Manufacturing optimization using a multi-tenant machine learning platform

Patent metadata

Field	Value
Publication number	US-2020371509-A1
Application number	US-202016883487-A
Country	US
Kind code	A1
Filing date	May 26, 2020
Priority date	May 24, 2019
Publication date	Nov 26, 2020
Grant date	—

Abstract

Techniques for manufacturing optimization using a multi-tenant machine learning platform are disclosed. A method for manufacturing optimization includes: obtaining physical sensor data, by a manufacturing device associated with a tenant of a multi-tenant machine learning platform; determining, by a machine learning spoke system associated with the tenant, a machine learning parameter based on at least the physical sensor data; preventing exposure of the first physical sensor data of the first manufacturing device to any other tenant of the multi-tenant machine learning platform; transmitting the machine learning parameter from the machine learning spoke system to a machine learning hub system of the multi-tenant machine learning platform; and updating, by the machine learning hub system, a multi-tenant machine learning model based at least on the machine learning parameter.

First claim

What is claimed is:

1 . A method comprising:
obtaining first physical sensor data, by a first manufacturing device associated with a first tenant of a multi-tenant machine learning platform; determining, by a first machine learning spoke system associated with the first tenant, a first machine learning parameter based on at least the first physical sensor data; preventing exposure of the first physical sensor data of the first manufacturing device to any other tenant of the multi-tenant machine learning platform; transmitting the first machine learning parameter from the first machine learning spoke system to a machine learning hub system of the multi-tenant machine learning platform; and updating, by the machine learning hub system, a multi-tenant machine learning model based at least on the first machine learning parameter.

2 . The method of claim 1 , further comprising:
obtaining second physical sensor data, by a second manufacturing device associated with a second tenant of the multi-tenant machine learning platform; determining, by a second machine learning spoke system associated with the second tenant, a second machine learning parameter based on at least the second physical sensor data; preventing exposure of the second physical sensor data of the second manufacturing device to any other tenant of the multi-tenant machine learning platform; transmitting the second machine learning parameter from the second machine learning spoke system to the machine learning hub system; and updating, by the machine learning hub system, the multi-tenant machine learning model based at least on the second machine learning parameter.

3 . The method of claim 1 , further comprising:
executing the multi-tenant machine learning model to determine a manufacturing optimization; and adjusting a second manufacturing device, associated with a second tenant of the multi-tenant learning platform, based at least on the manufacturing optimization.

4 . The method of claim 3 , further comprising:
transmitting the manufacturing optimization from the machine learning hub system to a second machine learning spoke system associated with the second tenant, wherein adjusting the second manufacturing device is performed by the machine learning spoke system responsive to receiving the manufacturing optimization from the machine learning hub system.

5 . The method of claim 3 , wherein the second tenant is subscribed to the multi-tenant machine learning platform to obtain the manufacturing optimization without contributing any machine learning parameter to the machine learning hub system.

6 . The method of claim 1 , wherein obtaining the first physical sensor data comprises obtaining data from a laser sensor of the first manufacturing device.

7 . The method of claim 1 , wherein determining the first machine learning parameter comprises executing computer-readable instructions, stored by the first printing machine and executable by one or more hardware processors of the first printing machine, that implement the first machine learning spoke system.

8 . The method of claim 1 , wherein the first manufacturing device is one of a computer numerical control (CNC) machine and a three-dimensional (3D) printer.

9 . A system comprising:
a first one or more processors; and a first one or more non-transitory computer-readable media storing instructions that, when executed by the first one or more processors, cause a first machine learning spoke system associated with a first tenant of a multi-tenant machine learning platform to perform operations comprising:
obtaining first physical sensor data from a first manufacturing device;
determining a first machine learning parameter based on at least the first physical sensor data;
preventing exposure of the second physical sensor data of the second manufacturing device to any other tenant of the multi-tenant machine learning platform; and
transmitting the first machine learning parameter to a machine learning hub system of the multi-tenant machine learning platform,
wherein the machine learning hub system is configured to update a multi-tenant machine learning model based at least on the first machine learning parameter.

10 . The system of claim 9 , further comprising:
a second one or more processors; and a second one or more non-transitory computer-readable media storing instructions that, when executed by a second one or more processors, cause a second machine learning spoke system associated with a second tenant of the multi-tenant machine learning platform to perform operations comprising:
obtaining second physical sensor data from a second manufacturing device;
determining a second machine learning parameter based on at least the second physical sensor data;
preventing exposure of the second physical sensor data of the second manufacturing device to any other tenant of the multi-tenant machine learning platform; and
transmitting the second machine learning parameter to the machine learning hub system of the multi-tenant machine learning platform,
wherein the machine learning hub system is further configured to update the multi-tenant machine learning model based at least on the second machine learning parameter.

11 . The system of claim 9 , further comprising:
a second one or more processors; and a second one or more non-transitory computer-readable media storing instructions that, when executed by a second one or more processors, cause a second machine learning spoke system associated with a second tenant of the multi-tenant machine learning platform to perform operations comprising:
receiving a manufacturing optimization, generated using the multi-tenant machine learning model, from the machine learning hub system; and
adjusting a second manufacturing device based at least on the manufacturing optimization.

12 . The system of claim 11 , wherein the second tenant is subscribed to the multi-tenant machine learning platform to obtain the manufacturing optimization without contributing any machine learning parameter to the machine learning hub system.

13 . The system of claim 9 , wherein the first one or more processors and the first one or more non-transitory computer-readable media are components of the first manufacturing device.

14 . The system of claim 9 , wherein the first manufacturing device is one of a computer numerical control (CNC) machine and a three-dimensional (3D) printer.

15 . A system comprising:
one or more processors; and one or more non-transitory computer-readable media storing instructions that, when executed by the one or more processors, cause a machine learning hub system of a multi-tenant machine learning platform to perform operations comprising:
receiving a first machine learning parameter from a first machine learning spoke system associated with a first tenant of the multi-tenant machine learning platform, the first machine learning parameter having been generated by the first machine learning spoke system based at least on first physical sensor data obtained from a first manufacturing device;
preventing exposure of the first physical sensor data of the first manufacturing device to any other tenant of the multi-tenant machine learning platform; and
updating a multi-tenant machine learning model based at least on the first machine learning parameter.

16 . The system of claim 15 , the one or more non-transitory computer-readable media further storing instructions that, when executed by the one or more processors, cause the machine learning hub system to perform operations comprising:

Assignees

• Markforged Inc

Inventors

• Mark Gregory Thomas

Classifications

• G06N3/092

  Reinforcement learning · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06N3/0499

  Feedforward networks · CPC title

• G06N3/0895

  Weakly supervised learning, e.g. semi-supervised or self-supervised learning · CPC title

• G06N3/098

  Distributed learning, e.g. federated learning · CPC title