Arduino anyone?

[AZRedhawk44]

The time of Christmas gift ideas has come about, and during the various forms of familial cross interrogations that are typical of my clan, I expressed an interest in Arduino controllers and small scale robotics/automation toys.

Anyone an Arduino junky here? Or anything similar?

I have a list of 10 increasingly difficult projects I want to build, after which I'm sure I'll want to add more projects or find sub-dependencies to get to these 10.  Foremost of the subdependencies will be a RepRap machine (a plastic "printer") to fabricate various plastic parts for my robotic minions:

1.  Simple Arduino board with LED lights that turn on/off through a timed sequence
2.  Use lights or possibly a stepper motor position (as a form of analog dial) to reflect the output of sensors:  thermistors, proximity, circuit voltage, etc.
3.  Synchronize the rotational speed of two or more rotors
4.  Perform range estimation and triangulation with one transmitter (remote) and three receivers (on the Arduino)
5.  Robot 1 - Wheeled or possibly tracked.  Senses and climbs curbs or other obstacles.  Follows me on a walk around the neighborhood (by me carrying the transmitter and it tracking via receiver)
6.  Robot 2 - Helicopter.  Hover 3 feet off the ground (via downwards facing proximity sensor), balance (via gyro and accelerometer sensors), and perform a controlled landing autonomously
7.  Robot 2 - Hover 3 feet off the ground, follow a radio beacon(held by me as I walk around), monitor battery life and land autonomously when low on power
8.  Robot 2 - small solar panel recharger installed.  Digital altimeter circuit and GPS circuits installed.  Attain 100ft AGL, follow a programmed set of GPS coordinates, return home and recharge autonomously, then follow that route again and return home
9.  Robot 2 - Altitude test.  See how high it can go before it reaches a "point of no return" and have it return home safely with a log for maximum altitude on a full battery charge.  Wind may make this one tricky, and it may be a series of attempts, pushing my luck and tolerance for lost helicopter(s)
10.  Robot 2 - Pre-program a series of safe landing spots (roofs of grocery stores and such, probably) as GPS coordinates.  Send the helicopter on a multi-day expedition where it repeatedly lands and recharges itself.  Have it land, on float pontoons, in my parents' swimming pool in their backyard, 8 miles away

Intermediary projects may involve teasing the locals via LED light patterns at night and such.

I'm also toying with a wireless ethernet shield for it, and having a fly-by-wire IP-based drone that responds to either immediate commands via joystick, or to batch-style command scripts that can be sent via "private" (as private as wireless can be, anyways) network.

[RevDisk]

For anyone starting off:

http://www.adafruit.com/index.php?main_page=product_info&cPath=17&products_id=170
or
http://www.adafruit.com/index.php?main_page=product_info&cPath=17&products_id=68

then

http://arduino.cc/en/Tutorial/HomePage

[zahc]

I've done lots of things with Arduino. Automatic fish feeder, several digital light meters/darkroom tools, robotic camera platform, so on.

My biggest gripe with the arduino platform is that the Arduino make file is not intended to be seriously used, so if you want to use Arduino's excellent libraries, you are forced to write your code in that abomination of an IDE, instead of being able to write in whatever text editor you want.

[RevDisk]

1.  Simple Arduino board with LED lights that turn on/off through a timed sequence

That's pretty simple.  Plug into pin 13 and ground

http://arduino.cc/en/Tutorial/Blink


2.  Use lights or possibly a stepper motor position (as a form of analog dial) to reflect the output of
sensors:  thermistors, proximity, circuit voltage, etc.

http://arduino.cc/en/Tutorial/AnalogReadSerial
(redirect to your out pin rather than serial, same concept)


3.  Synchronize the rotational speed of two or more rotors

Same rotors?  Same output = same speed.  Or is there another concern?


4.  Perform range estimation and triangulation with one transmitter (remote) and three receivers (on the Arduino)

Hrm.  Fixed sensor bracket, I assume?   As you need to know the distance between the receivers to triangulate.  Then it's just signal strength and direction, then do the math.  Simple RDF?  Or are you doing sound, light, motion, etc?


5.  Robot 1 - Wheeled or possibly tracked.  Senses and climbs curbs or other obstacles.  Follows me on a walk around the neighborhood (by me carrying the transmitter and it tracking via receiver)

Build normal robot.  Put an IR blinker on your belt spelling out a specific pattern.  Have robot track to pattern.  Curb = impact sensors.   


