Information processing system and autonomous system
AssigneeHitachi, Ltd.
InventorsKanako Esaki, Hiroyuki Mizuno, Tadayuki Matsumura
Life of this patent
Estimating the term…
Abstract
Provided is an information processing system for an autonomous system, the information processing system obtaining observation sensor data in the autonomous system and a target of the autonomous system, and including: an agent management unit that stores and manages an agent group; an attention unit configured to select an agent from the agent group based on the target; and an agent cooperation unit configured to generate a new agent in which inputs and outputs of selected agents are integrated and to train a generative model of the new agent based on the observation sensor data.
Claims6
An information processing system for an autonomous system, wherein the information processing system is configured to obtain observation sensor data regarding the autonomous system and a surrounding environment of the autonomous system and a target operation of the autonomous system, and the information processing system comprising:
- an agent management unit that stores and manages an agent group comprising a plurality of agents, each agent being associated with a corresponding portion of the autonomous system and having a generative model specifying a relational expression between inputs and outputs for that corresponding portion;
- an attention unit configured to select one or more agents from the agent group based on the target operation, wherein each agent of the agent group is configured to control movement of a corresponding portion of the autonomous system;
- and
- an agent cooperation unit configured to generate a new agent as a higher-level agent in which inputs and outputs of selected agents of the agent group are integrated in response to determining that an obtained target operation has not been learned and the attention unit has no basis for an agent selection,
- wherein the agent cooperation unit is configured to train a generative model of the new agent based on the observation sensor data and inputs and outputs of all agents in the agent group, and
- wherein the new agent is configured to control the selected agents of the agent group to achieve the obtained target operation.
Dependencies are inferred from the claim text.