Open-loop control assistant to guide human-machine interaction

Patent metadata

Field	Value
Publication number	US-10608907-B2
Application number	US-201815977063-A
Country	US
Kind code	B2
Filing date	May 11, 2018
Priority date	May 11, 2018
Publication date	Mar 31, 2020
Grant date	Mar 31, 2020

Abstract

An open-loop control assistance (“OLCA”) system can collect data, correlate and aggregate the data, and perform multi-dimensional analytics on the correlated and aggregated data. The OLCA system can then determine plurality of viable options for a next action to be taken by an operator in an open-loop control process, and can determine a specific option as an optimal choice for the operator to select. The OLCA system can present the plurality of viable options and a rationale explaining why the operator should select the specific option. The OLCA system can capture action(s) taken by the operator, and if the action does not correspond to the recommended action, the OLCA system can capture a reason regarding why the optimal choice was not selected. The OLCA system can analyze results from the action(s). The OLCA system can then fine-tune the open-loop control process based upon the results and the cause(s) thereof.

First claim

We claim:

1. An open-loop control assistance system comprising:
a processor; and
memory having instructions stored thereon that, when executed by the processor, cause the processor to perform operations comprising
collecting data from a plurality of sources,
correlating and aggregating the data to create correlated and aggregated data,
performing multi-dimensional analytics on the correlated and aggregated data,
determining, based, at least in part, upon the multi-dimensional analytics, a plurality of viable options for a next action to be taken by an operator in an open-loop control process,
determining a specific option of the plurality of viable options as an optimal choice for the operator to select,
generating a recommendation for the operator to select the optimal choice from the plurality of viable options,
presenting the plurality of viable options to the operator and a rationale explaining why the operator should select the specific option as the optimal choice from the plurality of viable options,
capturing an action taken by the operator, and if the action does not correspond to a specific action identified in the specific option, then further capturing a reason provided by the operator regarding why the optimal choice was not selected,
analyzing results from the action taken by the operator,
determining if the results are positive or negative, and at least one cause thereof, and
fine-tuning the open-loop control process based, at least in part, upon the results and the at least one cause.

2. The open-loop control assistance system of claim 1 , wherein collecting the data from the plurality of sources comprises collecting the data from a network automation platform.

3. The open-loop control assistance system of claim 2 , wherein the network automation platform comprises at least one data lake, a plurality of controllers, and a policy/rule database; and wherein the data comprises big-data from the at least one data lake, logs from the plurality of controllers, and at least one policy/rule from the policy/rule database.

4. The open-loop control assistance system of claim 1 , wherein performing multi-dimensional analytics on the correlated and aggregated data comprises applying an unsupervised learning technique implemented via machine learning to identify clusters among the data and to discover any hidden patterns and signatures contained therein.

5. The open-loop control assistance system of claim 1 , wherein the rationale comprises a plurality of system decision steps taken by the open-loop control assistance system to arrive at the specific option of the plurality of viable options being the optimal choice for the operator to select.

6. The open-loop control assistance system of claim 1 , wherein fine-tuning the open-loop control process comprises updating a process for generating the recommendation for the operator to select the optimal choice from the plurality of viable options.

7. The open-loop control assistance system of claim 1 , wherein fine-tuning the open-loop control process comprises updating a policy, a rule, a model, an algorithm, or a parameter used during at least one previous instance of the open-loop control process.

8. A method comprising:
collecting, by an open-loop control assistance system comprising a processor, data from a plurality of sources;
correlating and aggregating, by the open-loop control assistance system, the data to create correlated and aggregated data;
performing, by the open-loop control assistance system, multi-dimensional analytics on the correlated and aggregated data;
determining, by the open-loop control assistance system, based, at least in part, upon the multi-dimensional analytics, a plurality of viable options for a next action to be taken by an operator in an open-loop control process;
determining, by the open-loop control assistance system, a specific option of the plurality of viable options as an optimal choice for the operator to select;
generating, by the open-loop control assistance system, a recommendation for the operator to select the optimal choice from the plurality of viable options;
presenting, by the open-loop control assistance system, the plurality of viable options to the operator and a rationale explaining why the operator should select the specific option as the optimal choice from the plurality of viable options;
capturing, by the open-loop control assistance system, an action taken by the operator, and if the action does not correspond to a specific action identified in the specific option, then further capturing a reason provided by the operator regarding why the optimal choice was not selected;
analyzing, by the open-loop control assistance system, results from the action taken by the operator;
determining, by the open-loop control assistance system, if the results are positive or negative, and at least one cause thereof; and
fine-tuning, by the open-loop control assistance system, the open-loop control process based, at least in part, upon the results and the at least one cause.

9. The method of claim 8 , wherein collecting the data from the plurality of sources comprises collecting the data from a network automation platform.

10. The method of claim 9 , wherein the network automation platform comprises at least one data lake, a plurality of controllers, and a policy/rule database; and wherein the data comprises big-data from the at least one data lake, logs from the plurality of controllers, and at least one policy/rule from the policy/rule database.

11. The method of claim 8 , wherein performing multi-dimensional analytics on the correlated and aggregated data comprises applying an unsupervised learning technique implemented via machine learning to identify clusters among the data and to discover any hidden patterns and signatures contained therein.

12. The method of claim 8 , wherein the rationale comprises a plurality of system decision steps taken by the open-loop control assistance system to arrive at the specific option of the plurality of viable options being the optimal choice for the operator to select.

13. The method of claim 8 , wherein fine-tuning the open-loop control process comprises updating a process for generating the recommendation for the operator to select the optimal choice from the plurality of viable options.

14. The method of claim 8 , wherein fine-tuning the open-loop control process comprises updating a policy, a rule, a model, an algorithm, or a parameter used during at least one previous instance of the open-loop control process.

15. A computer-readable storage medium comprising computer-executable instructions that, when executed by a processor of an open-loop control assistance system, causes the processor to perform operations comprising:
collecting data from a plurality of sources;
correlating and aggregating the data to create correlated and aggregated data;
performing multi-dimensional analytics on the correlated and aggregated data;
determining, based, at least in part, upon multi-dimensional analytics, a plurality of viable options for a next action to be taken by an operator in an open-loop control process;
determining a specific option of the plurality of viable options as an optimal choice for the operator to select;
generating a recommendation for the operator to select the optimal choice from the plurality of viable options;
presenting the plurality of viable options to the operator and a rationale explaining why the operator should select the specific option as the optimal choice from the plurality of viable options;
capturing an action taken by the operato

Assignees

• At & T Ip I Lp

Inventors

• Chen Chuxin
• Dome George
• Oetting John

Classifications

• G06N5/025

  Extracting rules from data · CPC title

• G06N5/022

  Knowledge engineering; Knowledge acquisition · CPC title

• G06N20/00

  Machine learning · CPC title

• H04L41/0883

  Semiautomatic configuration, e.g. proposals from system · CPC title

• H04L41/0823

  characterised by the purposes of a change of settings, e.g. optimising configuration for enhancing reliability (for optimising operational conditions of wireless networks H04W24/02) · CPC title