Four Lessons an Adult Engineer Can Learn From Children’s Toys
Children’s toys aren’t just for children. Inside every toy, there are hundreds of hours of engineering experience just waiting to be discovered.
My parents remind me frequently that no electromechanical device has been safe around me since I was five years old. My fascination always seems to be with putting things together or taking them apart. I very seldom ever played with the toys the way the manufacturers intended. I want to know what is inside—what makes them tick.
And when I got a chance to get my hands on Elecrow’s Crowbits modular learning system, my reaction was to take the modules apart and discover their secrets immediately. I don’t recommend taking the modules apart since you must remove the pogo pins for reassembly, and they are challenging to reinsert without damage.
As a matter of journalistic disclosure, I have purchased Elecrow products in past Kickstarter campaigns. However, for this article, the company provided a complementary CrowPi kit that I will regift and donate to a local elementary or middle school after writing.
Even though Crowbits is technically designed to inspire engineering curiosity in young people, my experience with the kits also reminded me of four valuable lessons an adult EE can learn from these toys and others like them.
About Crowbits and Makeblock Kits
Before we jump into those takeaways, here’s a little background on the kits themselves. Crowbits come in kits divided into three difficulty levels: no-programming required, programming required, and advanced application.
Each of the three kits contains lego-compatible modules magnetically connected via pogo pins and divided into categories color-coded by function:
- Blue: power and logic modules
- Yellow: input modules
- Green: output modules
- Orange: special modules
You can use the parts to make everything from a simple earthquake alert sensor that lights an LED when disturbed to a cellphone. You read that right—a cellphone!
The Crowbits kit includes magnetic pogo pin connectors. Image used courtesy of Elecrow
These modules might remind you of the Makeblock Neuron line of snap modules I covered a few years ago. I was so enamored by the lessons in modular design I learned from the Neuron that I made a Super Simple Sensor System. And I already have some ideas about future products that I might design using the lessons learned from the Crowbits experience.
The Makeblock Neuron kits. Image used courtesy of Makeblock
Both sets of modules are programmed using a spinoff of MIT’s Scratch programming language called Letscode.
With that context in place, what can a practicing engineer possibly learn from these children’s toys?
Lesson 1: If It Works and It's Safe, Find a Way to Make It Cheaper
Remove Superfluous Microcontrollers
Based on appearances alone, it seems that the Makeblock Neuron provided some design inspiration for the Crowbits modules. Makeblock’s Neuron has a feature that I love: there are precisely four pins on each interface, and there is a microcontroller in absolutely every module. This allows data to pass through all modules.
Not just any microcontroller will do; Makeblock microcontrollers need at least two and sometimes three available USART lines with independent buffers. Each microcontroller reads two UART lines, writes two UART lines, and requires one more serial data bus available for UART/SPI/I2C to connect with sensors on the block.
A simple representation of a basic UART (universal asynchronous receiver/transmitter) communication interface
That’s not a particularly cheap option. But it allowed the engineers to create the Makeblock ecosystem easily. A block can be whatever it wants to be as long as it has those two serial lines entering and exiting the block (one UART on the left, one on the right). But it also increases the overall cost of the system. After all, if the block is just a rheostat or a momentary switch, does it need to have a microcontroller? Probably not.
The Crowbits engineers made some different design decisions and advancements that decreases the BOM cost and increases the modules’ usability. They still have data moving between some modules, but sometimes that data is I2C, sometimes it is SPI, sometimes it is UART, and for plenty of modules, there is no data at all—just a power-delivery switch.
That decision significantly decreases the cost for some PCBs at the expense of increased design time.
Reduce Connector Cost
I noticed that Makeblock used magnetic pogo pins (molded into a single package) in their design. In contrast, Crowbits used pogo pins and magnets that slip into separately-molded connectors.
I don’t know the exact cost savings here, but several years back when I tried to source the magnetic pogo pins used in the Makeblock design, I found them to be approximately $1 US per pair at 1 ku quantity (factory-direct pricing). That’s quite a bit since individual pogo pins and magnets are only a few pennies per part. Off-board connectors are some of the most expensive project features, so I imagine this decreased the project’s overall cost in a small but significant way.
