The arduino will on the insulin pump when the glucose level rise high and then and then when the levels fall to normal it will cut off the pump. This is like the many chemistry experiments to neutralise reactions.
However, I don't want to use vinegar and soda so plan on using glucose and insulin.
There is a problem to using this because I need a method for measuring glucose continuously. Does anyone have an idea of how this could be done in a cost efficient way.
Also how does the project sound itself? I plan on entering this into the Google science fair.
The only way I know to measure glucose continuously is with an electrochemical cell. But with that there would be many other compounds that might also register. What you're asking for is not simple and definitely won't come cheap.
Skdr786:
The arduino will on the insulin pump when the glucose level rise high and then and then when the levels fall to normal it will cut off the pump. This is like the many chemistry experiments to neutralise reactions.
However, I don't want to use vinegar and soda so plan on using glucose and insulin.
There is a problem to using this because I need a method for measuring glucose continuously. Does anyone have an idea of how this could be done in a cost efficient way.
Also how does the project sound itself? I plan on entering this into the Google science fair.
My wife is a diabetic and she has an insulin pump. Currently the blood glucose measurement is done using a separate meter using blood taken from her fingers. The glucose measurement is sent automatically to the insulin pump. There is an better pump that continuously measures blood glucose using a sensor inserted just below the skin but it is not generally available.
At the moment the whole insulin delivery process cannot be completely automated since the amount of insulin to dispense depends not just on the current blood glucose level but also on the carbohydrate ingested and this has to be estimated by the diabetic and entered into the pump.
The insulin pumps are very very expensive with much of the cost coming because the pump is considered safety critical and requires major testing by the authorities (of course at the developer's expense). You will incur the same development costs.
The skill set required to design the whole system is probably beyond that available by a single person and the whole project will take years. If you are on your own then by the time you get even a basic system designed, prototyped, quality tested and into production the chances are that you will be years behind the times.
I'm not saying that you should not have a go at the project but think in terms of years and not months.
I do have a shield that does this...for a commercial bioreactor.
and I do have a sketch to act as a master/server bridge on Modbus TCP to a YSI2700 select with an autosampler.
Two of these boards with Mega2560 run the whole reactor environmental controls and feeding and harvest and sterilization..I do slave these via Modbus TCP as well.
I use a YSI2700 select. This analyzer is based on glucose oxidase membrane technology.
A State engine in the Mega that runs the YSI sends commands to that "robot"
I2c slave to my master Megas runs the YSI..
I2c slaves for the stepper controllers do the feeding and Harvest.
for bigger fermenters I can talk RS485 to the motor controllers of VFDS.
If I were of a mind to it, I would rip the guts out of the YSI and run that machine directly with an Arduino...a very commendable project and if done right you could sell quite a number of them at about $3000
Bottom line GLucose Oxidase with a polarographic amperemetric sensor is the method.
