WearEXG

• 

  Seamlessly interact with the surroundings.

It is painful to watch many of our known quarters whose life suddenly change due to some accidents or strokes or other brain injury/diseases. Once a robustly independent person become entirely dependent on relatives or healthcare assistant even for a simple daily task. We want to enable disabled or immobile patients with some degree of independence to interact seamlessly with surrounding home devices, and even play games and a better way to interact with other people.

Our device will allow people to generate actionable commands to operate assistive devices and smart-home devices with simple gestures. It will also read physiological parameters, including stress, anxiety, and sleep patterns.

We combined simple EXG biosensor electrodes, bluetooth modules, embedded signal processing system and machine learning tools including Matlab, Python, Github to accomplish this task.

Accurate classification of various physiological signals generated from brain wave EEG, and other bio-signals like EOG, EMG, ECG was challenging. Also recalibration was required for different users.

We are able to control an external robot with just simple eye movements. We were also able to play powerpoint slides with simple eye blinks. We are very close to develop a sleep alert system for drivers while our wearable device can detect early signals of falling asleep.

To combine various hardware, embedded system and machine learning tools and generate a real-world use case tasks using simple technologies. Developing shortcuts and creative ways to classify signals and generate commands. To redesign a robot from legos and other simple materials and circuits boards.

It is important to develop this pre-prototype into a real-world product from a design-thinking approach, and develop miniature wearable devices that are easy, comfortable to wear, are compact and allow people interact with a range of existing and upcoming smart home IoT devices and robots. To test the prototype with several subjects and plan a path for a lean product development based on user feedback.