Having able to build a decent weather station and a web-based interface to display light, temperature and humidity data is one small meaningful accomplishment myself but I were not confident that the data quality especially the dependence of a single sensor and the location. The data with one sensor is a hit-or-miss record at best. In this tutorial, my objecive are:

• Collect data of light, temperature and humidity from at least two sensors.
• Integrate other open source data such openweathermap.org to compare the local values to a more general area.
• Request data from openwathermap.org for current and forecast data to have more meaningful data.

The microcontroller is ESP8266. This is a low-cost and well-developed chip with WiFi connection. However, ESP8266 has one analog to digital converter (ADC) so I have to add an expansion board to expand to collect more analog data.

An new feature of this version is a larger battery and a plastic box to isolate electrical signal. The plastic box will have several open windows to let the air exchanged.

Beside two sensors with an insolated box, the data will display as the average and standard deviation of the sources. The tempeature data will be collected from 4 sensors. A strong focus of this version is on display and analyzing. The plotly graphing is a simple and light to dispay 1000 points of data. Too much points is a source of distraction. Also, I prefered to limit the display to a certain period of time rather number of point such as in last 24 hours. Looking the data by the limit of data points is easier to implement by SQL query.

Time estimation, about 4-5 hours a day (as I did it):

• Testing sensors: 1 days
• Soldering: 1.5 days
• Debugging: 5 days

This is my first attempt for this configuration, so there was time budget for planning and testing rather excecuting. If everything planned out, one day for building and another day for debugging would be sufficient.

Tesing sensors will be present in Step 3. The purpose is to ensure each module works and the reading sensors in a closer range possible.

Hopefull, that amount of time did not discourage you. The complete version of my weather station is below:

and the live charts are present here.

In the next page, a list of parts is presented.