MindStream Monitor Update & Future Plans

robbyg · Feb 4, 2020

Silver14SS said:
What component in the process of automatic dosing correction concerns you? Is the concern inaccurate test results, inaccurate dosed amount, or bad logic in determining the amount to dose?

My main concern is inaccurate test results either by the machine or user error.
eg the guy who left the calibration fluid attached to his Trident and could not figure out why his Alk was not rising when he dosed. He at least figured it out after a couple of days. The machine would keep on going for maybe weeks. Also no machine is going to know that something is not right with your tank by just looking at it. I can and I can use that info to find out what the problem is. The list goes on and the solution is simple, take one minute out of your day to just check and correct.

Silver14SS · Feb 4, 2020

Inaccurate test results would result in inaccurate dosing whether you dump it in by hand, update your dosing programming, or have the tester updating the dosing programming. If you don't trust your measurement device, seems like it would not be worth continuing to use that device. If manually testing once per day and making the correction then, why bother with a MS, Trident, etc.?

The only time I manually Alk test is when I replace the Alkatronic reagent and recalibrate everything to make sure I didn't mess anything up. Even then I'm trusting that I didn't have a user error in completing the Salifert test and that it's correct.

robbyg · Feb 4, 2020

Silver14SS said:
Inaccurate test results would result in inaccurate dosing whether you dump it in by hand, update your dosing programming, or have the tester updating the dosing programming. If you don't trust your measurement device, seems like it would not be worth continuing to use that device. If manually testing once per day and making the correction then, why bother with a MS, Trident, etc.?

The only time I manually Alk test is when I replace the Alkatronic reagent and recalibrate everything to make sure I didn't mess anything up. Even then I'm trusting that I didn't have a user error in completing the Salifert test and that it's correct.

If my ALK is at 8.5 today and the machine tests it tomorrow and tells me its 7.5 then I am not just going to up my dosage to bring it back up like a machine would do. I would run that test again and if still says 7.5 I would break out a Hanna test kit and check it with that.
If it still read 7.5 with the Hanna then I would dose, but if I came back 8 hours later and it was still at only 7.6 I would start to take a much broader look at what could be going on.

None of that would happen with Auto testing and Dosing. Its just getting a number and assume that so long as it is not way out on left field it should run with that number and dose accordingly. No double checking and no idea that the problem is slowly getting worst.

People can do whatever they want, but even for someone using a tank that is probably rated in the top 5% of the highly automated tanks, I have my limits on what I will rely on a machine for.

Silver14SS · Feb 4, 2020

robbyg said:
If my ALK is at 8.5 today and the machine tests it tomorrow and tells me its 7.5 then I am not just going to up my dosage to bring it back up like a machine would do. I would run that test again and if still says 7.5 I would break out a Hanna test kit and check it with that.
If it still read 7.5 with the Hanna then I would dose, but if I came back 8 hours later and it was still at only 7.6 I would start to take a much broader look at what could be going on.

None of that would happen with Auto testing and Dosing. Its just getting a number and assume that so long as it is not way out on left field it should run with that number and dose accordingly. No double checking and no idea that the problem is slowly getting worst. People can do whatever they want, but even for someone using a tank that is probably rated in the top 5% of the highly automated tanks, I have my limits on what I will rely on a machine for.

I think most automated testers repeat the test at least once, hopefully more, when there is a large variance like the 1dKH in your example. That should prevent changes to a dosing schedule resulting from one errant test. I agree that continued, large, unexplained variances like that require human intervention.

The only device I recall seeing a reoccurring drift to test results was MS. Tying dosing to an Alk reading that continued to get less accurate day by day does sound like a recipe for disaster, but I don't recall seeing this issue in Trident, KHG, KHD, etc. Any issues I've seen with those devices have been obvious, "hey something isn't working" type issues.

To each their own, but if I was doing manual daily testing, an automated tester would have no value to me.

robbyg · Feb 4, 2020

Silver14SS said:
To each their own, but if I was doing manual daily testing, an automated tester would have no value to me.

I am not talking about manual testing for daily results. I am only talking about the Auto tester changing my dosing amounts automatically. That is what I do not trust!
I want to have my water auto tested and then I will interpret the results and adjust my dosing levels myself.

TheHarold · Feb 4, 2020

robbyg said:
I am not talking about manual testing for daily results. I am only talking about the Auto tester changing my dosing amounts automatically. That is what I do not trust!
I want to have my water auto tested and then I will interpret the results and adjust my dosing levels myself.

Mine adjusts dosing up to 10% from my programmed value. If that is not enough, it will error out and send notification.

The fallback value it will go to in case of unusual readings is always a "safe" amount.