from luxury devices to consumer sensors

This article was supposed to be about our study of the data between the Nebo sensor and reference air monitoring sensors, carried out with the help of Dr. Achim Dittler. However, the professor with whom we conducted the study was currently busy with scientific work and was unable to agree on the conclusions of the obtained data. And while he is busy with scientific work, my introductory text has gone beyond the limits of one sheet. Then let it be a separate article.

How air quality information is collected

These two devices you see in the photo, one of which looks like a space container, are often called a reference-class monitor. The smaller one is a BAM. Reference-class means that they are of the highest scientific quality, a kind of gold standard. These instruments are very expensive. And speaking of price, they can cost anywhere from tens of thousands to hundreds of thousands of dollars, and that's just for the purchase. Add tens of thousands for maintenance. So these devices are used in developed countries where the government “can afford” them, or in less developed countries with grants from developed foundations, which is also good. Sometimes the data from such stations is not published, so as not to spoil the statistics or “not to scare the population.”

Can everything be seen from satellites? Well, no.

And so you type in “Air quality Murmansk” in the search and get a link to the first weather services. Very convenient – real-time data.

Only next to the value number there is an asterisk, which hides the mysterious inscription “Modeling using satellite data”. In areas with less coverage, meteorological services use this clever method to provide data to users and increase the coverage of territories. You can often see interpolated surfaces reflecting air quality. Less costs – calculated on the server and issued data for hundreds of thousands of users. But I would treat such sources with a degree of skepticism and have already written about this before. short article. That is why our entire team is in favor of increasing the number of monitoring stations: the more sensors on the ground, the more accurate the modeling and forecasting.

What if there is no money?

Due to the lack of data from the ground, devices that can be called low-cost sensors or in scientific circles LCS (low-cost sensors).

It is a kind of consumer grade product, which means that you or I can buy such a sensor for our own use using funds from the family budget. The main thing is to explain to your wife why you need this toy. And as you can see in the photo, they are noticeably smaller.

Cheaper means better? So what's the problem?

In the early stages of development, inexpensive sensors were not very accurate, which caused criticism from the scientific community. It is worth saying that this was not unfounded. At the beginning of its appearance, LCS equipment had a large error due to the peculiarities of optical measurement technology, which is affected by factors such as humidity and negative temperatures. Laser technology is limited by the temperature range: from +10 to +40 degrees.

About technology

Inside the sensor is a laser that emits a narrow beam of light. Air is drawn into the sensor through a fan or pump and passes through the laser beam.

Dust particles scatter light, which is captured by a photodetector. Electronics analyze the data and translate it into the concentration of PM2.5 particles in the air.

And the sensors began to be improved

Idea one: to adjust the data relative to the reference class.

Swiss-German company IQAir was one of the first to use this method, installing a sensor next to BAM. They calibrated their sensor data against reference ones. Airly has gone one step further and calibrates data using AI against nearby “reference stations.” PurpleAir has inserted two optical sensors for reliability and also uses calibration.

Idea number two: use the first option and…

The German company Palas combined mathematical adjustments with the elimination of technical limitations of optical technology. They used a dehumidifier and heated air in their Fidas 200 sensors, which improved accuracy. However, such devices cost about $20,000.

We wanted cheaper, more compact sensors and to use both methods. That's how our window sensor Nebo came into being: we built air heating into it and, not without fear, decided to test its accuracy. We did this in Germany and Russia.

Just for fun: in one of our project presentations, we calculated that instead of installing just one “reference” station, we could install over 100 inexpensive stations with good data quality in a city. This approach would provide broad coverage and exceed the data quality of a single reference station.

Well, that's the end of the introduction.

That's it? No.

We will soon receive the data from the professor and publish it. It will even be interesting. I thank Professor Achim Dittler for fruitful discussions and contributions.

Is that all? Just a little more.

Our team loves data and hardware, but we avoid marketing. For me, it's a real pain. I've spent the last month collecting an endless list of “green” funds, writing applications and preparing pitches – it's long, painful and simply killing. If there are enthusiasts among you who love promoting products and finding funding, we'd be happy to meet you! We need a co-founder to complement our startup.

We will be glad to see you in our team!