Apparatus and method for implementing a mobility aid device

1. A non-transitory computer readable storage medium comprising executable instructions to:

receive sensor data measurements associated with usage of a mobility aid device from a plurality of sensors;

extract a plurality of sensor feature parameters from the sensor data measurements; and

based on the sensor feature parameters, classify the usage of the mobility aid device into a usage state classification, wherein the usage state classification is selected from proper use, improper use, and not used;

wherein the sensor data measurements include angular rate data measurements, force data measurements and linear acceleration data measurements; and

wherein the executable instructions to classify the usage of the mobility aid device include executable instructions to calculate a plurality of acceleration vectors from the angular rate data measurements, force data measurements and linear acceleration data measurements.

2. The non-transitory computer readable storage medium of claim 1 , further comprising executable instructions to provide real-time feedback to a user of the mobility aid device, wherein the real-time feedback includes visual feedback configured for a graphical user interface of a display screen, and wherein the visual feedback provides a graphical representation of the sensor data measurements.

3. The non-transitory computer readable storage medium of claim 2 , wherein the executable instructions to provide the real-time feedback include executable instructions to generate an audio tone indicative of the usage state.

4. The non-transitory computer readable storage medium of claim 1 , wherein the sensor feature parameters include at least one of a fundamental frequency amplitude parameter and a spectrum energy parameter, extracted based on applying a Fast Fourier Transform to the sensor data measurements.

5. The non-transitory computer readable storage medium of claim 4 , wherein the sensor feature parameters further include at least one of a ratio of the fundamental frequency amplitude parameter to the spectrum energy parameter, a mean of a time-series of the sensor data measurements, and a maximum slope of a time-series of the sensor data measurements.

6. The non-transitory computer readable storage medium of claim 1 , wherein the executable instructions to classify the usage of the mobility aid device include executable instructions to create sensor feature vectors.

7. The non-transitory computer readable storage medium of claim 1 , wherein the executable instructions to classify the usage of the mobility aid device include executable instructions to utilize a Naïve Bayes classifier model.

8. The non-transitory computer readable storage medium of claim 1 , wherein the executable instructions to classify the usage of the mobility aid device include executable instructions to utilize a decision tree-based classification technique.

9. The non-transitory computer readable storage medium of claim 1 , wherein the executable instructions to extract the sensor feature parameters include executable instructions to extract a swing characteristic from the angular rate data measurements.

10. The non-transitory computer readable storage medium of claim 1 , wherein the executable instructions to extract the sensor feature parameters include executable instructions to extract a user's change in direction from the angular rate data measurements.

11. The non-transitory computer readable storage medium of claim 1 , wherein the executable instructions to classify the usage include executable instructions to utilize a learning technique.

12. A non-transitory computer readable storage medium, comprising executable instructions to:

receive sensor data measurements associated with usage of a mobility aid device from a plurality of sensors;

extract a plurality of sensor feature parameters from the sensor data measurements; and

based on the sensor feature parameters, classify the usage of the mobility aid device into a usage state classification, wherein the usage state classification is selected from proper use, improper use, and not used;

wherein the executable instructions to classify the usage of the mobility aid device include executable instructions to determine a subset of the plurality of sensors to be used in the mobility aid device, based on an optimal combination of sensors that provides accurate usage state identification and reduced overall energy consumption; and

wherein the subset of the plurality of sensors is derived based on an energy-aware cost function applied to the plurality of sensors, wherein the cost function includes a decision matrix related to a plurality of alternate configurations, and a calculation of an energy-aware benefit for each of the plurality of alternate configurations.

13. The mobility aid device of claim 12 , wherein the plurality of sensors comprise at least one of a three-axis accelerometer, a single-axis gyroscope and a pressure sensor.

14. The mobility aid device of claim 13 , wherein the plurality of sensors are configured to detect at least one of a downward force, a linear acceleration and an angular rate associated with the usage of the mobility aid device.

15. A mobility aid device comprising:

a plurality of sensors including a three-axis accelerometer, at least one single-axis gyroscope and at least one pressure sensor; and

a computational unit coupled to the plurality of sensors, wherein the computational unit is configured to:

process outputs of the plurality of sensors and detect an applied force, a linear acceleration and an angular rate associated with usage of the mobility aid device;

classify usage of the mobility aid device into a usage state, wherein the usage state is selected from a first group, and members of the first group include proper use, improper use, and not used; and

provide feedback regarding the usage state of the mobility aid device;

wherein the computational unit is further configured to receive data from a wearable sensor, and wherein the data from the wearable sensor allows the computational unit to select the usage state from a second group that is larger than and including the first group.

16. The mobility aid device of claim 15 , wherein the computational unit is further configured to generate an audio tone indicative of the usage state.

17. The mobility aid device of claim 15 , wherein the computational unit comprises an audio-visual interface module configured to generate an audio output and a visual output of the usage state associated with the mobility aid device, wherein the visual output includes a graphical representation of the sensor data measurements.

18. The mobility aid device of claim 15 , wherein the computational unit comprises a sensor processing and inference module configured to extract the sensor feature parameters from the sensor data measurements and derive the usage state corresponding to the usage of the mobility aid device, based on the sensor feature parameters.

19. The mobility aid device of claim 18 , wherein the sensor processing and inference module is configured to perform a plurality of feedback cycles associated with the usage of the mobility aid device until proper usage is detected, wherein each of the plurality of feedback cycles includes:

providing a present usage state to a user of the mobility aid device;

extracting sensor feature parameters from sensor data measurements acquired by the selected subset of the plurality of sensors; and

deriving an updated usage state based on the sensor feature parameters.

20. The mobility aid device of claim 15 , wherein the computational unit is further configured to deactivate a subset of the plurality of sensors.

21. The mobility aid device of claim 15 , wherein the computational unit is configured to utilize a Naïve Bayes classifier model to process the outputs of the plurality of sensors.