Anyone do any graphing?

IVB

Senior Member
Wondering if folks are doing any graphing with their systems, and if so, what do you graph.  After 5 years I finally got my SQLServer back up & running as I wanted to test the efficiency of the new insulation I just put down, and quantifying heat loss in the rooms with big windows at the times i'm not there.
 
I just got it operational a few hours ago so this is the first graph. Was a fully manual process to create this, but before automating anything I wanted to build it manually to see what i'd want to do.
 
 
ibN9cTLh.png
 
I'm not doing any type of graphing yet.  Currently, I'm in the process of dismantling an old security/automation system that's been in use for almost 15 years.  My plan is to use a "server farm" for lack of a better description of five Raspberry Pi 2 B+ computers to handle all the required processing.  To give you a better understanding why I'm not using the old system for my current endeavor, one RPI 2 can substantially outperform the 20th century controller that I'm currently using.  Thus, even if it were possible to interface everything to the old controller, it would simply drop dead from being overwhelmed with processing requirements.
 
But anyway, getting back to graphing. I plan to to use Thing Speak  https://thingspeak.com/  which is essentially a IOT cloud app to do all my data analytics.  I plan to use wireless humidity and temperature sensors installed though out the home and outside which will report back to a RPI2 which in turn submit the data to Thing Speak.
 
Currently, I'm converting UPB control over to the RPI2 from the old controller but data analytics is definitely on my TODO list as well as voice recognition via Amazon echo.
 
I've attached a link to give you a better understanding of my analytics undertaking.  The only difference is that I will be using remote temp/humdity sensors  http://www.raspberrypi-spy.co.uk/2015/06/basic-temperature-logging-to-the-internet-with-raspberry-pi/
 
Interesting approach, i'm jealous of the humidity sensors, thats my/Elk big weakness. I re-architected and fully over-engineered my system 2 years as I also use the system to record & stream TV/BluRay/DVD, plus DropBox, so I have very little incremental cost for a local approach. Although I did put SQLServer on my OneDrive and might look into Azure in case I want to switch to a self-controlled cloud based approach.
 
Here graph weather, 1-wire temperature  / humidity sensors, water flow via 1-wire counters, lightning, rain and similiar for OmniPro 2 temperature and temperature humidity sensors.
 
For the furnace and air conditioner and none I keep the flow of air running at low speed 24/7 for even room temperatures. 
 
Also utilize an electronic air cleaner with charcoal and regular filters that are replaced every month.
 
Used to graph a lot but not as much today.  I also have always utilized one of those temperature guns which show me hot or coldspots.  (previously utilized the gun for server rooms et al).
 
I still keep the stuff locally even though I have Amazon, Microsoft and Google stuff.  I have recently been able to use 1-wire sensors connected to a Raspberry Pi2.  Works fine.
 
Just reading the text file with a script to pull temperature / humidity readings.  I can do the RRD stuff for some graphing.
 
The temperatures are pretty even except for the  bedroom over the garage which is always a bit warmer in the summer and cooler in the winter. 
 
Playing with the microrouter and bit banging an RTC clock was able to add a 1-wire sensor inside of the device.  The Ubuntu 64 bit Homeseer server is also a LAMP and Oracle VB box these days.
 
The Omnistat here manages the inside weather.  The weather station manages the outside weather (and there are a few 1-Wire sensors).  Still cannot automate outside weather though. 
 
I have a whole house humidifier and during the middle of last winter cranked on it a bit and it got a bit too tropical jungle like.
 
So here turned it down a bit and purchased a small Honeywell humidifier (uses UV - not liking it though) just for the master bedroom which worked other than having to fill it with water.
 
I was  wondering if anyone went beyond just graphing their data such as doing customized programming based upon the interpretation of the data.
 
