Motion estimation apparatus, motion estimation method, path generation apparatus, path generation method, and computer-readable recording medium

A path generation apparatus comprising:

• one or more memories storing instructions and a model;
• and
• one or more processors configured to execute the instructions to:
• generate first motion analysis data representing actual motion of a mobile object in a first environment, the mobile object being an autonomous vehicle or an autonomous robot;
• analyze the first environment, based on environment state data representing a state of the first environment, and generating environment analysis data;
• input the environment analysis data to the model for estimating motion of a mobile object in the first environment and estimating the motion of the mobile object in the first environment to generate motion estimation result data;
• determine whether the first motion analysis data is outside a confidence interval over a predetermined range, wherein the confidence interval is based on motion estimation result data estimated by the model;
• in response to determining that the first motion analysis data is outside the confidence interval over the predetermined range, instruct a learning device for learning the model to relearn the model,
• if the model is relearned, regenerate path data representing a path from a current position to a destination, based on motion estimation result data generated using the relearned model;
• generate information for controlling movement of the mobile object, based on the motion estimation result data and the path data, and control the mobile object to move from the current position to the destination;
• and
• in the current position, if the motion estimation result data is higher than a risk threshold, generate the path data since the path is to be modified, or in the current position, if the motion estimation result data is equal to or lower than the risk threshold, the path is not to be modified.

A motion estimation method performed by a motion estimation apparatus, wherein

• the motion estimation apparatus storing a model and comprises:
• first sensors configured to measure a state of a mobile object, the mobile object being an autonomous vehicle or an autonomous robot;
• second sensors configured to measure a state of a first environment surrounding the mobile object;
• wherein the motion estimation method comprises:
• analyzing mobile object state data measured by the first sensors, and generate first motion analysis data representing actual motion of the mobile object in the first environment;
• analyzing environment state data measured by the second sensors, and generate environment analysis data representing a state of the first environment;
• inputting the environment analysis data to the model for estimating motion of the mobile object in the first environment, and estimate the motion of the mobile object in the first environment to generate motion estimation result data;
• determining whether the first motion analysis data is outside a confidence interval over a predetermined range, wherein the confidence interval is based on motion estimation result data estimated by the model;
• and
• in response to determining that the first motion analysis data is outside the confidence interval over the predetermined range, instructing a learning model to relearn the model.

A path generation method performed by a path generation apparatus storing a model, the path generation method comprising:

• generating first motion analysis data representing actual motion of a mobile object in a first environment, the mobile object being an autonomous vehicle or an autonomous robot;
• analyzing the first environment, based on environment state data representing a state of the first environment, and generating environment analysis data;
• inputting the environment analysis data to the model for estimating motion of a mobile object in the first environment and estimating the motion of the mobile object in the first environment to generate motion estimation result data;
• determining whether the first motion analysis data is outside a confidence interval over a predetermined range, wherein the confidence interval is based on motion estimation result data estimated by the model;
• in response to determining that the first motion analysis data is outside the confidence interval over the predetermined range, instructing a learning device for learning the model to relearn the model;
• if the model is relearned, regenerating path data representing a path from a current position to a destination, based on motion estimation result data generated using the relearned model;
• generating information for controlling movement of the mobile object, based on the motion estimation result data and the path data, and controlling the mobile object to move from the current position to the destination;
• and
• in the current position, if the motion estimation result data is higher than a risk threshold, generating the path data since the path is to be modified, or in the current position, if the motion estimation result data is equal to or lower than the risk threshold, the path is not to be modified.

A non-transitory computer-readable recording medium storing a program performed by a path generation apparatus to perform processing, the path generation apparatus storing a model, the processing comprising:

• generating first motion analysis data representing actual motion of a mobile object in a first environment, the mobile object being an autonomous vehicle or an autonomous robot;
• analyzing the first environment, based on environment state data representing a state of the first environment, and generating environment analysis data;
• inputting the environment analysis data to the model for estimating motion of a mobile object in the first environment and estimating the motion of the mobile object in the first environment to generate motion estimation result data;
• determining whether the first motion analysis data is outside a confidence interval over a predetermined range, wherein the confidence interval is based on motion estimation result data estimated by the model;
• in response to determining that the first motion analysis data is outside the confidence interval over the predetermined range, instructing a learning device for learning the model to relearn the model;
• generating information for controlling movement of the mobile object, based on the motion estimation result data and path data, and controlling the mobile object to move from a current position to a destination;
• and
• in the current position, if the motion estimation result data is higher than a risk threshold, generating the path data since the path is to be modified, or in the current position, if the motion estimation result data is equal to or lower than the risk threshold, the path is not to be modified.