Install mode and cloud learning for smart windows

Patent metadata

Field	Value
Publication number	US-10425376-B2
Application number	US-201514821366-A
Country	US
Kind code	B2
Filing date	Aug 7, 2015
Priority date	Jan 12, 2015
Publication date	Sep 24, 2019
Grant date	Sep 24, 2019

Abstract

A cloud learning system for smart windows is provided. The system includes at least one server configured to couple via a network to a plurality of window systems, each of the plurality of window systems having at least one control system and a plurality of windows with electrochromic windows and sensors, wherein the at least one server includes at least one physical server or at least one virtual server implemented using physical computing resources. The at least one server is configured to gather first information from the plurality of window systems, and configured to gather second information from sources on the network and external to the plurality of window systems. The at least one server is configured to form at least one rule or control algorithm usable by a window system, based on the first information and the second information, and configured to download the at least one rule or control algorithm to at least one of the plurality of window systems.

First claim

What is claimed is:

1. A cloud learning system for smart windows, comprising:
at least one server configured to couple via a network to a plurality of window systems, each of the plurality of window systems associated with multiple control systems arranged in an hierarchical order each of the plurality of window systems having at least one electrochromic window and sensor, the multiple control systems including a first control system local to each window of a window system, the first control system comprising:
a power supply control module configurable to supply a constant current from a power supply to the at least one electrochromic window, and to stop the supplying of the constant current to the at least one electrochromic window when a sense voltage of the at least one electrochromic window attains a sense voltage limit;

the at least one server configured to gather first information from the plurality of window systems, and configured to gather second information from sources on the network and external to the plurality of window systems;
the at least one server configured to form at least one rule or control algorithm usable by a window system, based on the first information and the second information, and configured to download the at least one rule or control algorithm to at least one of the plurality of window systems; and
the at least one server configured to generate a probabilistic shade model applicable to the plurality of window systems, based on the first information and based on the second information including weather information regarding cloud cover, and based on a geometric model associated with atmospheric clouds, ground clutter, building shapes, and building profiles.

2. The cloud learning system for smart windows of claim 1 , further comprising:
the at least one server configured to determine microclimate weather information from the first information, with the microclimate weather information accessible via a network connection to the at least one server.

3. The cloud learning system for smart windows of claim 1 , further comprising:
the at least one server configured to host a social network based on the plurality of window systems; and
the at least one server configured to collect and offer access to a plurality of rules or control algorithms from or for the plurality of window systems, as a function for the social network.

4. The cloud learning system for smart windows of claim 1 , further comprising:
the at least one server configured to identify manufacturing or aging variances in the plurality of windows of the plurality of window systems based on use of the first information over a span of time.

5. The cloud learning system for smart windows of claim 1 , further comprising:
the at least one server configured to compare operations, control algorithms or rules of differing ones of the plurality of window systems and configured to derive a recommended operation, control algorithm or rule for one of the plurality of window systems based on such a comparison.

6. The cloud learning system for smart windows of claim 1 , further comprising:
the at least one server configured to determine a classification of a user of one of the plurality of window systems, based on the first information; and
the at least one server configured to determine a recommended operation, control algorithm or rule for the one of the plurality of window systems based on the classification of the user.

7. A method for operating a smart window system with cloud learning, comprising:
receiving, at at least one server, information from a plurality of window systems and information available on a network external to the plurality of window systems;
forming, at the at least one server, at least one rule or control algorithm usable by a window system, based on the information from the plurality of window systems and the information available on the network;
sending, from the at least one server to a window system of the plurality of window systems, the at least one rule or control algorithm;
adjusting transmissivity of at least one of a plurality of windows of the window system, based on the at least one rule or control algorithm; and
generating a probabilistic shade model applicable to the plurality of window systems, based on the first information and based on the second information including weather information regarding cloud cover, and based on a geometric model associated with atmospheric clouds, ground clutter, building shapes, and building profiles.

8. The method of claim 7 , further comprising:
determining electricity usage of a building or portion of a building having the window system, based on information from an electric utility, wherein the at least one rule or control algorithm is based on the electricity usage and wherein the at least one rule or control algorithm limits transmissivity of at least a subset of the plurality of smart windows in the building or portion of the building.

9. The method of claim 7 , further comprising:
co-operating the at least one server and at least one application on at least one user device to share a collection of rules or control algorithms applicable to window systems.

10. The method of claim 7 , further comprising:
providing information from sensors of windows of the window system and further window systems to a weather forecasting service or server.

11. The method of claim 7 , further comprising:
generating an environmental model for each of the plurality of windows of the window system, based on information from at least one sensor of each of the plurality of windows and based on the information from the network including weather and daylight information, wherein the environmental model includes a shade model and wherein the at least one rule or control algorithm is based on the environmental model.

12. The method of claim 7 , further comprising:
revising the at least one rule or control algorithm at the at least one server or at the window system, responsive to further information, wherein the smart window system with cloud learning is adaptive.

Assignees

• Kinestral Tech Inc

Inventors

• Nagel Paul
• Barnum Wally
• Coffin Stephen
• Nichols Brandon
• Nagar Ashish
• Bojanczik Kamil
• Ziebarth Jonathan

Classifications

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06N7/005

  Physics · mapped topic

• H04L51/32 Primary

  Electricity · mapped topic

• G05B15/02

  electric · CPC title

• H04L51/52 Primary

  for supporting social networking services · CPC title