Intelligent interface for automation multitasking
Techniques for intelligently managing task automation across multiple tasks and multiple automated assistants are provided. In one aspect, an intelligent multitasking system includes: a context manager configured to use human-centric input data to determine a context of a user; an intent mapper configured to map human communication to automated task intents; and an interruption manager configured to schedule automated tasks for performance by automated assistants based on the context of the user and the automated task intents. The context manager can be hosted on a cloud having MQTT clients (e.g., IoT sensors) and an MQTT broker. The intent mapper can include a human-machine software communication interface that identifies verbal and/or non-verbal communications. A method for intelligent multitasking is also provided.
1 . A method for intelligent multitasking, the method comprising:
determining, utilizing human-centric input data, a current activity of a user, wherein the human-centric input data includes input from biometric sensors for the user;
mapping human communication from the user to automated task intents for automated assistants performing automated tasks, wherein the human communication includes an affirmative intent for each automated assistant from the automated assistants to perform each automated task from the automated tasks;
scheduling the automated tasks for performance by the automated assistants based on the current activity of the user and the automated task intents;
queueing a plurality of responses from the automated assistants with respect to the automated tasks; and
sending, based on the input from the biometric sensors, a first response from the plurality of responses to the user, wherein the input from the biometric sensors for the user defines an interruption scheduled for the sending of the first response from the plurality of responses from the automated assistants.
2 . The method of claim 1 , further comprising:
converting verbal communications into utterance vectors;
converting non-verbal communications into the utterance vectors; and
correlating the utterance vectors to matching utterance vectors related to a specific intent of the user.
3 . The method of claim 2 , further comprising:
tagging variations of the human communication that convey a same intent.
4 . The method of claim 1 , further comprising:
determining time slots in a schedule of the user to put the automated assistants in order to perform the automated tasks; and
deciding whether the time slots are proper times to interrupt the user.
5 . The method of claim 1 , wherein the human-centric input data further includes human-computer interactions and software application logs.
6 . The method of claim 1 , wherein the human communication is selected from the group consisting of: verbal communications and non-verbal communications.
7 . The method of claim 1 , wherein the input from the biometric sensors for the user is selected from the group consisting of: heart rate and blood pressure.
8 . The method of claim 1 , further comprising:
sending, during the interruption scheduled as defined by the input from the biometric sensors, the plurality of responses from the automated assistants with respect to the automated tasks.
9 . The method of claim 1 , further comprising:
postponing the interruption scheduled for the plurality of responses from the automated assistants with respect to the automated tasks based on an activation of one or more of the biometric sensors.