System and method for adaptive fusion of data from multiple sensors using context-switching algorithm

1. A method for adaptive fusion of sensor data from multiple sensors, comprising:

collecting the sensor data generated by a plurality of sensors in an environment;

determining a quality metric of the sensor data generated by each sensor of the plurality of sensors, the quality metric corresponding to a suitability of the sensor data for performing a given task;

selecting, via an assessment artificial intelligence program, one or more sensors from the plurality of sensors based on the determined quality metric of the sensor data that yields a desired accuracy for performing the given task; and

selecting for said each sensor of the plurality of sensors selected, via the assessment artificial intelligence program, a machine learning algorithm from a predetermined set of machine learning algorithms based on the determined quality metric of the sensor data that yields the desired accuracy for performing the given task;

receiving a classification output from each selected machine learning algorithm for each selected sensor, wherein the classification output is a result generated by performing the task; and

combining classification outputs using a decision-level fusion rule; wherein the decision-level fusion rule defines how the classification outputs are aggregated to arrive at a final classification to accomplish the given task;

wherein the predetermined set of machine learning algorithms comprises preselected machine algorithms, where the preselected machine learning algorithms are selected by evaluating their performance in performing the given task using a benchmark data set.

2. The method of claim 1 , wherein the plurality of sensors include at least one of a PIR sensor, a thermopile sensor, a microwave sensor, an image sensor, and a sound sensor.

3. The method of claim 1 , wherein based on the combination of classification outputs from the selected machine learning algorithm for each selected sensor, a characteristic of a light emitted by one or more tunable luminaires arranged to illuminate the environment may be adjusted.

4. A computer program product for adaptive fusion of sensor data from multiple sensors having a plurality of non-transitory computer readable instructions, the plurality of non-transitory computer readable instructions arranged to be stored and executed on a memory and a processor, wherein the plurality of non-transitory computer readable instructions are operative to cause the processor to:

collect the sensor data generated by a plurality of sensors in an environment;

determine a quality metric of the sensor data, using a data quality module, generated by each sensor of the plurality of sensors, the quality metric corresponding to a suitability of the sensor data for accomplishing a given task;

select, via an assessment artificial intelligence program, one or more sensors from the plurality of sensors based on the determined quality metric of the sensor data that yields a desired accuracy of accomplishing the given task; and select for said each sensor of the plurality of sensors selected, via the assessment artificial intelligence program, a machine learning algorithm from a predetermined set of machine learning algorithms, based on the determined quality metric of the sensor data, that yields the desired accuracy for accomplishing the given task;

receive a classification output from each selected machine learning algorithm for each selected sensor, wherein the classification output is a result generated by accomplishing the task; and

combine classification outputs using a decision-level fusion rule; wherein the decision-level fusion rule defines how the classification outputs are aggregated to arrive at a final classification to accomplish the given task;

wherein the predetermined set of machine learning algorithms comprises preselected machine learning algorithms, wherein the preselected machine learning algorithms are selected by evaluating their performance accomplishing the given task using a benchmark data set.

5. The computer program product of claim 4 , wherein based on the combination of classification outputs from the selected machine learning algorithm for each selected sensor, a characteristic of a light emitted by one or more tunable luminaires arranged to illuminate the environment may be adjusted.

6. The computer program product of claim 4 , wherein the assessment artificial intelligence program is selected from an artificial neural network or a support vector machine.

7. A system for adaptive fusion of sensor data from multiple sensors, the system comprising:

one or more tunable luminaires arranged to illuminate an environment;

a controller having a processor, the processor arranged to:

collect the sensor data generated by a plurality of sensors in the environment;

determine a quality metric of the sensor data generated by each sensor of the plurality of sensors, the quality metric corresponding to a suitability of the sensor data for executing a given task;

select, via an assessment artificial intelligence program, one or more sensors from the plurality of sensors based on the determined quality metric of the sensor data that yields a desired accuracy for executing the given task; and

select for said each sensor of the plurality of sensors selected, via the assessment artificial intelligence program, a machine learning algorithm from a predetermined set of machine learning algorithms based on the determined quality metric of the sensor data that yields the desired accuracy for executing the given task;

receive a classification output from each selected machine learning algorithm for each selected sensor, wherein the classification output is a result generated by executing the task; and

combine classification outputs using a decision-level fusion rule;

wherein the decision-level fusion rule defines how the classification outputs are aggregated to arrive at a final classification to accomplish the given task;

wherein the predetermined set of machine learning algorithms comprises preselected machine learning algorithms, wherein the preselected machine learning algorithms are selected by evaluating their performance for executing the given task using a benchmark data set.

8. The system of claim 7 , wherein the plurality of sensors include at least one of a PIR sensor, a thermopile sensor, a microwave sensor, an image sensor, and a sound sensor.

9. The system of claim 7 , wherein the assessment artificial intelligence program is selected from a support vector machine and an artificial neural network.