"I have not failed 10,000 times. I have successfully found 10,000 ways that will not work." 

-Thomas Edison

Honestly I knew many biosensors take decades of research,  I didn't expect mine to be great, but stupidly I thought it might -- like a lightbulb--obviously just pass or fail and It did not, so far calibrations attempted and/or equipment have lots of error....

• Just beginning to realize the depths of uncertainty involved
• I need a controlled testing method because currently I am juggling too many variables 
• I need to fix the calibration probably with more research and data collection (understand the relation to both BG and the Medtronic Guardian 3 data)
• I want to buy better thermistors and rebuild the ear piece
• Contemplating the significance of blood flow/ heart rate and if I should add a sensor for it 
  • (bad data seems to worsen or become more irradic around heart rate changes and higher BGs) 
  • Need more data

Between my own inexperience, inexpensive sensors, irregular blood sugars (lifestyle) I have not made much forward progress but I'm still slowly steadily trying.