Hacker News Reader

Comment by boplicity | Hacker News Reader

Change it once a week -- $52 a year seems cheap compared to the alternative of children ingesting lead.

Regularly replacing filters is how filters always work.

That's $52/year per faucet. I've got four faucets in my home that a child could reasonably drink from. That's $200/year and 48 filter changes assuming you don't miss any changes. And that doesn't help with showers and baths, or even pools and hose water.

Compare that to a whole-home filter: under $500 with a five year warranty. Filters are a couple hundred bucks every 100,000 gallons. The time savings alone is worth it, but the real value is knowing that you're not relying on $1 plastic filters.

Nothing about these $1 filters is good for children. The idea that a household with kids has the time and energy to change filters on every faucet once a month (ignoring all the time your kids spend at school with lead water there!) is laughable.

Do you drink water out of every faucet?

You just need it for the one faucet in the kitchen you'd be drinking water out of.

Would that completely eliminate all your lead consumption? No, when you are brushing I'm sure you swallow some water, etc, but it would eliminate at least 90% of it.

There are at least three faucets in my house that we get drinking water from: kitchen and the two upstairs bathrooms. I'm not walking downstairs to the kitchen to get water in the middle of the night. I'd expect the average single family home is similar.

That's still $150/year and 36 filter changes.

And have you ever met a child? They'll play with and drink almost any source of water. Telling them not to drink from certain faucets is Sisyphean.

Using a whole-house filter misses the potential for lead to come from the pipes in the house. Using an after-the-tap filter eliminates that risk.

A permanently installed filtration system isn't as cost effective as a whole home filter, but it's still more cost effective than what's been designed here, and requires an order of magnitude fewer filter replacements. In fact, you probably wouldn't even need a plumber to install the system for you.

I think the GP is getting at the fact that, as a design, it is impressive enough to get media attention but not robust and comprehensive enough to be an actual real-world product. So much more industrial design would have to go into the concept in order for it to be available in the marketplace.

It reminds me of those news stories that pop up from time to time about some high schooler inventing a robot to clean up plastic waste from the ocean. A neat concept, and laudable, but the design they've produced has a whole litany of flaws that they did not address because they don't have real world marine engineering experience - susceptibility to corrosion, bio-fouling, etc.

Yes. I've built actual solutions for problems. There are things that kids can do in the research realm that's "interesting", but product development is not one of those things. The focus is on finding a thing that "works" (IE, shows at least one true positive), but not a thing "that could be deployed to a million homes for a decade and show a TP rate of 99.9% and a FN rate of 0.01%". Getting a product to the market involves mostly things that don't involve the working of the device, but rather preventing the device from failing in the field.

The other problem is that there are 10 other kids who are doing things better (learning to contribute in the long term) who don't get glowing articles written about them.

you're working from the assumption that this could actually be deployed, and work as expected. It would not.

Is your reasoning just that most things don't work as expected, and this is a thing, so it probably won't work as expected? Or do you know something specific about the chemistry of this particular process?

my reasoning is based on watching large numbers of medical devices and other consumer devices that have been deployed to billions of people. In that sense, yes, I am applying a prior. But also, kids just don't learn the steps required to make reliable field devices. They learn how to set up an experiment and run that with lots of hands-on adjustments. That doesn't work in the field. You have to assume your users are idiots with harsh environments who are actively undermining your device.

All of that is true.

It’s a $1 prototype. As soon as manufacturing, labor, and distribution is involved it’s now a $10 filter.

Reply on news.ycombinator.com