these are some pics from the project in kenya 3 years ago  … there are hotspots on the image that link to some project pictures – see if you can find:

• apiary made from mud and beehives.
• greenhouse made with local materials
• rapid compost bins
• vermicompost
• double digging and gravity fed irrigation
• squabs and poultry
• river beds
• nursery
• termites
• and others.

some links are stale .. will soon fix these

it is easier with smart aquaponics garden … you can let it run, collect data then make your adjustments remotely … while at lake tahoe.

it took 2 (cron) timing adjustments to optimize flood cycles for the eco village garden. in future such adjustments will be automatic or through drop down menu’s rather than from the command line as in the example shown here.

the following graph is for the first three garden cycles (each cycle is 60mins) and the two adjustments required to optimize the bed flows.

100 means the growbed is 100% full (it is flooding) …  0% is empty (drained) … these are flood and drain gardens – the negative values are forms of electronic noise

the following graphs shows the garden a few hours later after optimization. the sensor on the blue grow bed (red graph) is noisy … physical calibration should fix this.

will post url for this shortly.

i have been able to call in some problems remotely for some issues to be fixed by brian at that end – mostly debris in the indexing valves (nets were put around the pumps and this has not been a problem since then).
created live graphs of the growbed sensor data and was able to catch various internal problems – (internal – means it would not really be easy to catch these the garden was not a smart garden … leave along figuring these out remotely).

the live graphs gave the first clues:

the plots are for level sensors in the gb … interpreted as:

• green gb appears ok
• blue gb fills slows and drains quick … the stand pipe needs a quick adjustment(built into design) to slow the water some, the sensor needs calibration which i can do remotely by conditioning the data,
• black gb did not fill long enough … i wonder why. – the stand pipe appears is well adjusted.

just to make sure everything was working as expected i checked out the graphical simulation for this … i can see the sensors and garden relationship better with this view:

every appeared to work as expected. (the graphic is still for a single growbed but the sensor values are overloaded with extra data from the 3 bed garden)

next, checked the raw arduino messages  through the serial port. the json data is enriched further by by the router before transmission to remote db servers. this is shown below – notice the padded fields mapping the new garden type.

still no problem or cause of the black gb operating as it is … i checked the crontab  listing and … i found my problem:

do you think you can find it in the screen below?

one of the jobs is set to go for 15 mins while the others are doing 10 mins

will fix this remotely  … then will continue with the smart garden diagnosis in another blog … i suspect i still have one more problem i have not found yet probably with the garden itself.

smart garden farming is fun …

straight from the laser cutting press.
version 1 is almost ready for production … working on enclosure.
8 sensors – water flow rate, fish tank level, gb level, drainage flow switch, reservoir level, leak detector, temperature and light
4 relay outputs for controlling fish pump, reservoir pump(or ro filter solenoid), fish feeder, grow lights.
wireless – wifi .. garden will join cloud using existing internet connection.
secure python django backend – garden can be untraceable
embedded linux arduino based – open source reliability
processing.js ajax client – instant graphical simulation of rich garden data
we will give away out some smart aquaponic controllers to our social fans and followers once we start production – now is a good time to join.

It’s no fun when one is a serial fish killer.

Aquaponic tasks are repetitive in nature and the response time for plants in a garden is relatively slow, and my default behavior skews towards laziness. A bad combination for any successful urban farming endeavor. I have had many unsuccessful gardens; the only times I have been consistent and produced results is when I have had assistance from my computer. When wirelessly dialed into sensors in an aquaponic system, a computer can autonomously take over mundane tasks such as feeding fish, maintaining proper system flow, refilling water, monitoring water quality, and controlling lighting and temperature – all while notifying me remotely through my cell phone from anywhere in the world.

When I say “computers” I really mean microprocessors, or more accurately, microcontrollers, also known as single-board computers (sbc). Personal computers (desktops, laptops) have a few microprocessors in them – one for the display, another for networking and listening to peripherals, and finally, one that does the actual thinking (which is really just overkill, meant to process documents, browse the web and play games). The first rockets to the moon had less microcontrollers than my phone, and were far less powerful – one for navigation, one for making sure the engines were working, another for talking to the humans on earth, etc.

Microprocessors are mini-brains, capable of making millions of decisions in under a second. Physically, they are interconnections of millions of electronic transistors that can be rearranged to perform different functions based on binary instructions passed to them. These binary instructions are a program code, more commonly known as an application, which can monitor and control an aquaponic system. Because the different components of an aquaponic garden are engineered by hand and are directly accessible as compared to traditional farming, electronic sensors can easily be embedded in the garden’s various components. Electronic sensors allow environmental measurements to be taken that can then be transformed into electrical signals, which become the inputs for a microcontroller – somewhat like a keyboard that senses when we depress keys. For example, a float-switch sensor will float when the water in a tank is high, electronically indicating that the tank is full, and visa-versa for when it’s empty. Floating, the switch is “on,” indicated in binary language as 1; when the water drops, the switch is “off,” or 0. Microprocessors love these digits, they will suck them up like there is no tomorrow.

In short, aquaponic garden environments are microcontroller-friendly. Things that need to get done to keep an aquaponic garden running smoothly involve turning on and off the pump, refilling and testing the water, feeding the fish, etc., and all of this binary output information can be transformed into a user-friendly app, which allows one to control the system wirelessly and remotely. Aquaponic gardens are made up of inherently symbiotic relationships, and I feel that extending these relationships to electronic microprocesses are the missing link between aquaponics and the future. This “low-tech” high-tech is key to conserving the earth – a contained, biologically self-sufficient food production system, mimicking nature while relying on microprocessors. Why else would we send a robot to Mars to test rocks and dust (the definition of reliable “LoHi-tech”)?

And so: smart aquaponics are simply aquaponic gardens with microcontroller smarts