The Cubelets App has two main functions: Remote Control and Personality Swap. We’ve already introduced you to the Personality Swaps, but have you begun to use Remote Control in your classroom? There’s a hidden feature I want to highlight for you because it’s not the first application people think of when they see a title like Remote Control: gathering data about our robot constructions.
(Before you continue, it’s a good idea to make sure you understand how data travels through Cubelets by either reading this
blog post or taking the
Cubelets 102 (free) online workshop.)
As you already know, you can easily gather information about how data is traveling through a Cubelets robot construction using the
Bar Graph Cubelet. The Bar Graph is also a screen-free way to gather data about your Cubelets constructions. It simplifies the numbers into a 1-10 scale, as opposed to numbers between 1-255, so it makes data flow conversations available for students who are still emergent mathematicians.
However, there is one thing
Remote Control can do that Bar Graph Cubelets cannot: collect information about every Cubelet in a robot construction at the same time. By screenshotting the data in Remote Control, students can very quickly gather static data to analyze later.
As students build more complex creations, especially by adding multiple SENSE Cubelets, it’s more important that they check their assumptions about how the data is flowing through their robot constructions. In general, the five main states of a two-SENSE robot are:
- two sensors at 255,
- two sensors at 0,
- two sensors at ~127 (about halfway),
- one sensor at 255 while the other sensor is at 0,
- and vice versa.
Continue reading → 
Console replaces the Blockly and C programming apps and lets you do both at the same time. It also lets you Personality Swap your Cubelets to change their behavior with pre-written programs. Most exciting for me, though, is the new Data Logger interface. Connect a Sense Cubelet or three and watch their block values change over time. Add an Inverse and watch the complementary graphs. See if your kid can make a sawtooth waveform with a distance sensor. Track temperature or light data and export it to a CSV or a Google Sheet. Experiment!
I’m at home this afternoon and I noticed that although the sun is streaming in a couple of south-facing windows, our little cactus was in the shade between the windows, a temporary dark spot as the sun tracks across the sky. I thought maybe I’d build a little self-driving car for the cactus. You know, like Uber for plants!
Cacti like to be in the sun, so I figured I’d start with a Drive Cubelet as the base and a couple of light sensors so that the robot knows where to go. But before I put a plant on top of anything, I knew I’d want to ease the back-and-forth motion of the robot so that it didn’t come to a jarring stop when it found light and bounce the plant right off. I thought it’d be elegant for the robot to slow its velocity along something like a sine wave. Sin() is a bit heavy for a microcontroller, so I found a web site that generated lookup values for a sine wave and tried pasting a few of those data values into a Blockly program (sine-down.cubelet) that sine-waves down from 127 (half speed) every few seconds. Then I found the magic of Console. I switched quickly back over to Data Logger to verify that my code was working and saw these little approximated sine waves.
Then I flashed the code into a Drive Cubelet and saw this:
Looking good. Next I attached a couple of light sensors and tried programming the Drive Cubelet as sort of a lopsided state machine. If a light sensor has light, then drive toward it for a half second, slow down, and begin again. It worked! The little platform is successfully moving the cactus to the brightest sunshine. It’d be fun to add a couple of distance sensors to make sure that the plant car doesn’t bump into anything or fall down the stairs, and maybe some down-low glow, but I’m pretty happy with this for now.
Where were we, anyway? Right, Console! Console is a huge upgrade to the high-ceiling Cubelets experience. I was just doing distributed robot programming using multiple languages and leveraging inter-robot messaging schemes. And it’s for kids! It’s super cool to be able to sketch out a program in Blockly and then pop into C to understand the exact code that got created. There are lots of ways to look at the same algorithm, and lots of ways to understand things. Give Console a try and let us know what you build.
For the Beginners: It’s easy to be intimidated by Cubelets when you first pull them out of the box. After all, they’re blocks…that teach computational thinking…but how?
