[Michael] sent us a link to his RC robotic platform. He started with the same RC toy as the iPhone Robot (CAR) but ended up with a blank slate waiting for more features. What he has is an Arduino with a motor driver, three bump switches for the front and rear, and a XBee module. On the other side of things he used a SparkFun USB to FTDI connector to interface another XBee module with a PC in order to use Processing.
In the clip after the break you can see the motor control needs to have more fine-tuning done so the beast isn’t just out of control when running. But there’s a ton of potential here. It should not be a problem to add at lease rudimentary video feedback from the device. The Arduino is currently only being used marginally, leaving plenty of space to add on-board sensing such as IR, proximity, or light.
It’s a clean start, we expect to see updates!
15 thoughts on “RC Truck Source For Robotics Platform”
The XBEE data rate is so low I doubt video feedback would be manageable. A couple years ago I streamed low quality audio across it; I think that about maxed out the bandwidth.
Granted, 802.15.4 isn’t suited to such data rates, but it’s certainly something I’d expect to see posted on HAD: utilizing the full potential of the XBEE/802.15.4/ZigBee bandwidth.
An excellent, old school robotics platform is the HandyBoard from MIT. I know many colleges use (or used) these boards to teach intro to EE/CPE. I was so inspired I bought the blank PCB, then purchased all the components and made my own. It’s got lots (8 or more) ditital I/O, 6 or so analog inputs, motor drivers built in, and an IR link. Interfacing an XBEE shouldn’t be that difficult, but one would most likely have to bit-bang the serial interface (I don’t recall if the 68HC11 has a UART – I doubt it does).
He might not be able to send video over the XBEE but sending low quality/lossy video over a second medium (and just using the XBEE for control & input data — like he is currently) could work.
“It should not be a problem to add at lease rudimentary video feedback from the device.”
“lease” I think should be “least”/
This is very similar to the capstone engineering project I’m working on right now (done april 7th).
We are, in fact, using handyboards (not very handy, let me tell you that) to control a remote controlled car through a wireless mesh network created by XBees.
The HC11 does have a UART, but interactive C (the software that runs on the handyboard by default) takes control over it. You need to poke some bits that aren’t referenced in any manual to disable control by interactive C. After you do this, you can poll your uart or write some assembly routines to do it via interrupts.
He could have saved $ and space by using a Synapse RFE instead of Xbee+Arduino. He wouldn’t have been able to use Processing, but even better, he’d be programming in Python.
I’ve used Xbee for robotics. It often drops data after it has slept for a minute or two. I’m not a fan. See Synapse product demo from PyCon 2009: http://pycon.blip.tv/file/1947528/
Let the Arduino hate comments begin!
Personally I’d pipe the video over a 2.4Ghz link and leave the zigbee for remote signals.
I have no problems with arduinos, but I have dislike for that steering method on serious mobile robots.
For experimenting and such you might have to make do and I understand that.
However, if it’s really going off road then make it differential (tank-like) steering and avoid all the K-turns you are going to make (or not) in tight spots.
That drove me nuts with my own wheeled ROV so i ended up designing my own platform from scratch and ditching the Powerwheels car I started out with.
Good luck getting video over one of those XBee things. They’re slow as sin.
Also, if you just used an MC with a custom pcb(not even SMT necessarily) instead of that big honking arduino board, you could fit it on one of those charge-on-remote keychain-sized RC cars you can get anywhere for $5 to $20. Make a bunch of them and use a swarming algorithm with IR or wireless communication.
Perhaps even automate one of those $20 IR controlled helicopters…dammit, I know what I’m doing this weekend.
Toy RC cars make really crappy platforms; been there, done that. They tend to be geared way too fast (because kids want to race and crash them) and the turning is at best coarse, even if you replace the cheap toy mechanism with a real servo. For about the same bucks you can build a differentially steered trike with surplus windshield wiper motors for drive; it will be powerful enough to drive up a hill or plow through grass, precise, and capable of turning on a dime.
I second the suggestion of using a second one-way channel for video and letting the XBee or something like it handle controls. There is no way you will get useful video at any resolution (at least useful for something that’s supposed to be in motion) through the XBee.
HA XBee is slow for you, that why you lear analog stuff first and then digital, not other way around, couple transistors will be enought for good analog video transmitter
can i get a parts list for this project. i have an rc platform and an arduino but i would like to get the parts list.i sent a request to michael but no answer.
Interesting choice in a platform. I am curious how the suggested windshield wiper motor would work out, as suggested by “localroger”.
I wonder if there is a Traxxas RC that would make a better platform for you. You could probably find an old used model, like the Hawk or something, that you can still buy parts for. Much strnger than those toy rcs.
This is a great little starter. My carbot wall-racer is stll waiting for some competition so…
Please be kind and respectful to help make the comments section excellent. (Comment Policy)