Crombot One: From a Starter Kit to My First Robot

Last month, having set out to have myself an “Auspicious August,” I took to looking to the skies for auspices.

While walking into work on that first day of August, expecting to spot a bird, I saw a drone fly overhead. This is all the more curious since I have seen nothing overhead but birds and planes since.

In keeping with my monthly theme, I took this as a sign that I should tinker with robotics, so I ordered a robotics starter kit.

Crombot One fits neatly into the Signal September theme since he is a machine made from signals: infrared commands from the remote, motor-control signals from the Arduino, ultrasonic echoes from objects, and reflected infrared light from the floor sensors, all interpreted and reacted to in order to create autonomous behavior.

Here is a short demo:

This whole robot-building business has already opened my eyes to a growing number of beautiful nuances in the mechanical and electrical world around me. When I showed Crombot One to my uncle, he explained how the circuitry could be made much smaller, and how the ability to pack vast numbers of circuits into tiny chips helps explain the importance of Taiwan’s semiconductor industry. That kind of miniaturization can involve increasingly complex and costly manufacturing.

That underscores how amateurish Crombot One’s wiring is, with its loose wires and poor organization. It is an educational practice project, after all, and as an educational practice project it has done its job of illuminating the potential of robotics.

While flight, that is, building a drone, is the natural vision after sighting that drone on August 1, I reckon I should learn the basics before taking to the skies. Crombot 1.0 gives me a simple terrestrial robot to play around with. For my next robot, Crombot 2.0, I would like to experiment with a different body. Flight remains enticing, but I wonder if a swimming robot would be a better next exercise. It also seems as though it could be much cheaper.

Even that is probably getting ahead of myself, though.

I think the cheapest and most rewarding next experiment is to continue developing Crombot One by mounting my phone on him and linking it to the Arduino over Bluetooth. The phone could handle camera and microphone input, then send commands for Crombot to respond. That would allow me to dive straight into experimenting with audio and visual sensory recognition and response.

Oh, and by the way, here is a longer demo of Crombot One I put together (I am practicing making videos):