6.  Robot 2 - Helicopter.  Hover 3 feet off the ground (via downwards facing proximity sensor), balance (via gyro and accelerometer sensors), and perform a controlled landing autonomously

Buy autopilot hardware.  Srly.  Same yourself some sanity.  They make them already for hobbyists.


7.  Robot 2 - Hover 3 feet off the ground, follow a radio beacon(held by me as I walk around), monitor battery life and land autonomously when low on power

Buy autopilot hardware and figure out the control code.

8.  Robot 2 - small solar panel recharger installed.  Digital altimeter circuit and GPS circuits installed.  Attain 100ft AGL, follow a programmed set of GPS coordinates, return home and recharge autonomously, then follow that route again and return home

Voltaic Systems for your solar panels.  Digital altimeter and GPS = Sparkfun. 


9.  Robot 2 - Altitude test.  See how high it can go before it reaches a "point of no return" and have it return home safely with a log for maximum altitude on a full battery charge.  Wind may make this one tricky, and it may be a series of attempts, pushing my luck and tolerance for lost helicopter(s)

...   Good luck.  Incorporate automated recovery system (ie, parachute)


10.  Robot 2 - Pre-program a series of safe landing spots (roofs of grocery stores and such, probably) as GPS coordinates.  Send the helicopter on a multi-day expedition where it repeatedly lands and recharges itself.  Have it land, on float pontoons, in my parents' swimming pool in their backyard, 8 miles away

...   Good luck, if you accomplish this, give me a call.  I'll have you an engineering slot at Sikorsky Innovations in under a week.

[AZRedhawk44]

3.  Synchronize the rotational speed of two or more rotors

Same rotors?  Same output = same speed.  Or is there another concern?


I figure that there has to be variations from one little hobby DC motor to the next.  I'll need a way to balance the rotational speed of them so they are all even.


5.  Robot 1 - Wheeled or possibly tracked.  Senses and climbs curbs or other obstacles.  Follows me on a walk around the neighborhood (by me carrying the transmitter and it tracking via receiver)

Build normal robot.  Put an IR blinker on your belt spelling out a specific pattern.  Have robot track to pattern.  Curb = impact sensors.   

Curb = proximity sensors, change in angle of treads or deployment of elevated tires that "lift" vehicle.  No IR sensor, I want it following a radio beacon because I later want to be able to over-ride a GPS plot via detection of an active radio signal, and "come to daddy."


6.  Robot 2 - Helicopter.  Hover 3 feet off the ground (via downwards facing proximity sensor), balance (via gyro and accelerometer sensors), and perform a controlled landing autonomously

Buy autopilot hardware.  Srly.  Same yourself some sanity.  They make them already for hobbyists.

I never much cared for sanity.  I'll give it a go.  I figure it's a function of keeping the accelerometer steady, the proximity sensor within "x" distance of my ideal altitude, and might not even require the gyro sensor.  Small DC components are relatively cheap.


7.  Robot 2 - Hover 3 feet off the ground, follow a radio beacon(held by me as I walk around), monitor battery life and land autonomously when low on power

Buy autopilot hardware and figure out the control code.

If I can get 6, then I can do 7.

8.  Robot 2 - small solar panel recharger installed.  Digital altimeter circuit and GPS circuits installed.  Attain 100ft AGL, follow a programmed set of GPS coordinates, return home and recharge autonomously, then follow that route again and return home

Voltaic Systems for your solar panels.  Digital altimeter and GPS = Sparkfun. 

Thanks for the lead.  I was having trouble figuring out what to use for an altimeter sensor.

9.  Robot 2 - Altitude test.  See how high it can go before it reaches a "point of no return" and have it return home safely with a log for maximum altitude on a full battery charge.  Wind may make this one tricky, and it may be a series of attempts, pushing my luck and tolerance for lost helicopter(s)

...   Good luck.  Incorporate automated recovery system (ie, parachute)


I figure I'll do 150 feet, then 250, then 350, and so on, evaluating remaining battery charge each time and upping the ceiling.  I'll be thrilled if I get to 1000 feet and back, and ecstatic if I get above 2000.  I kinda doubt it on a 1100 to 2000 mAh battery with 5-10 minutes flight life before power drain, though.  The good news is I can allow it to RAPIDLY lose altitude as long as it has power to stop itself in the last 100 feet or so before becoming piecesparts on the street.


10.  Robot 2 - Pre-program a series of safe landing spots (roofs of grocery stores and such, probably) as GPS coordinates.  Send the helicopter on a multi-day expedition where it repeatedly lands and recharges itself.  Have it land, on float pontoons, in my parents' swimming pool in their backyard, 8 miles away

...   Good luck, if you accomplish this, give me a call.  I'll have you an engineering slot at Sikorsky Innovations in under a week.