For example, I have one scenario involving lawn irrigation.  Just keep in mind that this is just theory.  I have not yet implemented it.  I was thinking about using  ThingSpeak or possibly Azure to interpret the data that I would submit  to a data analysis service and have it develop a customized lawn irrigation schedule.  I currently have 26 lawn irrigation zones and plan to expand to about 30 zones.  Some of the zones will get a full day of sun, others a partial day of sun and some zones will get very minimal sun such as zones underneath large trees.  I plan to use soil moisture sensors and water flow sensors on each zone.  The water flow sensor will be used to keep track of the gallonage used on each zone.
 
I was thinking of the possibility to develop custom watering schemes.  For example, ThingSpeak would  interpret the data I submit along with current weather conditions at my home (is sunny or raining?) as well as future weather conditions (will it rain tomorrow?).  It will then determine what zones need to be watered and how long to water the zones from the soil moisture sensor data etc.  Based on this data, it will develop a custom scheme to be executed based on local sunrise time.  For example, it may determine that zones 1,4,8,13,17 have to be watered today with zone 4 and 17 getting 10 minutes of watering as opposed to only 5 minutes for the other zones.  Tomorrow, other selected zones will be watered based on the data analysis.
 
If I use ThingSpeak Matlab analysis, it won't cost me anything.  But if I use Azures Streaming Analysis, I'll be paying about $11 a month based on the projected amount of data I will be sending to Azure.
 
Since this is the first time I've attempted such a project,  I'm open to all views, opinions, suggestions etc.
 
 
 
 
 
Just curious about using light / UV sensors too with your zone monitoring and what it is you are going to utilize to measure water flow irrigation zone and what it is you plan to utilize for moisture sensors (and how?)
 
BobS0327 said:
I was  wondering if anyone went beyond just graphing their data such as doing customized programming based upon the interpretation of the data.
 
For example, I have one scenario involving lawn irrigation.  Just keep in mind that this is just theory.  I have not yet implemented it.  I was thinking about using  ThingSpeak or possibly Azure to interpret the data that I would submit  to a data analysis service and have it develop a customized lawn irrigation schedule.  I currently have 26 lawn irrigation zones and plan to expand to about 30 zones.  Some of the zones will get a full day of sun, others a partial day of sun and some zones will get very minimal sun such as zones underneath large trees.  I plan to use soil moisture sensors and water flow sensors on each zone.  The water flow sensor will be used to keep track of the gallonage used on each zone.
 
I was thinking of the possibility to develop custom watering schemes.  For example, ThingSpeak would  interpret the data I submit along with current weather conditions at my home (is sunny or raining?) as well as future weather conditions (will it rain tomorrow?).  It will then determine what zones need to be watered and how long to water the zones from the soil moisture sensor data etc.  Based on this data, it will develop a custom scheme to be executed based on local sunrise time.  For example, it may determine that zones 1,4,8,13,17 have to be watered today with zone 4 and 17 getting 10 minutes of watering as opposed to only 5 minutes for the other zones.  Tomorrow, other selected zones will be watered based on the data analysis.
 
If I use ThingSpeak Matlab analysis, it won't cost me anything.  But if I use Azures Streaming Analysis, I'll be paying about $11 a month based on the projected amount of data I will be sending to Azure.
 
Since this is the first time I've attempted such a project,  I'm open to all views, opinions, suggestions etc.
 
in general i find that my / family life is too unpredictable for programming. Even the dimensions of analysis end up being volatile. The fixes to situations end up being physical in nature (ie, i'm going to need to change the vent size or even close one off in a few locations), which programming can't help with.
 
THen again, my house is coming up on 105 years old, its architecturally nice but man is it whacky. Screwy heat distribution is but one of a myriad of problems.
 
The fixes to situations end up being physical in nature (ie, i'm going to need to change the vent size or even close one off in a few locations), which programming can't help with.
 
