|
Motive and Methodology Motive and Methodology |
| |
Movite： |
Could the information from open-area-based meteorological station represent that inside the forest around the station? If not, |
| |
what's the difference? |
Purpose： |
Microclimate monitoring on different landscape patches by wireless sensor network in Taiwan |
Methodology： |
|
Total of 30 Moteiv Telos motes, 2 for TOSBase and MultihopBase |
| |
|
Measuring at an interval of 5 minutes, duty cycle is 2% |
| |
|
Study area: Sitou Tract, The Experimental Forestry of National Taiwan University |
| |
|
Monitoring points: Flower bed (3m height), Japanese cedar stand (Cryptomeria japonica; 3m and 8.45m height), Taiwanese fir |
| |
stand (Taiwania cryptomerioides Hayata; 3m height) and Ginkgo stand (Ginkgo biloba; 3m, 9m and 11m height) |
| |
|
The longest direct distance from a mote to base station is 420m |
| |
|
 |
|
 |
Location of the study area in Taiwan |
|
Mote location map
|
|
|
| |
|
| |
|
Remote Configuration |
|
Mote Housing and Deployment |
| |
The system allows users to remotely: |
|
The motes are coated and sealed with anti-moisture material |
|
Turn on/off the LEDs |
|
They are then situated in a special housing which shields the |
|
Adjust data collection cycle |
|
motes from rain as well as lets air circulate |
|
Adjust duty cycle |
|
The combined units are very mobile and easy to maintain |
|
Reset all motes |
|
The current deployment has been live for over 3 months without any |
|
Program new TinyOS image
|
|
major hardware malfunction |
 |
 |
 |
| 通風筒 |
安裝於樹幹的 mote |
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
| |
|
| |
|
|
Sensor Specification and Calibration |
| |
| |
|
Air temperature - Sensirion SHT15 |
 |
微氣象站內進行戶外率定 |
|
| |
- Resolution：<=0.04°C |
|
| |
- Accuracy： <= 0.3°C at 25°C |
|
|
Air humidity – Sensirion SHT15 |
|
| |
- Resolution ： <=0.5%RH |
|
| |
- Accuracy： <= 2.0%RH at 25°C |
|
|
 |
 |
|
|
Indoor calibration with Rotronic-MP101 inside a damp proof |
| box yieldedsatisfactory results for both temperature and relative |
| humidity. The results of outdoor calibration with Vaisala- |
| HMP45D as reference, however, were not asgood as expected. |
溫度戶外率定 |
相對濕度戶外率定 |
Variances were decidedly noticeable when temperature was |
 |
 |
|
| higher than 16°C. |
| |
| |
| |
溫度室內率定 |
相對濕度室內率定 |
|
| |
|
|
| |
|
|
|
|
Data Yield Rate |
| |
Data yield rate drops dramatically as hop count increases. This is consistent with many published
|
|
 |
跳接數與回傳比例關係 |
| results. Known issues with underlying TinyOS libraries are being worked on. We anticipate significant |
| improvements on data yield rate before the spring/summer months when microclimate fluctuation |
| becomes more pronounced and more delicate plant physiology analysis will is made. |
|
|
|
|
| |
|
|
Power Consumption |
| |
|
We used Duracell, Panasonic Alkaline, Panasonic Oxyride, Toshiba Alkaline and Sanyo 2500mA Ni-MH Batteries for actual deployment. |
| |
They all have almost the same life time before voltage dropped below 2.1V (from 3.0V) |
|
A pair of above AA batteries lasted for 7 days when duty cycle is 66% and 14 days when duty cycle is 5% |
|
We have built a prototype with a solar panel coupled with a pair of rechargeable batteries. It has been proven capable of providing perpetual |
| |
power source for this deployment |
|
 |
 |
 |
|
| |
|
| |
|
|
Microclimate Monitoring of Different Landscape Patches |
| |
|
| |
|
| |
|
|
Web-Based Monitoring http://sabo.fo.ntu.edu.tw/wsn2 |
| |
|
Latest data (refreshes every 30 seconds)
|
 |
 |
 |
 |
 |
 |
|
|
Temporal charts of temperature |
|
Temporal charts of relative humidity |
|
Temporal charts of power consumption |
|
|
Previous day’s data download (CSV)
|
|
Mote location map |
|
|
| |
|
|
Conclusion and Next Steps |
| |
|
According to monitoring results, temperature and relative humidity were significantly different outside and inside forest |
|
Anti-moisture coating and our special mote housing were very good against the high humidity in Taiwan (>95% every night). However, the |
| |
upcoming summer storms will present the true test |
|
Much work remains to increase data yield rate |
|
2 AA Alkaline batteries could last for about one month when duty cycle is 2%. However, it’s still not good enough for field research. A solar |
| |
panel-based power strategy is required |
|
Placing Sensirion SHT15 and Hamamatsu photodiodes side by side prohibits them from sensing accurately at the same time. |
| |
Considerations need to be taken for future layout |
|
| |
|
