The comparing between AW and MAX30102 is finished, intuitively, the larger window we use to average the MAX30102 signal, the smoother it get and turns out to have an acceptable accuracy.
Email alert is primarily finished.
With all that being said, the variance of MAX30102 measurement is still fluctuating too much with different groups. This shouldn't be the problem with frequency. We will keep on digging...
Email alert sends too many alert per measurements as we set the criteria to be "send an email when spo2 < threshold (90)".
Address two problems above. If time is enough, use the LCD to visualize the data (this shouldn't take long as there are a few tutorial online.