Sensing systems

• US 10,180,339 B1
• Filed: 05/09/2016
• Issued: 01/15/2019
• Est. Priority Date: 05/08/2015
• Status: Active Grant

• Current Assignee: Digimarc Corporation
• Original Assignee: Digimarc Corporation
• Inventors: Long, Scott M., Lord, John D.
• Primary Examiner(s): Desta, Elias
• Application Number:
  US15/149,477
• Time in Patent Office: 981 Days
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 20/10 using kernel methods, e.g. support vector machines [SVM]
• CPC Class Codes: G06N 3/045 Combinations of networks
• CPC Class Codes: G06N 3/049 Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
• CPC Class Codes: G06N 3/08 Learning methods
• CPC Class Codes: G06N 3/126 Evolutionary algorithms, e.g. genetic algorithms or genetic programming
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G11C 19/00 Digital stores in which the information is moved stepwise, e.g. shift registers
• CPC Class Codes: G11C 7/1006 Data managing, e.g. manipulating data before writing or reading out, data bus switches or control circuits therefor
• CPC Class Codes: H04L 67/125 involving control of end-device applications over a network
• CPC Class Codes: H04L 67/63 Routing a service request depending on the request content or context
• CPC Class Codes: Y02D 30/70 in wireless communication networks

Abstract

A system learns to automatically identify, and detect, contextual conditions that may serve as action triggers to help please a user (or avoid annoying a user). Among other features, a simple sensor arrangement is detailed which, in addition to producing a customary stream of high bandwidth sensor data, provides an output of low bandwidth data. This low-bandwidth data serves to identify a particular reference pattern with which the high-bandwidth sensor data is found to correspond. Such a sensor can employ reference patterns discovered through pseudo-random trials. A great number of other advantageous features and arrangements are also detailed.

Claims

1. A packaged sensing system including:

a sensing element having an output;

a memory having an input and an output, the input being coupled to the output of the sensing element, the memory being adapted to store an array of plural data representing plural physical quantities sensed by the sensing element; and

a matrix module having an input coupled to the output of the memory, said matrix module including circuitry that multiplies data presented by said output of the memory by a matrix of plural coefficients, the matrix module including an output that presents output data based on a result of said multiplication;

wherein the sensing element, the memory and the matrix module are affixed to a common semiconductor substrate; and

wherein said memory is adapted to store an array of four or more data representing a physical quantity sensed by the sensing element at four or more respective previous times.

2. The packaged sensing system of claim 1 in which the matrix of coefficients comprises plural rows of coefficients, and the result of the multiplication is a vector of plural element values, wherein the matrix module is further adapted to produce output data indicating which element of said result vector is the largest.

3. The packaged sensing system of claim 1 that includes additional circuitry for varying said coefficients.

4. The system of claim 1 in which the sensing element comprises a 3D accelerometer, a temperature sensor, a barometric pressure sensor, or a humidity sensor.

5. The system of claim 1 in which:

the memory comprises a shift register having an input coupled to the output of the sensing element, the shift register presenting a temporal vector of at least four sensor values on at least four outputs; and

the matrix module has inputs coupled to the at least four outputs of the shift register, and being adapted to perform a dot product operation between the temporal vector of sensor values and each of plural reference vectors that comprise said matrix of coefficients, the matrix module further being adapted to provide, on a circuit output, an integer datum indicating one of said reference vectors that yielded a largest dot product result.

6. The system of claim 5 in which the plural reference vectors have an order, and the matrix module is adapted to output, on said circuit output, an integer datum indicating an ordinal position of the reference vector yielding the largest dot product, within said order.

7. The system of claim 5 in which at least certain of the plural reference vectors are fixed, and not able to be altered.

8. The system of claim 7 in which at least certain of the plural reference vectors are stored in a memory that allows alteration of said vectors.