I'm actually doing this because I'd like to change career focuses.  And I find it interesting.  I've always been a bit of a DC electronics monkey, though AC stuff (and anything DC with meaningful Amps) have always scared me.  I'm tired of working on financial data and I'd like to transition to something a bit more real and tangible.

[AJ Dual]

You need to make a RepRap that can fabricate an Ardurino, which in turn makes more rep-raps.

[AZRedhawk44]

You need to make a RepRap that can fabricate an Ardurino, which in turn makes more rep-raps.

You josh, but there's talk of a RepRap prototype model with a print head capable of delivering sintered metal leads... Right now, the limitation on the current RepRap isn't the resolution.  It's the print head temperature and appropriate media to feed it.  Right now the limitation is various low temp polymers.

When it can feed silicon and copper at 1200x1200dpi, watch out.  While it's not microprocessor-small like Intel and AMD produce, that would be small enough and cheap/easy enough to make soldering irons a thing of the past.

[GigaBuist]

I've got one of their boards on my desk right now.  The first project is building a controller for an automated watering system that takes into account light conditions and temperature to determine when to start a watering cycle for a given crop.  It's more fun than it sounds. My last couple of nights were spent getting a .NET program talking to the thing and writing configuration parameters to the internal EEPROM for storage.

Fun stuff.

Later on I hope to build a device that can weigh my bed and detect when one of our dogs jumps on it.  Then it can yell at her for me.

I like the helicopter ideas. 

[AJ Dual]

You josh, but there's talk of a RepRap prototype model with a print head capable of delivering sintered metal leads... Right now, the limitation on the current RepRap isn't the resolution.  It's the print head temperature and appropriate media to feed it.  Right now the limitation is various low temp polymers.

When it can feed silicon and copper at 1200x1200dpi, watch out.  While it's not microprocessor-small like Intel and AMD produce, that would be small enough and cheap/easy enough to make soldering irons a thing of the past.

I know.

And "biological print heads" and other forms of matter too someday. You could print a gun, breakfast, or a new liver.

The post-scarcity economy where the only things of any value are basic raw materials and energy (with damn low profit margins) rare earths (better margins), and the very few things that matter-printing/stereolithography can't produce (high margins) will seem just damn strange to us now. (surgery*, personal services, music, art, hand-crafts, luxury/prestige items etc.)

And how do you pay for the high-margin complex products, or personal services (*assuming post-scarcity robotics aren't handling them) when everything else has almost zero value?

Worth finding out though.

[AZRedhawk44]

AJ, with on demand prototyping you still have design and assembly to take into account.

Just like the dominant OS software today is closed source, and dominant entertainment media is closed source (music and movies), plans for objects that rapid prototypers can produce will be sold at a particular cost.  I'd expect some sort of DRM-capability in there to "serial number" the design and even make it impossible to 3D scan the object and try to clone it, even bypassing the CAD plan.

Yes, RepRaps will be out there that people will self-support.  But those people are tinkerers at heart.

Commercial prototypers will be sold with the above mentioned limitations, for the great unwashed masses to use.

My prediction is that instead of paying $10 at Walmart for a widget, you'll end up at Walmart.com and buy the prototyper plans for the widget for $10, then still have the production costs of actually making the widget.  And you'll only be able to make 1 widget via that purchase unless you use "illegal" software like the current offerings for open source DVD playback or encryption cracking.

[RevDisk]

My prediction is that instead of paying $10 at Walmart for a widget, you'll end up at Walmart.com and buy the prototyper plans for the widget for $10, then still have the production costs of actually making the widget.  And you'll only be able to make 1 widget via that purchase unless you use "illegal" software like the current offerings for open source DVD playback or encryption cracking.

And no doubt there will be thousands of wonderful open source widget plans that no one will use despite being superior in most manners.  Only a handful of weirdos will use them.

Re dominant OS software, depends on your definition of "dominant".  Desktops?  Sure.  Embedded systems?   Dude, Windows CE is an oddity when it comes to handhelds or embedded systems.  Linux and derivatives are the overwhelming majority OS (excluding "OS" (ie just firmware) written specifically for that one device or cluster of devices) when it comes to stuff you see all the time but don't think about.  The microcontroller field has virtually no MS products, and less closed source apps every year.  This isn't done for philosophical reasons, just cost savings and efficiency gains.

