Monday, 18 August 2014

My Last visit to a Greenhouse for Wireless Sensor Network Experiment

Trip to a Greenhouse in Tamil Nadu- WSN experiment. 

It was 13th of Aug afternoon, we left our office (me and Srinivas) to our way to A greenhouse located on the border of Karnataka and Tamil Nadu. I had visited that greenhouse with my boss a day ago and today we were require to go over there to conduct a WSN(Wireless sensor network) experiment. That greenhouse was approximately 32km form our Bangalore office and we decide not to take bus but rather take scooty of one of our colleague. And here we were on scooty, on our way to greenhouse. 

Our sole objective of conducting WSN experiment in that greenhouse was to test the LOS and Non LOS range of the wireless modules that we are using and to test the capability of wireless transmission of data which we had tested earlier in our office. It is always true that to test the real efficiency of a system, test it out in the field away from resources rather than testing it in office. In the same greenhouse we had already deployed our GPRS/GSM equipped temperature logger, that is continually sending data to our IOT cloud ,which they can access through a login and android app, can get notification based on alert rules they had put in their login.
Yuktix IOT based temperature logger




So this was some thing that can measure the temperature and humidity of the location or a specific chamber where this logger is kept or maximum we can do is we can add 4-5 more sensors, keep them in different chambers and read the data from them. But for that we need to keep the limitation of I2C bus that without I2C extender, we cannot extend the range of the wire beyond a certain limit (although we can, but them we need to compromise the frequency of I2C bus). But again, what if we want to the measure the temperature and humidity variable of a chamber which is almost 500 meter away from the place where my datalogger is kept. Ideally they don't want to make every sensor unit equipped with GSM/GPRS as it wouldn't them meet their financial criteria. 

So we propose them, that we can deploy a wireless sensor network in their greenhouse and can get sensor data from the remote unit with in the range of our wireless module which is absolutely limited(it cannot be more than few km, apart from the fact that we use 5.8 Ghz modules with 14 Dbi or 22 Dbi patch antenna which will again be worthless if we take in count their cost, effort and other technical specifications like compatibility with our solution etc). SO we decided to hookup with Zigbee and continue our WSN experiment rather then going with RF modules which are again prone to noise.

We decided to start with 3 end-node(slave) sensor equipped with our Zigbee module, 1 repeater(coordinator) and 1 control unit or master . We started with testing the range of the module in reality in the field(they always say that it will work for 1-2 km but believe me in real time conditions without a patch antenna it wouldn't go beyond 500 meters). Although we had tested the same modules in our office region which had many obstructions and we got a range of 50 meter(Non-LOS).

Rhutvik doing WSN range experiment near office 
While when we tried out the same experiment in LOS in greenhouse, we got a range of approx 350-400 meter and we were very much satisfied with the same. We tested our experiment with a temperature and humidity sensor plugged to our DL box with Zigbee module attached and we were moving in the field with master module attached to the laptop.

Srinivas conducting WSN experiment in the field.
So at present we conducted the range experiment with only temperature and soil sensor. On our second visit, we had decided that we will test out the same experiment by adding Soil sensor(Soil Mositure and Temperature), Air humidity and Temperature and Lux sensor with 1 repeater and 1 master. Along with creating a sensor network, we will also find out the power consumption of repeater(we are trying to equip our Zigbee repeater with solar power ). Our sensor-end node are already low power consuming sensor nodes, which go in sleep mode after sending the data to the master but still we need to figure out the exact amount of the power consumption. Below is the glimpse of what we will be trying in our next visit.


No comments:

Post a Comment