An interesting conversation arose during the Texas Barnraising with a few participants, regarding the idea of developing a simple device to measure the quality of housing in El Salvador where I am based. I am currently part of an effort to improve housing design for social interest housing organizations such as Habitat for Humanity and Techo in El Salvador, by using bio-climatic analyses that take into account things such as the home orientation, green roofs and ventilation.
The plan will be to design a very simple sensor to compare and evaluate three things: how home designs improve the comfort level of families living in them, human perception of comfort, and other environmental variables related to where the home is located and that might be influenced by communities, like the amount of trees, or presence of water.
Characteristics of the device
We convened that the first step on this conversation was to think of what we needed for this project. Some of the ideas that have come up, that might help us limit the characteristics of the solution are:
- The device will be designed for low-cost housing (in El Salvador, this is less than USD 15 000). This is important due to certain limitations that the home should have: less options of where to set the sensors, which means less variability of temperature in different spaces.
- To be used mostly in rural areas with no Internet access in El Salvador. This may rule out the use of IoT solutions.
- To measure at least the minimum amount of usable environmental variables for doing some analysis. On this line, perhaps, is the idea of measuring variables that people might be aware of regularly, like temperature and humidity, because it opens the door to compare perceptions and automatic measurements.
- A very low cost and the possibility of local assembly. I am considering around $15-20 for each kit. A great personal inspiration for this is the Coqui, being currently developed mainly by Jeff
- A non-intrusive design for families who are not used to technological tools such as environmental sensors being put on a wall in their houses, and that uses the least amount of space possible.
My first proposal for this is to develop a simple sensor based off a bare ATmega328 microcontroller, which means a bare Arduino chip. The chip can be programmed in the Arduino, which reduces the cost. Other ideas are:
- Buliding the sensor into a wall clock to turn it into a more friendly accesory around the house. Perhaps some degree of interaction with the user would be pretty interesting as well.
- Data will be recorded in an SD card, so an important part of the development of the sensor will have to do with logging data correctly.
- Using cheap batteries. If done correctly, a low-voltage sensor can simply use AAA batteries.
My biggest challenge with this idea is thinking about interaction with data from the users. A researcher can make use of the data, but what about homeowners? What actions can they choose, based upon the available information? Part of this exploratory study will try to answer these questions, by talking to users and designing procedures alongside them.
The next step after this will be to develop a simple prototype and put it to the test through a couple of organizations in El Salvador which have expressed interest in this.
It is quite easy to build the environmental data logger you describe for $15. All of the parts for a Mini Pearl Logger can be ordered from eBay for $12.00. This device will log temperature, humidity, and barometric pressure every minute or every hour for many months on 3 AAA batteries.
For an additional $3.00 you can add an OLED display which can display the current time, temperature, and humidity. This can make it a useful household device even if the residents will not use the data being saved on the microSD card.
The Mini Pearl Logger uses an Arduino Pro Mini ($2.00). There might be no advantage to building a logger from a less expensive bare 328p chip. The Mini Pearl Logger seems to do exactly what you want and is a tested solution with complete assembly instructions and open source software.
This is beautiful! Thank you so much! I heard stories of your feats at the last Barnraising, and I am glad that you're responding to this note.
This is definitely what I'm looking for. I will definitely test one of these babies as soon as possible.
I'm glad this might help. One clarification to my comment above: Adding a display to the logger works best if the logger is plugged into the wall (house power). An always-on display is incompatible with the low-power timer that allows the logger to operate for months on small batteries.
More clarification on the display: For a battery powered logger, you could use an e-paper (e-ink) display which retains the image after it has been powered down. So when the low-power timer turns on the Arduino (e.g., every five minutes) which reads the sensor and writes to SD card, it can also update the display which will then be static until the Arduino powers up again. I think the e-paper display requires no power between updates.
For $4.00 you can buy an e-paper (e-ink) display (1.5" x1.5"). I am not certain that this one works as I describe, but I just ordered a few to test.
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
Definitely. Two things:
I'm trying to find a difference between using an Arduino Mini Pro or an ATmega328, because they seem to be fairly the same (I found the same costs online for both). I made a purchase of both and will try to make tests on manufacturing takinig into account cost, easiness of making the sensor and quality.
The display is definitely something that will make sense for a user. Maybe a seven-segment display is cheap and simple enough for showing the results, like old designs. I'm thinking of a very, very simple interface for people to make use of, perhaps to input their perception, and see how good they are at forecasting weather (rain, heat) in an effort to gamify it. I became inspired after this edX course: https://www.edx.org/course/backyard-meteorology-the-science-of-weather
The standard (last generation) Arduinos are circuit boards with the ATmega 328p microcontroller chip on them. A common hacker exercise eight years ago was to build your own Arduino board with an Atmega 328p chip and a pile of resistors, capacitors, wires, LEDs, etc. Today, factories in Asia make reliable clones of Arduino boards and sell them for $2.00 including shipping from Asia. If your goal is to exploit the power of a microcontroller, starting with a $2.00 board is a good choice.
Seven segment displays might be the least expensive way to display results, but LCDs, OLEDs, and e-ink allow any ASCII character to be displayed and OLEDs and e-ink allow graphics. A $4.00 display has room for 20 to 50 characters. This can improve the user experience.
If your goal is user interaction (the user inputs information and the device responds), a 328p Arduino might be too limited for this project. The memory limitations of the ATmega 328p might allow operating a sensor, SD card, real time clock, and display, but even those components are pushing the limits of the 328p. Adding an input device and the code that will allow something akin to gamification could push it over the top. This also might push the BOM over $15.00.
If your goal is user interaction (or other added features) you should probably upgrade from a 328p Arduino to an ESP8266. These boards cost $3.00 instead of $2.00 but have four times the memory, speed, and power of 328p Arduinos (chart here). They also have integrated WiFi, but you can ignore that when it is not needed. After you have upgraded your Arduino IDE to support ESP boards, the sketches for 328p Arduinos will run mostly unmodified on ESP boards. The extra memory will allow operating additional devices (other sensors, input buttons, sound generation) and also allow any code needed to support user interaction and nice things like cool fonts.
But if you just want to log temperature and humidity and maybe display the current values, you can do it for $15.00 with a 328p Arduino.
This is perfect. I ordered ten of these Arduinos and will start playing with them.
Perhaps we can begin by the simplest which is logging and a couple of interface actions like push a button and a LED, and then build interaction on top.
I appreciate a lot your feedback, Chris! I'll reach out to you when I have something going on to hear your thoughts! Thanks!
Reply to this comment...
Log in to comment
Login to comment.