7c8efd2228
- MCP server with stdio transport for local use - Search episodes, transcripts, hosts, and series - 4,511 episodes with metadata and transcripts - Data loader with in-memory JSON storage 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
94 lines
7.1 KiB
Plaintext
94 lines
7.1 KiB
Plaintext
Episode: 1579
|
|
Title: HPR1579: Crowd Sourced Air Quality Monitoring
|
|
Source: https://hub.hackerpublicradio.org/ccdn.php?filename=/eps/hpr1579/hpr1579.mp3
|
|
Transcribed: 2025-10-18 05:14:30
|
|
|
|
---
|
|
|
|
This episode of HBR is brought to you by AnanasThost.com.
|
|
Get 15% discount on all shared hosting with the offer code HBR15.
|
|
That's HBR15.
|
|
Better web hosting that's Aniston Fair at AnanasThost.com.
|
|
I'm a PhD student in the Robotics Institute at Carnegie Mellon University.
|
|
I have a couple of different things that I work on.
|
|
These range from projects over in Uganda and Haiti, also centered around air quality,
|
|
to actually working on these sensors and also working on different sensors or different,
|
|
basically algorithms to take very large amounts of data and process that efficiently.
|
|
Okay, so what's the name of this device?
|
|
This device is called the spec.
|
|
Spec.
|
|
Yes, basically we're trying to get at the idea of sort of small low-cost devices that we can scatter
|
|
around, kind of like you would scatter dust almost, and it sort of breaks the typical way of
|
|
sensing as like having air quality that's only measured in like one or two places around a city,
|
|
to being able to basically put one in every home.
|
|
So you're sort of crowd sourcing air quality?
|
|
Right, right.
|
|
So, okay, I'm going to describe it since it's on the radio and people don't know what I'm
|
|
looking at. It's a small black box with a little screen, and I guess it's color coding
|
|
the different air quality. So what are the different color codes that you get on this thing?
|
|
Well right now, this is sort of a prototype, so we have a scale of values that aren't yet correlated
|
|
to a specific reading. But basically we've taken readings in a few different places, and
|
|
what we have here is the colors basically indicate whether or not there's a high concentration
|
|
of particulate matter or a low concentration of particulate matter. It doesn't necessarily tell
|
|
you whether or not the air is safe to breathe, but particulate matter is a really good measure
|
|
for basically figuring out whether or not other harmful things might be present.
|
|
And I see that there's, I mean like if you look at at the right angle, there's like a graph or
|
|
something on the front that it's taking what's all that data, and where does that data go?
|
|
I'm assuming it could pipe out to your computer in the portal.
|
|
So the graph on the front of the screen is basically the past hour of data.
|
|
What this will let you do is basically see the recent trends in air quality.
|
|
One hour of data is kind of helpful to look at right away, because it can tell you,
|
|
for example, if you cooked dinner, did that significantly increase the
|
|
amount of particulate? Now it does store all of that data internally, and when you plug it into
|
|
your computer via USB cable, we have an uploader that uploads that to basically a central database,
|
|
so that you can review all of the data that you've collected in your home over time.
|
|
So what you said earlier when we were talking that it's pretty much all custom hardware,
|
|
like what are some of the things that go into making something like that?
|
|
So the core of the unit is a small inexpensive dust sensor. The dust sensor itself costs about
|
|
$2, and typically you find them in air purifiers and things like that,
|
|
basically so that they just switch on when the air quality gets really bad.
|
|
Now what we're actually using is a custom microcontroller to take that sensor data,
|
|
to take basically an arm chip and an algorithm that we're in the process of developing,
|
|
and take the fairly low resolution data that we're getting out of that,
|
|
and basically use a combination of statistics and machine learning to estimate the data
|
|
with a higher fidelity that we can get straight out of the sensor.
|
|
Okay, so the whole thing is running some kind of embedded system,
|
|
like an old PSD or an X or something, and we might, is that what you guys program,
|
|
or what is the brain of this thing with?
|
|
I mean, the brain of the thing is embedded, it's all inside,
|
|
but right now it's just basically running some low-level C code.
|
|
It doesn't really need anything fancier than that.
|
|
The whole thing is fairly low power, which helps to basically contribute to the low cost of the
|
|
device. Is it actually powered by USB right now? That's correct, it's just powered by a standard
|
|
USB cable. Okay, so the end goal, I guess, is to get a device like this in a lot of people's phones,
|
|
so that you can get an idea of the air quality. What, to what end, why do people need to know that,
|
|
or is it just for informational purposes, or can it contribute to something larger where maybe
|
|
you monitor an area? Well, there's a lot of reasons why people might or ought to
|
|
care about air quality. These range from, say, wondering why your child has asthma attacks and
|
|
trying to pinpoint the causes of that, to, say, wondering if the power plant or the drilling
|
|
site down the street is basically like blowing particles into your home when the wind blows
|
|
the right way. The trick about air quality is that air is clear unless it's like really bad,
|
|
and there's like smoke, but for the most part, you don't really have a good way of intuitively
|
|
knowing what the air quality is. Right now, the sensors that are available are all really expensive,
|
|
and the ones that basically have publicly available data, more often than not, these are on the
|
|
tops of buildings and such, so it's actually above all of the air that people are actually breathing.
|
|
And because air quality is sort of naturally continuous, what we really need is a lot of monitors,
|
|
not just a few, we need a lot spread out so that we can actually start to see sort of the dynamics
|
|
of the air quality in a specific area. The best place to go right now is CMUCreateLab.org.
|
|
That's the website for the CreateLab, which stands for Community Robotics, Education,
|
|
and Technology Empowerment, and that site will have information about the spec, which is the
|
|
air quality monitor, as well as a bunch of our other projects as well, so. Is it like open
|
|
the source? Can people start hacking on your code if they want to help? So our lab is sort of unique
|
|
and that we have a policy that anything that we work on is open source. Our philosophy right now
|
|
is that we basically want to get this stuff out there so that it helps people, and if there's
|
|
someone else that can come along and help us improve on that, that's all the better. We totally
|
|
welcome that. What's your name again? My name is Mike Taylor. Cool, thanks a lot for talking to me. All right, thank you.
|
|
You've been listening to HECCA Public Radio at HECCA Public Radio.org. We are a community podcast
|
|
network that releases shows every weekday Monday through Friday. Today's show, like all our shows,
|
|
was contributed by an HBR listener like yourself. If you ever thought of recording a podcast,
|
|
then click on our contributing to find out how easy it really is. HECCA Public Radio was founded
|
|
by the digital dog pound and the Infonomicon Computer Club, and it's part of the binary
|
|
revolution at binrev.com. If you have comments on today's show, please email the host directly,
|
|
leave a comment on the website or record a follow-up episode yourself. Unless otherwise status,
|
|
today's show is released on the creative comments, attribution, share a life, 3.0 license.
|