Previous Blog: Oct 23 - More PDP | Next Blog: Nov 6 - Electrical Design Review

This week, we started specing key parts of our design, such as a battery and Peltier module. We also looked analytically into heat dissipation and small form factor.

We also thought long and hard about what we wanted to do for our control algorithm - if we chose a design that is versatile like a patch, how can our sensors detect when a hot flash is happening in the body if we can’t predict what part of the body the user choses to stick the patch onto?

That would mean we can’t really collect aggregated temperature data for a particular part of the body (vs. something like the Embr Wave is only wrist-mounted, so it eliminates a whole host of unknown variables).

We brought these concerns to the teaching team and they advised us to refine our controls scope.

We were also asked to prepare calculations behind some of our engineering decisions for next time.

In the end, for the controls scope, we decided to use AI to predict the probability of a hot flash using sensor fusion. Ultimately, sensor input would be put into a weighted matrix algorithm and trained over time under supervision of the user.

(See Controls & Software (4A))