Skip to main content

iTracking

Progress Report #3 (11/15)

Goals:

  1. Figure out how to communicate data between 2 Google AIY kits.
  2. Finalize lens purchase
  3. Test convolutional neural nets with tensor flow.
  4. Figure out a signal processing algorithm to achieve clear signal (long-term goal that necessitates a prototype).

Progress:

This week we finalized part orders for a stepper motor system. We debated stepper and servo motors for our design as they are similar in concept, but with a few fundamental differences. Stepper motors are excellent at lower speeds (less than 2000 rpm) which we will certainly be operating under. Also, stepper motors are able to keep place of position (absolute position instead of only displacement) without an encoder which limits costs and extra parts to buy. The draw back of our decision is that stepper motors can run hot due to continuous current; we do not foresee problems with this as we are looking at small run times. However, this should be taken into consideration when designing the overall product.

To be done:

We will need to connect our stepper motors to the Raspberry Pi board and test their turning capability, especially when carrying another Google AIY kit. In addition to the motors, we will need to create a CAD design for a box to hold the camera and connect to the motors. It is also important to pick an appropriate lens to better focus and magnify specifically the eyes. We need to research and pick from several lens modification options: 1) designing a zoom lens (several lenses placed in series), 2) replacing the lens of the Raspberry Pi Camera v2 with a more appropriate one, or 3) swapping out the camera model entirely. We will make this decision based on cost and feasibility.