Task execution and delegation by autonomous mobile agents based on intent knowledge base
View Patent ↗A Grid application framework uses semantic languages to describe the tasks and resources used to complete them. A Grid application execution framework comprises a plurality of mobile agents operable to execute one or more tasks described in an intent based task specification language, Input/Output circuitry operable to receive input that describes a task in the task specification language, an analysis engine for generating a solution to the described task, and an intent knowledge base operable to store information contained within tasks of the plurality of mobile agents.
1. A Grid application execution system comprising:
at least one processor;
an input/output circuitry operable to receive as input one or more tasks described in an intent based task specification language, wherein the intent based task specification language describes the context in which the tasks are composed and the end goal the tasks are supposed to achieve;
a plurality of agents executable by the at least one processor to perform the one or more tasks,
wherein the plurality of agents includes a master agent that subdivides a task from the one or more tasks into one or more subtasks and delegates the subtasks to other agents by allocating the other agents to perform the specified subtasks,
wherein the other agents who receive the subtasks are subordinate agents to the master agent and have sufficient domain and resource knowledge to further subdivide the subtasks and delegate the subdivided subtasks to other agents,
wherein each agent of the plurality of agents has a personality comprising a set of characteristics and a pattern of resource use and result delivery, wherein each agent personality is matched against task or subtask requirements to determine the agent best suited to each task or subtask; and
an intent knowledge base operable to store task history, agent history, services information, and heuristics, wherein the plurality of agents use the intent knowledge base to evaluate, delegate, and perform the tasks.
2. A Grid application execution method performed by at least one processor, the method comprising:
receiving at an input/output circuitry of the at least one processor as input one or more tasks described in an intent based task specification language, wherein the intent based task specification language describes the context in which the tasks are composed and the end goal the tasks are supposed to achieve;
performing the one or more tasks by a plurality of agents executable by the at least one processor,
wherein the plurality of agents includes a master agent that subdivides a task from the one or more tasks into one or more subtasks and delegates the subtasks to other agents by allocating the other agents to perform the specified subtasks,
wherein the other agents who receive the subtasks are subordinate agents to the master agent and have sufficient domain and resource knowledge to further subdivide the subtasks and delegate the subdivided subtasks to other agents,
wherein each agent of the plurality of agents has a personality comprising a set of characteristics and a pattern of resource use and result delivery, wherein each agent personality is matched against task or subtask requirements to determine the agent best suited to each task or subtask, and
wherein the plurality of agents use an intent knowledge base to evaluate, delegate, and perform the tasks, the intent knowledge base operable to store task history, agent history, services information, and heuristics.
3. A computer program product embodied in non-transitory computer-readable media that includes program instructions, when executed by at least one processor, performs a Grid application execution method comprising:
receiving at an input/output circuitry of the at least one processor as input one or more tasks described in an intent based task specification language, wherein the intent based task specification language describes the context in which the tasks are composed and the end goal the tasks are supposed to achieve;
performing the one or more tasks by a plurality of agents executable by the at least one processor,
wherein the plurality of agents includes a master agent that subdivides a task from the one or more tasks into one or more subtasks and delegates the subtasks to other agents by allocating the other agents to perform the specified subtasks,
wherein the other agents who receive the subtasks are subordinate agents to the master agent and have sufficient domain and resource knowledge to further subdivide the subtasks and delegate the subdivided subtasks to other agents,
wherein each agent of the plurality of agents has a personality comprising a set of characteristics and a pattern of resource use and result delivery, wherein each agent personality is matched against task or subtask requirements to determine the agent best suited to each task or subtask, and
wherein the plurality of agents use an intent knowledge base to evaluate, delegate, and perform the tasks, the intent knowledge base operable to store task history, agent history, services information, and heuristics.