Autonomous drone play and directional alignment

A system, comprising:

• a processor;
• and
• a memory, including instructions that when performed by the processor enable the system to:
• in response to receiving a command for a remotely controlled device to perform a behavior, identify a first series of actions that are grouped in a sequence to define the behavior, wherein each action of the first series of actions is a user-directed action for the remotely controlled device to perform in the sequence taught by a user to a machine learning model;
• monitor a performance of the first series of actions when executed by the remotely controlled device, wherein the remotely controlled device performs each action of the first series of action at a base speed for base duration;
• receive feedback related to how the remotely controlled device performs the behavior, wherein the feedback is received from at least one of the user, a second device, or environmental sensors;
• update, according to the feedback, the machine learning model used by the remotely controlled device to produce a second series of actions to perform the behavior that is different from the first series of actions;
• and
• in response to receiving a subsequent command to perform the behavior, instruct the remotely controlled device to perform the second series of actions.

A non-transitory computer-readable medium containing computer program code that, when executed by operation of one or more computer processors, performs an operation comprising:

• in response to receiving a command for a remotely controlled device to perform a behavior, identifying a first series of actions that are grouped in a sequence to define the behavior, wherein each action of the first series of actions is a user-directed action for the remotely controlled device to perform in the sequence taught by a user to a machine learning model;
• monitoring a performance of the first series of actions when executed by the remotely controlled device, wherein the remotely controlled device performs each action of the first series of actions at a base speed for the first series of actions and for a base duration for the first series of actions;
• receiving feedback related to how the remotely controlled device performs the behavior, wherein the feedback is received from at least one of the user, a second device, or environmental sensors;
• updating, according to the feedback, the machine learning model used by the remotely controlled device to produce a second series of actions to perform the behavior that is different from the first series of actions;
• and
• in response to receiving a subsequent command to perform the behavior, instructing the remotely controlled device to perform the second series of actions.