Makeblock vs. Crowbits pogo pins and magnets
I don’t think Crowbits would exist without Makeblock paving the way. I appreciate each design decision. If I were making these blocks, I’d likely go with Makeblock’s choice because I don’t think I have the foresight to use Crowbits organizational structure; they’ve spent a great deal of time planning this design.
Lesson 2: Make It Adjustable
Electronics engineering is a challenging profession. There are so many competing requirements for any given design that it’s difficult to do anything other than focus on the task at hand. But often, if we take a step back, we can start to look at the bigger picture.
Crowbits and Makeblock make an infinitely adjustable product. They can replace any or all of their modules without affecting the performance of the system. If a part becomes unavailable for any reason, they can respin the board for that module for minimal cost with new features, and the end-user would be none the wiser. They can expand or shrink the product to meet market demand.
littleBits is another modular electronics kit that benefits from adjustable products. Image used courtesy of littleBits
The overall adjustability makes these two designs relatively unique in the electronics engineering space. Far too often, engineers fall into the trap of designing completely bespoke solutions for their clients. Every now and then, I’ll see a hierarchical schematic design where a particular subsystem is laid out a bit cleaner than the others. Or I’ll notice a PCB that allows me to clearly trace the signal path as it moves into and out of the board.
Those are usually times when an engineer has reused parts of previous designs. Subsystem reuse can save a great deal of time, and modular design allows team members to focus on a design’s individual aspects.
So the next time you create a project, ask yourself if you can integrate future-proofing for this design by incorporating an expansion port.
- Are there unused GPIO pins on the microcontroller or an extra USART data bus you can route to the board’s edge for later use?
- Perhaps you can route it to an unpopulated header?
- Is there sufficient board space to put on alternate land patterns in case the part is unavailable?
- Can you use an FCC-certified module instead of designing your own wireless subsystem?
Lesson 3: Make It Idiot Proof
Assume that no one will read your manual unless something goes wrong. People take things out of the box, start putting bits together, and pushing buttons. Unless you design features that actively prevent a user from doing something they shouldn’t, you can expect a user to do it.
Both Makeblock and Elecrow polarized their devices with magnets and pogo pins. It is difficult, if not impossible, to improperly connect modules together. Magnetically-attached modules will naturally attract toward a proper connection and repel an improper relationship.
Do whatever you can—use polarized connectors, make connectors with a different pitch, or color code—to ensure that your users don’t need to read the manual to make things work. That will reduce the time spent in RMA troubleshooting.
Lesson 4: Do Market Research
It’s a pretty good guess that if you have an idea, another engineer somewhere in the world has had it before. Before you start dumping your time, sweat, and dreams into the next big thing, do a bit of research to determine if others are willing to pay for your ideas.
Elecrow created Crowbits because they saw the success of Makeblock’s Neuron. They had a slightly different take on the implementation, but they knew there was a market for it because it was already on the market.
Both Makeblock and Elecrow did whatever they could to mitigate risk by starting production through a Kickstarter campaign. They didn’t make these products “on spec.” They made these products knowing there was a market for them.
If you make designs for a client, you know you’ll get paid. But if you make things for an unknown customer somewhere in the world, you could end up wasting a great deal of valuable time, money, and mental health. Do whatever you can to learn from the successes or failures of others and build on their ideas.
Barbies aren’t just for girls, firetrucks aren’t just for boys, and children’s toys aren’t just for children. Inside every toy, there are hundreds of hours of engineering experience just waiting to be discovered. So next time you see an exciting Kickstarter campaign or a new product comes into your inbox, open it and explore it before you give it to your nephew or a local school.
OK, this is an interesting article. I took things apart to see how they worked and then always put them back together so I could have fun with them working. From that I also learned about the evils of “design for one-time-only assembly. The plastic tabs that are intended to break off when a product is opened, the welded cases that forbid repairs, and over-molded connectors on very flimsy cables.
I am not familiar with these products but they do sound interesting, and the connection scheme seems like it is very robust. The cheaper kit use snaps, which certainly work and are really cheap, but tend to be rather demanding in position accuracy, and also result in a lot of force applied to connections. But they are a whole lot cheaper. less than $1 per hundred in bulk quantities.