For one project here utilized static pressure measurements and temperature stuff to fine tune a system.  It was a PITA and a bit time consuming. 
 
pete_c said:
Just curious about using light / UV sensors too with your zone monitoring and what it is you are going to utilize to measure water flow irrigation zone and what it is you plan to utilize for moisture sensors (and how?)
I just finished the  irrigation hardware/software and really didn't put too much thought into the next step of totally automating the project.  I was thinking about using a weather feed such as weather underground to determine whether it was sunny or overcast.  I have boilerplate code for weather underground from another project.  So, it would be an easy implementation for me.  But I will start exploring light/UV sensors as a possible alternative. 
 
I started a thread on the Raspberry Pi forum for suggestions on a water flow sensor.  Everybody has recommended the SeeedStudio sensor. http://www.seeedstudio.com/wiki/G3/4_Water_Flow_sensor They're about $15.00 each.
 
Moisture sensors will determine how dry the grass is and would feed that data to ThingSpeak.  ThingSpeak using that data and other data will determine the appropriate watering scheme.  There are quite a few Soil Hygrometer Detection Modules on the market that can be interfaced to the RPI (or the Arduino).  It's just a matter of determining which module would be the best fit. Here's a link to one example of monitoring soil moisture for a plant.  http://computers.tutsplus.com/tutorials/build-a-raspberry-pi-moisture-sensor-to-monitor-your-plants--mac-52875
 
I have found that there are a lot of RPI (and Arduino) projects posted on the Internet.  So, it's really a matter of  my "cherry picking" the best ideas from the best projects to be used on my endeavor.
 
I just did a brief search on UV/light sensors and there are quite a few sensors on the market accompanied by a fair amount of projects using the sensors.  I came across one project where a guy used the sensor in an advanced weather station project.  So, I will probably use the sensor in my project as opposed to Weather Underground.
 
Wondering if you will be utilizing a local on site weather station?
 
I have purchased stuff from Seed Studio.
 
Will you be installing the water flow sensors near the manifold and  solenoids outside?
 
Like to know a bit about numbers you will be getting from your soil sensors and how many you will be using per zone.  (I am an innate bean counter - if that makes any sense?)
 
Very interesting project Bob.
 
I would like to keep asking questions and now feeling a bit guilty here about hijacking IVB's thread about graphing.
 
pete_c said:
The fixes to situations end up being physical in nature (ie, i'm going to need to change the vent size or even close one off in a few locations), which programming can't help with.
 
For one project here utilized static pressure measurements and temperature stuff to fine tune a system.  It was a PITA and a bit time consuming. 
 
For my house, i'm currently in the "I need a sledgehammer, not a scalpel" stage. 105 years old = not remotely at fine tuning for HVAC.
 
Will you be installing the water flow sensors near the manifold and  solenoids outside?
 
 My yard is an extremely odd  shaped yard which covers a little over an acre that utilizes popup spray heads as opposed to rotor heads  The sensors will be installed after the sprinkler valves.
 
Wondering if you will be utilizing a local on site weather station?
If I do decide to utilize Weather Underground, the API allows me to set the latitude and longitude of my home.   WU will obtain data from the closest  facility to my home which is in another town about four miles away from.
 
Like to know a bit about numbers you will be getting from your soil sensors and how many you will be using per zone.  (I am an innate bean counter - if that makes any sense?)
 
I'm really not sure what sensor I'll use at this point since the project is still in its infancy stage.  One sensor indicates humidity (moisture) level of the soil.  Data could be used to indicate how fast/slow the zones are drying out which in turn would allow me to adjust watering schemes accordingly. To identify zones that are drying out faster than others possibly due to being in an area getting a full days sun.
 
 
Will you be watering all year round or do you have a winter where you live?
 
Are you utilizing your own water source (well / lake?) or do you pay for your water?
 
 
 
Googling this morning using the terms "the science of irrigation"; first hit was this paper. 
 
I have no idea when it was written as it just came up as a PDF only.
 
The Science of Irrigation: What is ET?

