Author Options:

Provide 1000 cheap, wireless climate data loggers - Citizen Science Contest Answered

The SciStarter Citizen Science Contest is live! This is your opportunity to help millions of citizen scientists contribute to real scientific discovery. Make their experiences better by coming up with solutions to some real annoyances that hinder their participation. To get you started, here is a specific--and very real--challenge sent to us by project organizers.


Background: Wildlife of Our Homes provides an opportunity for citizen scientists to help researchers study the species that live alongside us everyday - bacteria, fungi, and insects. By using a sampling kit and answering a few questions, volunteers help researchers create an atlas of microbial diversity in homes across the country.

The Problem: Project organizers would love to collect climate data in each of the 1000 homes where volunteers are sampling microbes from 4 common surfaces. Unfortunately, climate sensors are expensive, and more importantly, project organizers don't have an easy way to transfer data from those home sensors (temperature, humidity, etc) to an online database. Currently, they must physically retrieve and download the data.

The Challenge: Find a way to log climate data and wirelessly transmit the data to the project organizers.

Enter now! Contest closes January 21, 2013


Mouser has a nice mini-page for rf-wireless modules. http://www.mouser.com/applications/rf-wireless-technology/
For instance: http://www.mouser.com/catalog/specsheets/Pan1720Brochure.pdf ,could probably handle this with the addition of a humidity sensor.

My only question is what do you mean by wireless, do you mean it has to transfer date across miles and miles, or do you mean it can transer data 30 feet to a computer.

This is an exciting challenge! Though, there are a few trouble spots that really need clarification from the sponsor.

Sample accuracy is one - do a search on DigiKey, Mouser, Farnell/Element 14, etc... for humidity / temperature sensors. You'll see a variety of components with a very wide range in price; $5 - $50 per piece in 1k quantity. Many people are tempted to select the least expensive component to keep price of the solution as low as possible. If a RH reading with an error margin of +/- 10% is acceptable... then cheap parts are fine. Typically, there is a direct correlation between component cost and precision; more precise = more expensive.

So... what is 'good enough' with regard to data accuracy?

Alternately, the device could have a USB interface on the sensor
with an "auto run" file that would tell the computer to upload the data file to your server. You would need to build a website front end to collect the user data to associate with the data file.

If the Detector had a screen that could display a QR code, up to 2.9kB of binary data could be captured ba cell phone picture and texted back to headquarters. The device could be moved to a different location in the house and the process repeated.
Instructions for the use could also be displayed on the screen so the user would know what number to send the text message to and when it was finished collecting its data and ready to generate the QR code.

Hi folks. This is Rob Dunn. In our ideal world we want to sense those variables that are relevant to species living in your house. We know that temp and humidity are potentially important. It seems likely that some other things (O2 or CO2 for example) are also important. Which things are most important depends on the model of how we deploy these. If we put one in each house and we end up with thousands of houses they need to be quite cheap ($20). Another model would be to do a few houses per state (with variation among house types) and have the price be a little more expensive. The second question is how intensively to sample each house. Because we are sampling life in multiple points in each house, it would be interesting to have climate/chemical data from each one of those points. That said, adding sampling points/house adds to cost. Finally, the third question is how the data get to us. The standard model is that people are sent and send back dataloggers that are downloaded by us. That is a pain and can fail in a number of ways. The cooler approach would be to have dataloggers that connect (and here I show my ignorance) to peoples phones or computers in one way or another and every so often send us the data. What does the ideal scenario look like (taking cost into account). I think the ideal scenario is one in which we can get citizens/dataloggers to send us data and where 500 or so people have one datalogger inside and one inside, but in a hundred or so homes we get people to set up one datalogger at each of our sampling points.

Sounds like you should bootstrap off of the Nielsen ratings system that is deployed. Maybe have an environmental sensor module off of that...maybe they already have...

Thanks for all of your questions. We're working with project organizers to get you the information you need. Stay tuned!


5 years ago

It would be really helpful if there were some listed requirements. I can't find any. What variables, what ranges, what accuracy, what per unit cost? Is the contest to design such a sensor system or to donate it? Assuming it isn't donate 1000 sensors, what skill in assembly is ok?

This reads more as a request for quotes than a competition.

What exactly would you expect the sensors to sense about the climate?

Air pressure?
Pollen count?
Wind direction?
Light levels?

And at what price point? And in what location? Obviously the easiest place to measure these things would be next to a router. This problem's already been solved, I don't see how a hobbiest will be able to make it cheaper, value will come with volume. Some of the products by Ciseco Systems will probably do what you want for this.