In retrieving data from our prototype, we noticed that it had stopped recording in the middle of the night, around 12:00AM. Additionally, only one sensor contributed viable data, and analysis of that data found no detectable pattern or trend.
So why all these problems?
First, why would the system randomly shut off at midnight? We inquired with the greenhouse whether there would be any reason for this, and they informed us that some of their outlets are on a timer. Ah-ha! At least this is an easy fix; we just have to use a different outlet, or switch our system to battery power.
Secondly, why was only one sensor contributing data? We suspect this had to do with sensor placement, and plan to experiment with different placements to see what works best.
Finally, why was there no trend in the data we retrieved? We don’t have a solid answer for this, but we did notice that the sensors had begun corroding over the single night we left them at the greenhouse. Our current theory is that this process was accelerated by the pH and other chemical properties of the nutrient solution the plants are watered with. We found another sensor to use that should be corrosion-proof.
We plan to test our next iteration in the lab, and will update you all on the results!