Skill platform system, skill modeling device, and skill dissemination method

1. A skill platform system comprising:

a processor; and

a storage device in which a program is stored, the program controlling the processor to function as:

a data acquisition unit configured to acquire data on work of workers at a first site in real space to construct a first digital twin that reproduces in cyberspace the work of the workers in real space, wherein the data includes data on input to the workers and data on output from the workers, the data on output being data indicative of lines of sight of the workers and/or behavior content of the workers;

a data extraction unit configured to extract data on work of skilled workers from the data of the first digital twin to construct a second digital twin that reproduces the work of the skilled workers in the cyberspace;

a model creation unit configured to perform, using a computer model, modeling of a skill of the skilled workers by extracting, from the data on input and the data on output in the second digital twin, relations between the data on input and the data on output having high correlation, and by creating information indicative of the extracted relations as the computer model; and

a skill provision unit configured to provide the skill of the skilled workers to a second site in the real space by using the computer model created by the model creation unit.

2. The skill platform system of claim 1 , wherein the program further controls the processor to function as

a customization unit configured to customize the computer model allocated to the second site, based on evaluation data indicative of evaluation at the second site for the skill of the skilled worker provided by the skill provision unit.

3. The skill platform system of claim 1 , wherein the skill provision unit is configured to estimate an action to be taken at the second site, using the computer model, from input information on work at the second site.

4. The skill platform system of claim 1 , wherein the model creation unit is configured to create as the computer model a neural network by using the data on input and the data on output as learning data.

5. The skill platform system of claim 4 , wherein the neural network comprises at least one of a fully connected neural network, a convolutional type neural network, a recurrent type neural network, and a self-encoding neural network.

6. The skill platform system of claim 1 , wherein the skill provision unit is capable of providing the skill of the skilled worker in a field unit or a process unit.

7. The skill platform system of claim 1 , wherein the data acquired by the data acquisition unit comprises information on an event recognized by the skilled worker with five senses, a captured image of the skilled worker, information on an action of the skilled worker, information on a condition of a device that the skilled worker handles, or information on an achievement of the skilled worker.

8. The skill platform system of claim 1 , wherein:

a first computer model corresponding to a first field or a first process is created with data on work of a first skilled worker engaged in work of the first field or the first process, and a second computer model corresponding to a second field or a second process is created with data on work of a second skilled worker engaged in work of the second field or the second process.

9. The skill platform system of claim 8 , wherein:

the data acquisition unit is configured to acquire data on a plurality of sites including the first site; and

the first skilled worker and the second skilled worker are workers of different sites.

10. The skill platform system of claim 1 , wherein the data acquisition unit is configured to acquire the data by a simulator.

11. The skill platform system of claim 10 , wherein the simulator includes a tactile sensation supply simulator for giving tactile sensation to at least one of a finger and a palm of the skilled worker.

12. The skill platform system of claim 10 , wherein the data acquisition unit is configured to cause a trouble on the work to occur in the simulator, and to acquire data on an action of the skilled worker coping with the trouble.

13. The skill platform system of claim 1 , wherein the data indicative of the lines of sight of the workers and/or the behavior content of the workers comprises data indicative of the lines of sight of the workers captured by a camera and tracking pupils of the workers in real time.

14. The skill platform system of claim 1 , wherein

the model creation unit includes one or more work-specific model creation units provided for respective works of plural fields and processes and is configured to

create the computer model for each work of the plural fields and processes, and

adaptively increase or decrease the one or more work-specific model creation units.

15. A skill dissemination method executed by a computer, the method comprising:

acquiring data on work of workers at a first site in real space to construct a first digital twin that reproduces the work of the workers in cyberspace, wherein the data includes data on input to the workers and data on output from the workers, and the data on output being data indicative of lines of sight of the workers and/or behavior content of the workers;

extracting data on work of skilled workers from the data of the first digital twin to construct a second digital twin that reproduces the work of the skilled workers in the cyberspace;

modeling, using a computer model, a skill of the skilled worker by extracting, from the data on input and the data on output in the second digital twin, relations between the data on input and the data on output having high correlation, and by creating information indicative of the extracted relations as the computer model; and

providing the skill of the skilled worker to a second site in the real space by using the computer model created by the modeling.

16. The method of claim 15 , wherein the modeling comprises creating as the computer model a neural network by deep learning using the data on input and the data on output as learning data.

17. The skill dissemination method of claim 15 , wherein

the modeling includes

creating the computer model for each work of plural fields and processing, by one or more work-specific model creation units provided for respective works of the plural fields and processes, and

adaptively increasing or decreasing the one or more work-specific model creation units.