Honestly, I think we're at least two decades from "rapid prototyping" machines everywhere.  It'll follow the normal progression.   R&D labs, university labs, DIY hobbyists, limited commercial success, long period of being awesome and no one knowing about it, killer app found, EVERYWHERE. 


Back to the original thread, sounds good.  Now let's see the build.   

[AZRedhawk44]

On another tangent...

Did anyone see Jay Leno's commercial prototyper?

In one demo of its capabilities, he did a 3D scan of a crescent wrench.

The machine made it, even with the adjustable claw and screw.  The screw turned.  It was a perfect, 3-piece copy of the crescent wrench, done in plastic, and built 100% assembled.  Just friggin' amazing.

http://www.popularmechanics.com/cars/jay-leno/technology/4320759

Video halfway down the page.

I won't be able to start my stuff until after Christmas.  Everyone in the family is on a "no buying stuff for yourself" prohibition in preparation for the gifting orgy that will assuredly come on December 25th.  Though, I did only ask for 1 Arduino and specific sensors for my robotic minions.  I'll need another Arduino and motors/sensors/etc to build a RepRap.  I could probably get started on that one before Christmas.  Hmmm.

[AJ Dual]

AJ, with on demand prototyping you still have design and assembly to take into account.

Just like the dominant OS software today is closed source, and dominant entertainment media is closed source (music and movies), plans for objects that rapid prototypers can produce will be sold at a particular cost.  I'd expect some sort of DRM-capability in there to "serial number" the design and even make it impossible to 3D scan the object and try to clone it, even bypassing the CAD plan.

Yes, RepRaps will be out there that people will self-support.  But those people are tinkerers at heart.

Commercial prototypers will be sold with the above mentioned limitations, for the great unwashed masses to use.

My prediction is that instead of paying $10 at Walmart for a widget, you'll end up at Walmart.com and buy the prototyper plans for the widget for $10, then still have the production costs of actually making the widget.  And you'll only be able to make 1 widget via that purchase unless you use "illegal" software like the current offerings for open source DVD playback or encryption cracking.

The current RepRap is not even at the phase when the Altair came out in terms of the personal computing revolution.

I'm talking what matter printing, and self-assembly will be like in 30 years from now, when most of us will still be alive. It's like trying to imagine today's internet in 1979.

As to design, yes, human talent will be required for some things. OTOH, opensourcing/crowdsourcing might make that free for many things too. People on forums might design consumer goods just for the prestige, the way that freeware games and utilities are now. (and Aurduino code is availible)

Or, you might see evolutionary design for some things. Humans need not design anything, they can test on-screen representations of the desired objects, select yes/no if it looks good enough, and the computer uses that input and churns out ever more refined prototypes. The few final variations can be printed for real-world acceptance testing.

http://www.nasa.gov/centers/ames/research/exploringtheuniverse/borg.html

[AZRedhawk44]

Does anyone on this forum have a RepRap Mendel or Darwin model, or know someone who has one?

I need the plastic parts fabricated.

[zahc]

Linux and derivatives are the overwhelming majority OS (excluding "OS" (ie just firmware) written specifically for that one device or cluster of devices) when it comes to stuff you see all the time but don't think about.

Last time I flew, I saw a hip-looking guy in DFW airport with an Apple laptop, Droid phone, and Kindle. I wonder if he even knows what linux/Unix is, realizes how all his favorite devices (and likely the airport router and servers of the websites he uses) all use it.

It may not be the year of Linux on the desktop, but the desktop is dying a death of a thousand cuts anyway.

[erictank]

Does anyone on this forum have a RepRap Mendel or Darwin model, or know someone who has one?

I need the plastic parts fabricated.

+1.  This looks like an awesomely-cool thing to get into.

If I get one before anyone else, I'll post here on APS - will print RepRap parts for members at cost+shipping.

Caveat - I have no idea when I'll actually be getting one for myself, so don't hold your breath or anything .

[AZRedhawk44]

If I get one before anyone else, I'll post here on APS - will print RepRap parts for members at cost+shipping.

I won't.

"Free as in Freedom, not Free as in Beer."

The parts take about 100 solid operating hours on a Mendel.

That's a long time to keep an extruded plastic print-head at 250 degrees C.  And to not have my RepRap making stuff I want it to make.  That's worth something.

I'll reflect a similar price as what I end up paying.

It's 100 operating hours of the machine, but it does not take 100 man-hours by any stretch.

I'm willing to pay $1 per operating hour, plus cost of plastic material, plus shipping.  I'm guessing there's about 1.5 to 2 pounds of ABS plastic consumed to make the parts.  I see ABS plastic wire going for about $15-$20 per pound.  And I'd extend that same price to anyone else that wanted RepRap parts.

