Predictive power management in a wireless sensor network

• US 9,721,210 B1
• Filed: 11/26/2013
• Issued: 08/01/2017
• Est. Priority Date: 11/26/2013
• Status: Active Grant

• Current Assignee: Invent.ly LLC
• Original Assignee: Invent.ly LLC
• Inventors: Brown, Stephen J.
• Primary Examiner(s): Rampuria, Satish
• Application Number:
  US14/090,099
• Time in Patent Office: 1,344 Days
• Field of Search: 700286-306
• CPC Class Codes: G05F 5/00 Systems for regulating electric variables by detecting deviations in the electric input to the system and thereby controlling a device within the system to obtain a regulated output
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/046 Forward inferencing; Production systems

Abstract

An apparatus comprising a power source, one or more sensors, a transceiver, and a memory. The power source may be configured to store energy to power the apparatus. The one or more sensors may be configured to receive captured data from one of a plurality of sources. The transceiver may be configured to send and receive data to and from a wireless network. The processor may be configured to execute computer readable instructions. The memory may be configured to store a set of instructions executable by the processor. The instructions may be configured to (A) evaluate an expected power usage budget calculated using a predictive model of future energy consumption and (B) (i) store the captured data in the memory in a first mode and (ii) transmit the captured data to a remote storage device in a second mode. The first mode or the second mode is selected based on characteristics of the captured data received from the sensors.

Claims

1. An apparatus comprising:

a power source configured to store energy to power said apparatus;

one or more sensors configured to receive captured data from one of a plurality of sources;

a transceiver configured to send and receive data to and from a wireless network;

a processor configured to execute computer readable instructions; and

a memory configured to store a set of instructions executable by the processor, said instructions, where executed, are configured to (A) evaluate an expected power usage budget calculated using a predictive model of future energy consumption and (B) (i) store said captured data in the memory in a first mode and (ii) transmit the captured data to a remote storage device in a second mode, wherein (a) said first mode or said second mode is selected based on characteristics of the captured data received from the sensors, (b) said instructions include a machine learning process configured to optimize a predictive model and (c) said predictive model is configured as (i) a series of energy input functions, (ii) one or more energy consumption functions with scaling factors that are optimized based on historical data based on prior patterns and (iii) a least squares minimization of a difference between model predictions and actual historical data or various gradient descent techniques.

2. The apparatus according to claim 1, wherein said sensors receive data from an external device.

3. The apparatus according to claim 1, wherein said sensors receive data from forces acting on said apparatus.

4. The apparatus according to claim 1, wherein said instructions are configured to be updated through the wireless network.

5. The apparatus according to claim 1, further comprising:

a remote computing device removed from the apparatus configured to selectively couple to the apparatus and collect stored data and transmit new instructions to the apparatus.

6. The apparatus according to claim 1, wherein the instructions contain a set of threshold conditions for storing sensed data in the first mode.

7. The apparatus according to claim 1, wherein the instructions contain a set of threshold conditions for transmitting said sensed data in the second mode.

8. The apparatus according to claim 1, wherein the instructions contain computational logic for computing a threshold condition based on one or more types of said sensed data and one or more data points over time.

9. The apparatus according to claim 8, wherein said computational logic includes a cascade of threshold conditions that determine a powering on or off of additional features of said sensors.

10. The apparatus according to claim 9, wherein the cascade has a power-aware prioritization of sensor data capture and communications to lower a duty cycle to average power.

11. The apparatus according to claim 8, wherein the computational logic determines said threshold condition for storing and transmitting data based on (i) currently available power, (ii) predicted future energy harvesting opportunities based on forecasting a future environment, and (iii) predicted energy consumption based on applying the threshold condition to the future environment.

12. The apparatus according to claim 1, wherein other factors that are uploaded by the remote computing device to the sensors include a schedule of when sensor data is captured and stored.

13. The apparatus according to claim 1, wherein the instructions are configured to coalesce data by only recording changes when compared to a base state or expected state.

14. The apparatus according to claim 1, wherein said instructions provide a computationally efficient process to only record an average value of the sensors over a period of time and then starts recording events if a measurement is significantly above average.

15. The apparatus according to claim 1, wherein the apparatus is self-powered through said power source.

16. The apparatus according to claim 1, wherein said power source is implemented as a battery.

17. The apparatus according to claim 1, wherein said power source is replenished through energy harvesting of light, vibration, thermal variation, thermal gradient, infrared, chemical, RF sources or some combination.

18. The apparatus according to claim 1, wherein the optimization is performed to maximize a utility within a set of cost parameters, or alternatively, to maximize utility per unit of cost.

19. An apparatus comprising:

a power source configured to store energy to power said apparatus;

one or more sensors configured to receive captured data from one of a plurality of sources;

a transceiver configured to send and receive data to and from a wireless network;

a processor configured to execute computer readable instructions; and

a memory configured to store a set of instructions executable by the processor, said instructions, where executed, are configured to (A) evaluate an expected power usage budget calculated using a predictive model of future energy consumption and (B) (i) store said captured data in the memory in a first mode and (ii) transmit the captured data to a remote storage device in a second mode, wherein (i) said first mode or said second mode is selected based on characteristics of the captured data received from the sensors, (ii) in order to adapt to a duty cycle, sampling rates, adaptive and conditional powering, data storage policies and data transmission policies of said sensors, (a) each energy use by the apparatus is characterized as an energy cost function and (b) each data captured is characterized as a utility or benefit function, (iii) the adaptive and conditional powering of additional sensors is determined based on a set of initial sensor readings and (iv) energy harvesting is a negative cost or energy contribution in the predictive model.