Technical Info

NEMA 4, Polycarbonate box

NEMA 4, Polycarbonate box

I’m looking into a Nanospark application where, given the environment (automated greenhouse), I figured we’d want to use a NEMA 4 box to protect the tablet and electronics from humidity and water.  So the question naturally arose, will the wifi signals work with the box? Wouldn’t you know it, I happen to have a NEMA 4 box on my desk.  It’s polycarbonate with hinged clear cover (model YH-060604-05), made by a company called PolyCase.  I pull out the gasket, drilled a small hole for the Nanospark cable, put the gasket back in, tucked the iPod touch 4 in the case, latched the lid- ready to test.  On the Nanospark housing I put an alcohol sensor module (MQ-3) and connected the vcc of the module to digitalOutput2. It works perfectly.  To be fair, I’m about two rooms and 30 feet away from the router.  So I was starting to think of further tests, when something that’s second nature to my generation finally triggered in my skull.  Google it.  Turns out polycarbonate is an ideal enclosure material when working with wifi. Here’s an example of what I found: “One of the major reasons why polycarbonate enclosures work so well in this telecommunication industry is that the polycarbonate material does not block the wifi “waves” and the whole system can be placed inside of the enclosure.” (from www.integraenclosures.com) So, if we can take this automated greenhouse somewhere, you’ll probably see it in a NEMA 4 polycarbonate enclosure. Of course, the same would apply if you’re going to put a Nanospark system in a dusty, grimy industrial...

Read More

Color Detection With Multiple Calculation Areas

Color Detection With Multiple Calculation Areas

We had a farmer from Australia contact us with a fascinating idea, color detection.  Could we put an iPod touch on the chemical tank of  his tractor and use the Colour Detector app to trigger the weed killer to spray only when it sees a weed?  We loved this idea not only in cost savings to the farmers, but also how good this must be for the environment. To make it happen though, we needed to tighten up Colour Detector’s gathering of RGB values.  So we separated the screen out into a grid, each looking for and reporting it’s own RGB value.  Here’s our proof of concept report. Attempt 1: The first attempt at color detection with multiple calculation areas I used six selection areas. Each area was a 160px by 160px square. Using three rows and two columns the screen was divided perfectly since the 3.5 inch retina displays measure 320 by 480 pixels. With six selection areas the results were less than favorable. When looking at solid colors or nearly homogeneous patterns the RGB reading would be pretty similar from one square to the next. The performance in this case is passable. When looking at areas that were not homogeneous though, the readings would fail to reflect the presence of an object such as a tuft of grass. I assume that the problem lay in the fact that each calculation area was too large. Each area was looking at 25600 pixels! Even if a pretty sizable object were to be inside a calculation area is would only be a portion of the total pixels. Further the other pixels are already a mix of red, green, and blue just simply in varying degrees. The colors from the object would likely fail to significantly influence the color average. Below are a few screen shots of the first attempt. Attempt 2: For the second attempt I decided to up the number of calculation areas. My hope was that the smaller calculation areas would lead to more “sensitive” averages since there would be fewer pixels involved. More sensitive calculation areas should improve performance both with homogeneous subjects and heterogeneous subjects. I split the screen into fifteen calculation areas with five rows and three columns. The 320px by 480px screen did not split up as nicely. In order to have equal calculation areas I chose to have small 1px columns and rows...

Read More

Nanospark – A ‘Smart’ Control System (Our First Demo)

Nanospark – A ‘Smart’ Control System (Our First Demo)

What is Nanospark?  Nanospark is the foundation for whatever monitoring or control system you’d like to create.  What’s really slick about Nanospark is the interface.  It connects to an iOS device, like an iPod touch or iPhone, which lets you interact with the sensors, actuators or other equipment through an app.  Of course, you can program an app to do whatever you need with those sensors.  You can use functions like text messaging or GPS location to enhance your system with useful features.  We worked up a simple demo to help everyone visualize how Nanospark works. To help you replicate the demo, if you’re so inclined, we put together a walk-through.  It’s not step-by-step, but will help someone who already has a basic knowledge of Objective-C and Xcode. If the demo sparks an idea in you please try it and tell us about it (on Facebook or...

Read More