First week is over. Overall it going well. 4 of the 6 plants are excelling; super boof auto and gorilla z auto). The 2 strawberry gorilla autos are not doing very well.
Right now for strawberry gorilla, I believe some of my issues may be genetics specific. Both plants exhibit same issues. Each one is in a different medium. The only constant is strawberry gorilla auto genetics. I am going to continue growing them for another week to evaluate options. One I am fairly sure is going to die. The other, while overall looking terrible, has some new growth starting that is a better healthier green. Hopefully it is just a rough first week. But these are autos. A rough start usually means a rough entire grow. Not necessarily that they will keep having issues, but due to short fixed veg time frame, these 2 have significantly lost growth time.
The DIY IoT automation side of this grow/project is progressing nicely. I worked on the PAR sensor and have no deployed it for its first 24 hour “soak test” to ensure optimal operation for 24+ hours. If the soak test last, it means it’s pretty much ready to be deployed. The PAR sensor is quite good if I am being honest. I was able to verify my ss550 LED at ~1000 umol/s it is rated for at a specified height.
Now is when these sensors start providing real value. First week of growth always requires low light levels/high light height to accomplish providing somewhere around 150-250umol for seedlings. Without the PAR sensor to use at the beginning, I do what I always do and adjust height and light levels. Unscientific, but it’s how most growers are doing it due to PAR sensors like apogee for the dli500/600 PAR sensor starting at 5+ and anything below this range usually is not a true PAR sensor and no longer of value outside of “consistency” checks. They are lab grade, they are meant to be, this explains their price. Apogee and other comparable sensors are also spot checkers. They are used to check, validate and log, but they are not meant to do it fast or at a rate that is useful beyond average DLI. They can’t be integrated into automation systems. They largely are siloed devices that work alone and must manually transfer logs for analysis. The Dfrobot PAR sensor is a real time, continuously polling, connected, that can be interacted with in real time and used to control automation. It is also industrial grade with ip67 rating compared to apogees ip65 (it’s a spot check probed not meant for automation integration). Basically, apogee are amazing sensors, but they fundamentally are unable to do what I want. This Dfrobot par sensor, while not lab grade, is nearly lab grade, which in turns bring the sensor significantly down in price compared to a “lab grade” sensor that cost 4x. I am not a researcher or a scientist. Lab grade is amazing, but I literally cannot abstract additional value for the statistically significant accuracy difference (remember statistics and our p value, statistically significant is often a very small number , but in the real world, it does not matter since I am not doing actual research nor needing perfect controls with only a single variable change. I am only one person. ;) Uhh… I think I got lost on a tangent, ultimately PAR sensor confirmed 251 umol/s, which is the maximum these should be getting. But 4 are thriving, 2 are not. I’m not adjusting the lights for the 2 struggling to the detriment of the others
Now that the PAR sensor is running 24 hour soak test and about to be deployed full time, I’ll be focusing on the other two sensors; the 4 in 1 soil (ec, ph, soil temp, soil moisture) sensor and NPK sensor. All three of the sensors mentioned are connected by a single Ethernet wire for communication and power to simplify deployment. The 3 sensors will ultimately plug into a shared hub that allows for easy plug n play for these sensors and future ones.
Hopefully by next weeks post I’ll have full logging of all sensor data to my timescale database which will store all sensors logged data for future analysis.