Blogs are Wrong

So, I built my first ladybugbot last night.  It costs me about $5 US to build one–~$.50 each for the motors, $1 each for the toothbrushes, $1 for the plastic/makerbot time, and the electronics are ‘free’, since I’m cannibalizing them, but I happen to know their factory cost is $.75.  It takes me about 15 minutes to solder and hot glue each one.

For a cute little robot that can scoot around my floor, this is pretty good.  Right now, though, it’s nothing more than a cute, R/C toy.  I’d like to do something more.

First of all, it bugged me that I had to do 15 minutes of assembly per bot, and that I had to print out a top and a bottom, a la last night’s design, http://www.thingiverse.com/thing:12975.  So, I made a single piece design that would print in 45 minutes, and should only take 5 minutes to assemble.  Here’s my v3 design on thingiverse:  http://www.thingiverse.com/thing:13004

So I started thinking about what a bunch of cute, cheap bots could do, and what I could put together in a day of soldering.  I knew that I could get CDS cells in Sham Shui Po, here in hong kong, so I thought I could make some ‘bots with like seeking/fearing behavior.  Wouldn’t it be cool to have a little swarm of bots that freak out and hide under the furniture when you turn on the light?  Or a solar powered bot that seeks out light?  This isn’t anything that BEAM roboticists haven’t done a million times before, but hey–everyone should have their own swarm, right?  In the future (I’m designing the board now), I can put a small AVR on board so I can start programming more interesting behavior.

So today, I fired up the makerbot and started cranking out the next iteration of bots.  For the first time, I tried the gcode copy-and-paste trick to make a bunch of objects in series using the automated build platform.  It was pretty good–I got 5/6 shells and one ugly plastic blob.  By the end of the afternoon, I had eight shells–my goal was 10 $5 ‘bots in the swarm, and I was feeling pretty good.

I started thinking a bit more about a simple little light chasing circuit, so I made this one with a dual comparator, some CDS cells, and two NFETS to drive the motors.  There’s nothing novel here–this is right out of what the frog’s eye tells the frog’s brain, but it’s fast and cheap, and that’s all I care about.  Video of the build coming tomorrow.

Well, I’m wicked proud of this ‘bot. Dig it:

This cute little bristle bot is about 2″ large, costs ~$5 to make, and I’m amazed at how well it zips around my kitchen floor.  It’s just two vibrating motors, two toothbrush heads, and a remote control pulled out of another robo-toy.  I really want to see a little swarm of these guys with a little bit of intelligence, even if it’s just light-seeking or light-fearing.

I prototyped this guy in four days in hong kong, starting from scratch with my first bristlebot:

Then moving on to an R/C, steerable ‘bot made entirely out of mangled toothbrushes and hot glue

And then moving on up to 3D printed bodies.  Square bodies don’t work so well, because the robot can only drive forward and turn, so it gets stuck.  My third iteration is the circular ladybugbot

Here’s some video of the ladybugbot being adorable:

So, that’s the ladybugbot.  It’s stunningly cheap and fast to build–the motors are fifty cents each, the bodies take about an hour to print out, and my latest and greatest prints come out of the makerbot complete with bristles and snap-mounts for the motors.  I’d like to end up with some $5 ‘bots that have some smarts–enough to do basic communication and navigation. The future is coming, and it’s small, round, and pink!

If you like it, make one for yourself!  You can use pretty much any vibrating motors you can get your hands on–I’ve been ripping mine out of cheap little vibrating toys, but there’s lots on ebay, in junk stores, and on hobbyist electronics sites like seeedstudios.com

I ripped the electronics out of a deskpet (http://www.mydeskpets.com/) to make my remote controller, and that gave me a cheap remote, motor controller and battery.

The 3D printed files are all up on thingiverse here:  http://www.thingiverse.com/thing:12975

And the toothbrushes you can get at your preferred toothbrush retail location.  Hold everything together with some dabs of hot glue, and away you go.  If you make one or improve on it, let me know–I’d love to see more little pink robots in the world.

photo credit:  Flickr usr bluman

I don’t like the last thing I wrote on Moleskine, don’t like it one bit.  There’s no moral to the story, just a grumpy old man kvetching about how he can’t get his favorite esoteric notebooks.  It’s time for a do-over.

Moleskine is giving us a glimpse of the future, a glimpse of what products and production and stock shelves will look like when it’s drop-dead-simple to make a new product.  The magic of Moleskine is in the production–they have a set of machines that can literally make anything within a certain class of product.

Take a look at this moleskin-esque notebook supplier in Guangzhou:  They have a production line that can make just about any style of notebook, and they specialize in custom orders and low volume production.  A notebook costs $.5, and they have a minimum order of 1000 notebooks.

http://www.alibaba.com/product-gs/448946289/OEM_high_quality_hardcover_notebook.html

So, with production like this, it costs $500 to go from a design to production.  They deliver seven days after you pay them.  That’s insanely accessible.  The barrier to entry is so low, it can’t even really be considered a barrier.  It takes far more  money and time to design the product than it does to bring it into production, but even the design effort isn’t a big deal.  If you’re a company like Moleskine, that has a brand and a way (web stores, retail shelves, etc) to put new products in front of people, why not experiment with a lot of new products?  The cost of designing, of experimenting, of trying out a new idea is so low (and Moleskine’s ~1000% profit margins on their $16 dollar notebooks are so high), that even if a new product sells terribly, it’ll probably break even.

What’s interesting here is that Moleskine’s retail presentation hasn’t caught up with the flexibility of its production.  They have 400 products that they’re actively selling, but you haven’t heard of most of them.  Their store shelves have no advertising, no way of telling consumers about a new product.  They just have an undifferentiated pile of products on a shelf.  Who knows, maybe that’s enough.  The bottom line, though, is that Moleskine is a business where the bottleneck isn’t the design or the production, and Moleskine is pushing it for all it’s worth, cranking out new, different products at a mind-boggling rate.

I have mixed emotions when I look at the Moleskine shelf.  On one hand, I see a pile of exceedingly similar products, where the product I love is buried somewhere in a pile of shit I don’t care about.  One the other hand, I see a product meritocracy; a world where the barrier to entry for a new product is almost nil, and it’s so easy and cheap to try out a new idea that it happens all the time.  Where the products people love rise to the top and the others fall to the wayside.  Where sales aren’t correlated to advertising efforts, but to the utility of the product.  The low barrier to entry does for journals what online blogging did for publishing–it’s cheap, it’s easy, and anyone who cares can contribute and get ideas out into the world.  There’s one thing the Moleskine shelf lacks, though, and that’s a way to search for the product you want in a field of similar, but inferior products.  It needs a search engine.

My Moleskine craze has just begun.  More on Moleskines soon.

photo credit:  flickr usr flyingturtle