Servo Magazine (February 2018)
Microcontrollers — They’ve Finally Made it to Cracker Jacks
Kids of all ages know that Cracker Jacks have a prize inside. Maybe a decoder ring. Maybe a toy soldier. Maybe a rub-on tattoo. Whatever the prize, it’s by default a low-cost commodity item. If something has made it to the status of a Cracker Jack prize, it’s probably nearing extinction in the free market.
Of course, microcontroller boards aren’t really available in Cracker Jack boxes — at least, not yet. However, they’re cheap enough now where I consider them not worth repurposing — that is, they’re disposable. Take the popular Arduino, Raspberry Pi (strictly speaking, a microprocessor board) and, say, the Wi-Fi enabled NodeMCU. They’re all available for $5-$10 in single-unit volumes. At those prices, it’s not worth spending an hour unsoldering and disconnecting prototype circuits.
Aside from the issues of whether to repurpose commodity microcontroller boards, the affordability of these boards means that there is no excuse for the robotics experimenter to become fluent in their operation. One challenge is picking a processor board and going with it without constantly looking back.
For example, lately, I’ve been spending a lot of time interfacing Arduinos to the ESP8266 to web-enabled devices. The NodeMCU incorporates the ESP8266 and essentially provides web access for free. Even more attractive is the NodeMCU’s support for the interpreted programming language Lua which — for beginners — is simpler to learn and use than the flavor of C supported by the Arduino.
The great upside to virtually free embedded hardware is a future where everything has embedded intelligence. Forget bar codes being replaced by dumb RFID chips. Bar codes on everything from fruit to lawnmowers will be replaced with intelligent chips that tell the user the history of the item from inception to stacking on the store shelves.
Imagine if the chip embedded in the carton of eggs could warn you that the temperature rose to 70 degrees for an hour during transport. Or, that the cut of beef originated from a cow that was given a double injection of growth hormone. Or, that the box containing your potentially new LCD TV was subjected to a 9G drop during delivery. That’s all data you’d probably be willing to pay a few cents for, right?
The only downside to ubiquitous embedded processing power is probably loss of privacy. Sure, your washing machine might be able to query your pants to determine if they’re sweaty and grimy enough to need a good washing. But do you want your employer to check out your clothes as you walk into the office? Maybe you like to wear your shirts two or three times before laundering them — but not if everyone else in the office wears a freshly washed shirt every day.
I’ve been trying to devise a way for my running shoes to indicate when the cushion has bottomed out — but I don’t want to be inundated with Google ads for running shoes.
At pennies per processor, what data would you want automatically posted to the web by embedded processors? Would you let Amazon or some other online store monitor your refrigerator, your clothes, and your daily activity?
In short, do you see the potential for good outweighing the potential for harm? SV