It’s good that you built it, but it seems to me that now you have a prototype, before you start investing in a patent or business to sell scaled-up versions, it’d make more sense to invest $100 in a CO2 air sensor and a Raspberry Pi with a switch to randomly turn it on/off: to verify that it decreases CO2 as much and as long as expected, whether you can tell when CO2 levels have been lowered, and whether this has any measurable behavioral effects. The value of such information is very high: there is no point in scaling up a design which isn’t working at its basic task of lowering CO2 levels, and commercialization will be difficult if it does nothing observable (especially given our questions about how and whether CO2 does anything) and, perhaps more importantly from a marketing perspective, if the user can’t feel it doing something.
I have the relevant air sensor, it’d be really hard to blind it because it makes noise, and the behavioral effects thing is a good idea, thank you.
It’s not currently with me.
I think the next thing to do is build the 2.0 design, because it should perform better and will also be present with me, then test the empirical CO2 reduction and behavioral effects (although, again, blinding will be difficult), and reevaluate at that point.
I have the relevant air sensor, it’d be really hard to blind it because it makes noise, and the behavioral effects thing is a good idea, thank you.
Just randomizing would be useful; obviously, your air sensor doesn’t care in the least if it is ‘blinded’ or not. And if it’s placed in a room you don’t go into, that may be enough. As well, maybe you can modify it to have a flap or door or obstruction which opens or closes, greatly changing the rate of CO2 absorption, and randomize that; or if you have someone willing to help, they can come in every n time units to replace the filler or not, giving you both blinded & randomized comparisons between high-CO2-removal vs low-CO2-removal conditions based on whether they pulled out the used filler or not, since the fan presumably still makes the same noise regardless of whether it has brand-new filler removing CO2 at maximum rates or expired tired filler removing only a little CO2. (Remember, experiments work fine comparing 100% removal rates to, say, 10% removal rates; it doesn’t have to be exactly ‘on’/‘off’, that’s just a bit more statistically-efficient because it has a slightly larger effect size, and you have to remember the estimate is a bit lower than the ‘true’ estimate because the ‘off’ condition has 10% of the benefits of the ‘on’.)
Verifying that the thing scrubs CO2 at the expected rate is definitely a good idea. Verifying the behavioural effects is much harder—you’d need to avoid unblinding, and ideally have several different people with varying levels of age, fitness etc, and then you’d get affected by weather, unless your house is very well sealed...
How portable can this scrubber be? If you’re somewhere cold and not getting enough air at night and it’s your house, you could install a heat recovery ventilator. There is evidently a big market for portable air conditioners, despite their inefficiency; the description of this thing (water, air, pumps out sludge) sounds a lot like a washing machine.
It’s good that you built it, but it seems to me that now you have a prototype, before you start investing in a patent or business to sell scaled-up versions, it’d make more sense to invest $100 in a CO2 air sensor and a Raspberry Pi with a switch to randomly turn it on/off: to verify that it decreases CO2 as much and as long as expected, whether you can tell when CO2 levels have been lowered, and whether this has any measurable behavioral effects. The value of such information is very high: there is no point in scaling up a design which isn’t working at its basic task of lowering CO2 levels, and commercialization will be difficult if it does nothing observable (especially given our questions about how and whether CO2 does anything) and, perhaps more importantly from a marketing perspective, if the user can’t feel it doing something.
I have the relevant air sensor, it’d be really hard to blind it because it makes noise, and the behavioral effects thing is a good idea, thank you.
It’s not currently with me.
I think the next thing to do is build the 2.0 design, because it should perform better and will also be present with me, then test the empirical CO2 reduction and behavioral effects (although, again, blinding will be difficult), and reevaluate at that point.
Just randomizing would be useful; obviously, your air sensor doesn’t care in the least if it is ‘blinded’ or not. And if it’s placed in a room you don’t go into, that may be enough. As well, maybe you can modify it to have a flap or door or obstruction which opens or closes, greatly changing the rate of CO2 absorption, and randomize that; or if you have someone willing to help, they can come in every n time units to replace the filler or not, giving you both blinded & randomized comparisons between high-CO2-removal vs low-CO2-removal conditions based on whether they pulled out the used filler or not, since the fan presumably still makes the same noise regardless of whether it has brand-new filler removing CO2 at maximum rates or expired tired filler removing only a little CO2. (Remember, experiments work fine comparing 100% removal rates to, say, 10% removal rates; it doesn’t have to be exactly ‘on’/‘off’, that’s just a bit more statistically-efficient because it has a slightly larger effect size, and you have to remember the estimate is a bit lower than the ‘true’ estimate because the ‘off’ condition has 10% of the benefits of the ‘on’.)
Verifying that the thing scrubs CO2 at the expected rate is definitely a good idea. Verifying the behavioural effects is much harder—you’d need to avoid unblinding, and ideally have several different people with varying levels of age, fitness etc, and then you’d get affected by weather, unless your house is very well sealed...
How portable can this scrubber be? If you’re somewhere cold and not getting enough air at night and it’s your house, you could install a heat recovery ventilator. There is evidently a big market for portable air conditioners, despite their inefficiency; the description of this thing (water, air, pumps out sludge) sounds a lot like a washing machine.
Not really. There’s scads of behavioral measures you can collect passively.
No you don’t, and blinding is easy if you think about it for a few seconds, see the comment I left well before yours.
No, you don’t, you are letting perfect be the enemy of better
This is a feature, not a bug.