Controlling an engagement state of an agent during a human-machine dialog

1. A method of controlling an engagement state of an agent during a human-machine dialog, the method comprising:

receiving a spoken request that is a conditional locking request, wherein the conditional locking request uses a natural language expression including (1) a locking command and (2) a locking condition according to which the agent remains in a locked state;

storing the locking condition in a format that can be evaluated when needed by the agent;

entering a conditionally locked state in response to the conditional locking request;

in the conditionally locked state, receiving a multiplicity of requests without a need for a wakeup indicator; and

for a request from the multiplicity of requests:

evaluating the locking condition upon receiving the request; and

processing the request if the locking condition is true.

2. The method of claim 1 , wherein the locking condition is time-based and the natural language expression includes at least one of the words “after,” “before,” “until,” “when,” and “while”.

3. The method of claim 1 , wherein the locking condition is location-based and the natural language expression includes at least one of the words “in,” “outside,” “near,” “next to,” “where,” “wherever” and “within”.

4. The method of claim 1 , further comprising:

receiving a spoken request that is an unlock request; and

exiting the conditionally locked state in response to the unlock request.

5. The method of claim 1 , further comprising:

receiving a timeout; and

exiting the conditionally locked state in response to the timeout.

6. The method of claim 1 , wherein:

the conditional locking request is spoken by a first user; and

the method further includes ignoring a wakeword spoken by a second user, the first user and the second user being different users.

7. The method of claim 6 , wherein the ignoring of the wakeword spoken by the second user is performed as a result of the first user having a higher ranking than the second user.

8. An automotive virtual assistant subsystem supporting human-machine dialog, the subsystem comprising software instructions for:

receiving a spoken request that is a conditional locking request, wherein the conditional locking request uses a natural language expression including (1) a locking command and (2) a locking condition according to which the subsystem remains in a locked state;

storing the locking condition in a format that can be evaluated when needed by the subsystem;

entering a conditionally locked state in response to the conditional locking request;

in the conditionally locked state, receiving a multiplicity of requests without a need for a wakeup indicator; and

for a request from the multiplicity of requests:

evaluating the locking condition upon receiving the request; and

processing the request if the locking condition is true.

9. The automotive virtual assistant subsystem of claim 8 , wherein the locking condition is time-based and the natural language expression includes at least one of the words “after,” “before,” “until,” “when,” and “while”.

10. The automotive virtual assistant subsystem of claim 8 , wherein the locking condition is location-based and the natural language expression includes at least one of the words “in,” “outside,” “near,” “next to,” “where,” “wherever” and “within”.

11. The automotive virtual assistant subsystem of claim 8 , further comprising:

receiving a spoken request that is an unlock request; and

exiting the conditionally locked state in response to the unlock request.

12. The automotive virtual assistant subsystem of claim 8 , further comprising:

receiving a timeout; and

exiting the conditionally locked state in response to the timeout.

13. A non-transitory computer-readable recording medium having computer instructions recorded thereon for controlling an engagement state of an agent during a human-machine dialog, the computer instructions, when executed on one or more processors, causing the one or more processors to implement operations comprising:

receiving a spoken request that is a conditional locking request, wherein the conditional locking request uses a natural language expression including (1) a locking command and (2) a locking condition according to which the agent remains in a locked state;

storing the locking condition in a format that can be evaluated when needed by the agent;

entering a conditionally locked state in response to the conditional locking request;

in the conditionally locked state, receiving a multiplicity of requests without a need for a wakeup indicator; and

for a request from the multiplicity of requests:

evaluating the locking condition upon receiving the request; and

processing the request if the locking condition is true.

14. The non-transitory computer-readable recording medium of claim 13 , wherein the locking condition is time-based and the natural language expression includes at least one of the words “after,” “before,” “until,” “when,” and “while”.

15. The non-transitory computer-readable recording medium of claim 13 , wherein the locking condition is location-based and the natural language expression includes at least one of the words “in,” “outside,” “near,” “next to,” “where,” “wherever” and “within”.

16. The non-transitory computer-readable recording medium of claim 13 , wherein the operations further comprise:

receiving a spoken request that is an unlock request; and

exiting the conditionally locked state in response to the unlock request.

17. The non-transitory computer-readable recording medium of claim 13 , wherein the operations further comprise:

receiving a timeout; and

exiting the conditionally locked state in response to the timeout.