By Melissa Baum-Haley, M.E., MWDOC Water Use Efficiency Programs Specialist
 
This summer, MWDOC is featuring a new series of articles in eCurrents on “The Science of Irrigation.” Each article will strive to better educate and inform our readers about the principles and technologies behind watering their lawn and
landscapes. There is more to establishing and maintain a lush, healthy landscape than simply turning-on the sprinklers and letting them run.
 
A little history: the common lawn sprinkler dates back to the 1870s. By the 1930s, residential irrigation was frequently advertised in magazines. Homeowners were promised savings in time, money, and an increased appearance in their lawn during times of drought with the installation of a permanent in-ground irrigation system. Now almost all newly constructed homes include an in-ground irrigation system with an automatic sprinkler timer.

How much water is used for irrigation? According to the Environmental Protection Agency, the national volume of water used to supplement lawns and landscapes is more than 7 billion gallons per day – enough to fill 280,000 residential swimming pools.

So, why do we irrigate? Plants need water. Irrigation is used to supply plants with water when there is either not enough rainfall, or the rainfall doesn’t occur at the right time. The goal of efficient irrigation is to apply only that water lost through
evapotranspiration that is not already replaced by rainfall.

Think back to your elementary school science class. Your teacher stood before you and drew the water cycle on the board. It was explained that water evaporates and condenses into clouds; it bursts out in rainfall; it is used by plants or creates runoff into a pond next to the little house on a hill; then it evaporates again and we have a simple cycle. If only everything needed no more than a third grade explanation.  This brings us to the first “grown up” concept in the science of irrigation, EVAPOTRANSPIRATION, also referred to as ET. Evapotranspiration is the amount of water lost by plants through evaporation or transpiration. Since it is very difficult to actually separate evaporation and transpiration mathematically, these two terms are combined.

Evapotranspiration rates are based on a number of weather factors including wind, humidity, temperature, and solar radiation. All of these characteristics of weather vary throughout the year and from season to season, as do the water requirements of plants. Of course, every plant uses water differently, so each plant has a specific value, called a “crop coefficient,” that represents the amount of water the plant might need relative to its seasonal growing stage. A plant’s specific evapo-transpiration rate will either be increased or decreased based on it crop coefficient.With plant-specific evapotranspiration rates changing from week to week, month to month, and season to season, it becomes clear why your irrigation schedule needs to change as well. The amount of water that needs to be replaced in a plant is not static.

Is all this discussion of weather and adjusting your irrigation system leaving your head in the clouds? Well, have no fear, because the invention of the “Smart Timer” has made watering our plants and grass much easier and less confusing. The most common type of “Smart Timer” is the weather-based irrigation controller. These devices use an estimation of ET to schedule irrigation and are typically programmed with site- and landscape-specific conditions to make them more efficient than time-based systems alone (i.e. the standard irrigation, automatic, “set-it-and-forget-it” time clock).

Each Smart Timer works differently depending on the manufacturer and technology used, but typically can be programmed with site-specific conditions such as soil type, plant type, sprinkler type, sun and shade, etc. The controllers are designed to either replace the typical timer or act as an add-on device to the timer. There are two basic types of Smart Timers, those with on-site weather sensors or those that receive weather information from a remote weather source. Depending on how the controller obtains its weather information, signal fees could be involved, but there are many controllers for which there are none. Additionally, Smart Timers can have accessories to make them more accurate (such as a rain shutoff sensor), while others come as a complete package and need no additions. The EPA is even in the process of developing a certification for Smart Timers as part of their WaterSense program.
 
And it gets better. Many local and regional water agencies offer customer rebates for the purchase and installation of Smart Timers and other related landscape irrigation products. Check out www.mwdoc.com or www.SoCalWaterSmart.com to find out about customer rebate options in Orange County and greater southern California.
 
So, lesson one in the Science of Irrigation series is complete.
 
In the next installment, we will discuss water efficiency versus uniformity – two irrigation terms that are often confused.
 
Your homework assignment is to hold on to your empty food cans for the field exercise in August.
 
Back
Top