If I find a better deal ($0.50/hr + costs) I'll certainly take it, and market forces will encourage me to lower my costs or simply not make RepRap parts beyond my conscientious "sure, I'll make one just to make sure someone else has a RepRap too" obligatory unit.

RepRaps are still rare commodities and finding one state-side is a difficult thing.  I would be happy to make parts for people to increase the number of RepRaps out there.  But I won't do it for free.  Anyone not willing to foot the cost of materials assembly isn't likely to actually follow through and make use of their machine to further reproduce RepRaps.

As such, I'm willing to motivate someone to make parts for me, by paying them $1/hr per machine-hour.

[CNYCacher]

Arduino is quite possibly the coolest thing I've come across.

For a guy like me who was messing with analog and digital circuits in his early teens, and also has been into programming forever, this is awesome!

I'll be buying my first board really soon.  Today possibly.

[AJ Dual]

When I can make an ersatz Magpul FMG for one of my Glock slides with the polymer a RepRap puts out, and register it as an SBR, I'll be interested. 

[CNYCacher]

For anyone starting off:

http://www.adafruit.com/index.php?main_page=product_info&cPath=17&products_id=170
or
http://www.adafruit.com/index.php?main_page=product_info&cPath=17&products_id=68

then

http://arduino.cc/en/Tutorial/HomePage

Just ordered the second one.

A friend of mine is also interested in this now, too.  I told him AZRedHawk44's idea of landing the copter in his parent's pool.  He has a pool.  He live 45.8 miles away. :)  We are going to try to get this done within 3 years.

[AZRedhawk44]

Just ordered the second one.

A friend of mine is also interested in this now, too.  I told him AZRedHawk44's idea of landing the copter in his parent's pool.  He has a pool.  He live 45.8 miles away. :)  We are going to try to get this done within 3 years.

LOL... it'll take 40-60 days of flight/charge sessions just to get there.  That's ~100 recharge landing sites you'll have to map and program into its memory.  And 100 times for it to get damaged on landing or stolen by someone who thinks it's super-cool and they should own one too.

I figure I'll be lucky to get 2 flight sessions a day, each of which get me half a mile.

[CNYCacher]

LOL... it'll take 40-60 days of flight/charge sessions just to get there.  That's ~100 recharge landing sites you'll have to map and program into its memory.  And 100 times for it to get damaged on landing or stolen by someone who thinks it's super-cool and they should own one too.

I figure I'll be lucky to get 2 flight sessions a day, each of which get me half a mile.

It's not my fault your helicopter sucks. :)

Hmmm. . . .  My parents house is 6 miles away. . . they only have a hot tub though.

[AZRedhawk44]

It's not my fault your helicopter sucks. :)

Hmmm. . . .  My parents house is 6 miles away. . . they only have a hot tub though.

You think > 1 mile flight distance is possible on a 2000mAh battery is possible?  I dunno.  I really don't.  I've never taken an RC craft for distance... I'm just thinking of 5-10 minutes reliable flight time before the rotors start to struggle to keep it in the air.  Assuming 10mph average speed, that gives between 0.8 and 1.6 miles traveled in that operational 5-10 minute window.  Add in altitude climb on take-off (200 feet sounds safe to me, to avoid buildings, trees, water towers and power lines), and controlled landing.  Add in time to fine-tune the GPS landing coordinates.  Add in time to fight a headwind.

I ordered a sanguino reprap mobo and some smaller components to start.

[CNYCacher]

You think > 1 mile flight distance is possible on a 2000mAh battery is possible?  I dunno.  I really don't.  I've never taken an RC craft for distance... I'm just thinking of 5-10 minutes reliable flight time before the rotors start to struggle to keep it in the air.  Assuming 10mph average speed, that gives between 0.8 and 1.6 miles traveled in that operational 5-10 minute window.  Add in altitude climb on take-off (200 feet sounds safe to me, to avoid buildings, trees, water towers and power lines), and controlled landing.  Add in time to fine-tune the GPS landing coordinates.  Add in time to fight a headwind.

Before anything I say is taken too seriously, I should probably make it clear that my experience with model aircraft of any type of power is precisely zero.

I really don't know how far anything should fly given a certain sized battery. . . although I had hoped it could be a little further than that

[RocketMan]

It may not be the year of Linux on the desktop, but the desktop is dying a death of a thousand cuts anyway.

I've heard that said repeatedly for at least the last fifteen years.  Still waiting...