Two levels of implementation are necessary
1. Custom mobile application
- A custom mobile application will be developed which will analyze the map (integrated to google maps) to the destination.
- It should also analyze the driving/commuting pattern via GPS to understand if the person is approaching a requirement to turn to the left/right on the way.
2. Model for training and prediction
- A model is trained on a series of GPS data to determine this requirement of turning left or right, well in advance.
- This model predicts a change in direction requirement (turn), and also if the person is closing in on the destination (approach).
- This will be different from the normal directional indications. Here, the model should predict a change in direction well in advance so that the user is able to know it pre-maturely.
- This model will be deployed on the Arduino Nano33 IoT module.
Once the turn and approach are predicted and determined by the EdgeImpulse model (using map data and current GPS data),
our custom mobile application will be able to determine a specific vibration pattern that is required to alert the person for an upcoming change in direction or an approach to the destination.
The user speed can be acquired from the IMU onboard the mobile or via a dedicated IMU interfaced with the Arduino module.
The trained model also takes into account the current speed and heading of the user (from GPS or IMU data) for predicting the distance and time required for a change in direction
The intensity and pattern of the vibration can be used to convey these data to the person using appropriate patterns.
By leveraging this possibility, appropriate vibration values within the dynamic range of the motors will be implemented by Arduino.
The module will be worn on the wrist of the user.
The person will be initially required to get acquainted with each of the vibration patterns that will denote the direction or approach to his destination.
Thanks to the versatility of our custom mobile application, it will allow the user to customize various vibration patterns or intensity, according to his/her comfort.
This will be particularly useful for commuters who are in a hurry as they won't be required to halt and check for directions in their navigational aids such as mobile phones,
as our wearable device will intimate them of the directions and estimated turns very well in advance through the use of differential vibrational patterns produced by the micro-motor assembly.