9. A sensing system including:

a sensing element having an output;

a memory having an input and an output, the input being coupled to the output of the sensing element, the memory being adapted to store an array of plural data representing plural physical quantities sensed by the sensing element; and

a matrix module having an input coupled to the output of the memory, said matrix module including circuitry that multiplies data presented by said output of the memory by a matrix of plural coefficients, the matrix module including an output that presents output data based on a result of said multiplication;

wherein the sensing element, the memory and the matrix module are affixed to a common substrate; and

wherein the matrix of coefficients comprises plural rows of coefficients, and the result of the multiplication is a vector of plural element values, wherein the matrix module is further adapted to produce output data indicating which element of said result vector is the largest.

10. The sensing system of claim 9 that includes additional circuitry for varying said coefficients.

11. The sensing system of claim 9 in which the sensing element comprises a 3D accelerometer, a temperature sensor, a barometric pressure sensor, or a humidity sensor.

12. The sensing system of claim 9 in which:

the memory comprises a shift register having an input coupled to the output of the sensing element, the shift register presenting a temporal vector of at least four sensor values on at least four outputs; and

the matrix module has inputs coupled to the at least four outputs of the shift register, and being adapted to perform a dot product operation between the temporal vector of sensor values and each of plural reference vectors that comprise said matrix of coefficients, the matrix module further being adapted to provide, on a circuit output, an integer datum indicating one of said reference vectors that yielded a largest dot product result.

13. The sensing system of claim 12 in which the plural reference vectors have an order, and the matrix module is adapted to output, on said circuit output, an integer datum indicating an ordinal position of the reference vector yielding the largest dot product, within said order.

14. The sensing system of claim 12 in which at least certain of the plural reference vectors are fixed, and not able to be altered.

15. The sensing system of claim 14 in which at least certain of the plural reference vectors are stored in a memory that allows alteration of said vectors.

16. The sensing system of claim 9 in which the sensing element, the memory, and the matrix module are affixed to a common semiconductor substrate, and wherein said memory is adapted to store an array of four or more data representing a physical quantity sensed by the sensing element at four or more respective previous times.

17. A sensing system including:

a sensing element having an output;

a memory having an input and an output, the input being coupled to the output of the sensing element, the memory being adapted to store an array of plural data representing plural physical quantities sensed by the sensing element; and

a matrix module having an input coupled to the output of the memory, said matrix module including circuitry that multiplies data presented by said output of the memory by a matrix of plural coefficients, the matrix module including an output that presents output data based on a result of said multiplication;

wherein the sensing element, the memory and the matrix module are affixed to a common substrate; and

wherein the sensing element comprises a 3D accelerometer, a temperature sensor, a barometric pressure sensor, or a humidity sensor.

18. The sensing system of claim 17 in which the matrix of coefficients comprises plural rows of coefficients, and the result of the multiplication is a vector of plural element values, wherein the matrix module is further adapted to produce output data indicating which element of said result vector is the largest.

19. The sensing system of claim 17 that includes additional circuitry for varying said coefficients.

20. The sensing system of claim 17 in which:

the memory comprises a shift register having an input coupled to the output of the sensing element, the shift register presenting a temporal vector of at least four sensor values on at least four outputs; and

the matrix module has inputs coupled to the at least four outputs of the shift register, and being adapted to perform a dot product operation between the temporal vector of sensor values and each of plural reference vectors that comprise said matrix of coefficients, the matrix module further being adapted to provide, on a circuit output, an integer datum indicating one of said reference vectors that yielded a largest dot product result.

21. The sensing system of claim 17 in which the plural reference vectors have an order, and the matrix module is adapted to output, on said circuit output, an integer datum indicating an ordinal position of the reference vector yielding the largest dot product, within said order.

22. The sensing system of claim 17 in which at least certain of the plural reference vectors are fixed, and not able to be altered.

23. The sensing system of claim 22 in which at least certain of the plural reference vectors are stored in a memory that allows alteration of said vectors.

24. The sensing system of claim 17 in which the sensing element, the memory, and the matrix module are affixed to a common semiconductor substrate, and wherein said memory is adapted to store an array of four or more data representing a physical quantity sensed by the sensing element at four or more respective previous times.