the best practices for the at least one other system;

responsive to determining that the best practices for the at least one other system fit the first user interaction, generating a rationale explaining why the best practices for the at least one other system fit the first user interaction; and

presenting to the user the rationale explaining why the best practices for the at least one other system fit the first user interaction.

9 . The system of claim 8 , the executable operations further comprising:

receiving a second user interaction from the user;

determining, by the best practice determiner, whether the second user interaction fits at least one of the best practice rules; and

responsive to determining, by the best practice determiner, that the second user interaction fits the at least one of the best practice rules, storing the second user interaction as a sample interaction comprising training data.

10 . The system of claim 8 , the executable operations further comprising:

determining differences between the first user interaction and the best practice rules; and

outputting the determined differences to best practice descriptions of best practices for the system.

11 . The system of claim 8 , wherein;

the system is a smart vehicle system in a vehicle;

the first user interaction from the user is a user control of a steering system changing a steering angle; and

the executable operations further comprise detecting, using a steering sensor, the first user interaction.

12 . The system of claim 11 , the executable operations further comprising:

responsive to the steering sensor detecting the first user interaction, determining, by the best practice determiner, whether an adaptive cruise control for the vehicle is enabled while the first user interaction is detected; and

responsive to determining, by the best practice determiner, that adaptive cruise control for the vehicle is enabled while the first user interaction is detected, determining, in real time based on the best practice rules, that best practice during use of the adaptive cruise control is for the user to keep control of steering system and for the vehicle not to drift into other lanes,

wherein the modified user interaction comprises a suggestion pertaining to safe operation of the vehicle.

13 . The system of claim 12 , the executable operations further comprising:

accessing, using a transceiver, best practices for at least one other smart vehicle system;

determining whether the best practices for the at least one other smart vehicle system fit the first user interaction by comparing the first user interaction to the best practices for at least one other smart vehicle system;

responsive to determining that the best practices for the at least one other smart vehicle system fit the first user interaction, generating a rationale explaining why the best practices for the at least one other smart vehicle system fit the first user interaction; and

presenting to the user the rationale explaining why the best practices for the at least one other smart vehicle system fit the first user interaction.

14 . The system of claim 8 , wherein the best practice determiner uses a transformer model architecture.

15 . A computer program product, comprising:

one or more computer readable storage mediums having program code stored thereon, the program code stored on the one or more computer readable storage mediums collectively executable by a data processing system to initiate operations including:

receiving a plurality of best practice rules pertaining to best practices for interacting with a system, wherein the system is a smart vehicle system in a vehicle;

receiving a first user interaction from a user, wherein the first user interaction from the user is a user control of a steering system changing a steering angle, and wherein the first user interaction is detected using a steering sensor;

determining, by a best practice determiner, whether the first user interaction fits at least one of the best practice rules;

responsive to determining, by the best practice determiner, that the first user interaction does not fit the at least one of the best practice rules, generating a modified user interaction by modifying the first user interaction based, at least in part, on the received plurality of best practice rules; and

presenting to the user the modified user interaction.

16 . The computer program product of claim 15 ,

wherein the program code is executable by the data processing system to initiate operations further comprising:

receiving a second user interaction from the user;

determining, by the best practice determiner, whether the second user interaction fits at least one of the best practice rules; and

responsive to determining, by the best practice determiner, that the second user interaction fits the at least one of the best practice rules, storing the second user interaction as a sample interaction comprising training data.

17 . The computer program product of claim 15 ,

wherein the program code is executable by the data processing system to initiate operations further comprising:

identifying system documentation for at least one other system;

determining best practices for the at least one other system by the best practice determiner analyzing the system documentation for the at least one other system;

determining whether the best practices for the at least one other system fit the first user interaction by comparing the first user interaction to the best practices for the at least one other system;

responsive to determining that the best practices for the at least one other system fit the first user interaction, generating a rationale explaining why the best practices for the at least one other system fit the first user interaction; and

presenting to the user the rationale explaining why the best practices for the at least one other system fit the first user interaction.

18 . The computer program product of claim 15 ,

wherein the program code is executable by the data processing system to initiate operations further comprising:

determining differences between the first user interaction and the best practice rules; and

outputting the determined differences to best practice descriptions of best practices for the system.

19 . The computer program product of claim 15 ,

wherein the program code is executable by the data processing system to initiate operations further comprising:

responsive to the steering sensor detecting the first user interaction, determining, by the best practice determiner, whether an adaptive cruise control for the vehicle is enabled while the first user interaction is detected; and

responsive to determining, by the best practice determiner, that adaptive cruise control for the vehicle is enabled while the first user interaction is detected, determining, in real time based on the best practice rules, that best practice during use of the adaptive cruise control is for the user to keep control of steering system and for the vehicle not to drift into other lanes,

wherein the modified user interaction comprises a suggestion pertaining to safe operation of the vehicle.

20 . The computer program product of claim 15 , wherein the best practice determiner determines whether adaptive cruise control is enabled.

Assignments (1)

Continuity (1)

References Cited (36)