I am contemplating buying a RoboSapien for fun and experimentation.
I have found a site that lists the IR codes for the robosapien:
And a Palm program that you can use to control it:
What I was wondering was, would it be possible to use a cheap and
cheerful USB IRDA adapter to send signals to it from a PC instead?
That way I could write an API to control the device and make it do
more interesting stuff!
Any thoughs would be greatly appreciated!
great idea! I've been planning to do something similar myself. In fact,
I've already bought the RoboSapien.
For the IR transmitter, I was planning to build a small circuit myself that
attaches to a single bit of the parallel port (or, alternatively, one of the
control lines of the serial port). The circuit would just be an 39.2kHz
oscillator driving an IR LED, and controlled by this port pin of the PC.
This reduces the real-time requirements of the software to a time resolution
of 1200Hz, which makes it possible to do it on a very low-end PC/laptop
or, alternatively, on a very advanced/inefficient operating system ;-)
IRDA is an interesting alternative, in that hardware is readily available
or even already built-in in many laptops and PDAs. From the standards
(at http://www.irda.org /) it seems that an IRDA transmitter does not
do any modulation itself, it's just on/off pulses. That gives a lot of
but requires that you do the 39.2kHz modulation in software. Not a big
deal for current computers, although under windows I'm not sure it will
be trivial either. The main worry with standard IRDA hardware is that
you need to somehow figure out how to control it from software.
One possibility would be to use the LIRC drivers, http://www.lirc.org /.
They support both self-built hardware and IRDA, and WLIRC
http://ruslat.hotbox.ru/WLiRC/WLiRC.htm allows driving a simple
LED+resistor from a serial port bit, even under windows. However,
both LIRC and WLIRC seem to be limited to Windows95/98 or NT4,
and the computer I am intending to run this on is running XP, which might
complicate things. Maybe I'll reconsider and use an old laptop.
Another thing that I'd like to do is to add some feedback. The Robosapien
is not exactly accurate as a feedforward robot, at least not if you are used
to thinking about stepper motors, servos and the like. Reproducability is
poor, even relatively simple things like walking in a straight line is not
depending on how the arms are positioned and what it is carrying. Therefore,
I am planning to use a simple webcam and some software, maybe in combination
with a simple light on the Robosapien to allow the PC to figure out where
Robosapien is, and in which direction it is moving. This should make it
to have the PC carry out more complex tasks, like sending the Robosapien to
a predefined location with some accuracy.
A final modification I am planning is a charging station. I want to make
strips beneath the "feet" connected to the battery compartments that are
located there, and a thin plate with metal springs that the Robosapien can
onto. By putting its feet on the springs, and having a charger connected to
springs, I can make sure the Robosapien will get charged. This will allow
to keep it awake indefinitely by sending at least once an hour or so a
the Robosapien that prevents it from switching itself off. For this to work,
the webcam thing to make sure the Robosapien can be reliably guided back to
the charging station.
Does any of this make sense? Any comments/ideas?
Some interesting ideas there. I went out and bought one on Saturday
and I am impressed with it so far. I am a programmer, and I don't know
much about electronics. I could build a circuit from a diagram at a
push, but thats about it.
I was thinking of using one of those USB-IR transmitters to control
robosapien such as a redrat:
http://www.redrat.co.uk/ - Its $99 though!
Or a TIRA
As regards feedback, I think its a great idea. I would like to use a
I am a lecturer in Dublin. I think I will try and get my college to
buy one and have a student do the work as a final year project!
My Robosapien does not seem to walk that well. In particular on a
wooden perfectly smooth floor, he seems to get stuck. He seems to
prefer carpet. Does yours do that?
$99 for an USB-IR transmitter is rather expensive, the WLIRC
transmitter consists of one resistor and one LED, which ought to
be a lot cheaper, and shouldn't be difficult to build at all ;-)
In the "Pictures" section of http://www.robosapien.tk/ there's a
set of pictures about building a wireless cam into the RS, including
type of camera etc. The receiver of such a cam could feed straight
into a TV card on the PC. I have some code in Delphi that will
fetch a picture from a TV card.
The idea about having a student work on it sounds good, that's a
fringe benefit of being a lecturer that I hadn't considered yet.
My RS walks well on our smooth wooden floor, although it is
not perfectly smooth any more (we had a rather large dog). When
it get's stuck, is that because the feet keep sliding, or do the motors
stop moving the legs/body?
Thanks for the pointer - I've found a similar mod. for the i-Cybie which
uses a MAX233 (doesn't need external capacitors) at
http://www.aibohack.com/icybie/sic_rs232.htm . I think a slight rewire of
this might do the job. All being well, I should then be able to take
control using an umbilical from a standard PC.
If the PC as a controller is viable, then the next step will be to get a
Compact-Flash capable SBC running a Linux kernel that the RoboSapien can
wear as a backpack. This would add plenty of opportunities for additional
Kinda busy till next week, expect some news then.
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.