Technical Perspective:

• Provides a wearable device for data collection for determining activities of a user.
• provides a technique for determining complex activities & improves accuracy in determination of said activities.
• Further reduces position-based dependency for determining activities of the user.

Prototype:

• The present invention is worn on the upper portion of the body to the waist, & it amplifies and senses activities on the upper portion of the body.

Sensors:

• Said device uses a single type of sensor i.e. strain sensor to determine different activities.

Activity of the user:

• Present invention further uses a combination of smart suspender and machine learning algorithm to realize the activities listed hereinbelow:
• There are 14 different normal day-to-day activities such as standing, exercising (arms front stretch, arms lateral stretch and arms lifted), walking, sitting, eating, drinking, washing hands, on phone, texting over phone, coughing and deep breathing.
• Said features related to the movements of the user is one or more of mean, standard deviation, mode, maximum, minimum, signal magnitude area, autoregression coefficients, correlation coefficient, energy, signal average energy, interquartile range, entropy, skewness, and kurtosis.

Machine Learning Model:

• One or more machine learning models is K-Nearest Neighbours (KNN), Long Short-Term Memory (LSTM), Random Forest (RF), Support Vector Classifier (SVC), Support Vector Machine (SVM), Logistic Regression (LR), Neural Networks (NNs), Decision Tree (DT), and Gradient Boosted Decision Tree (GBDT).

Industrial Perspective:

• By using wearable device, present invention is useful in a wide variety of applications, such as health monitoring in secure manner & cost-effective manner.