c6jones720cybernetix

New Boards

2007-01-25T11:15:00.000-08:00

The servo motor and servo controller boards have now been manufactured for me. I collected them last week. Ive soldered the connectors onto them and I am now waiting for a some new PIC and Picaxe processors to arrive. As soon as this happens I will actually have everything I need to drive my robot joints.

The annoying thing is that while I was waiting for my boards to be manufactured, I figured out how to build the multiplexer using USB electronics. One USB multiplexer could have easily done the work of 8 or more RS232 multiplexers! Oh well, thats a job for another day...

Natural Language Processing

2006-12-31T08:47:00.000-08:00

I recently had a look at natural language processing and speech recognition. It makes for an interesting interface between people and machines. I had already experimented with voice control of robots but that was before I was able to program the Microsoft Speech SDK properly.

After doing some research I have come up with a possible template for a speech processing NLP module for conversing with and controlling a robot. Im hoping the system will be capable fo the following:

Distinguishing between conversation types, Question, Statement and Command
Acquiring semantic knowledge by parsing Statements.
Retrieving facts from semantic knowledge base by answering simple Questions
Performing Commands by comparing what was said against a list of acceptable commands
Asking a user for the definition of a word that does not appear in the lexicon and updating it

To do this Im planning to save semantic memory in the form of human readable files which can be parsed for content. Effectively you can tell the robot whatever you like, you can ask it a question about what you said later too.

For understanding the general gist of the conversation I am using an Augmented transition Network or ATN. This contains a finite state machine whoose states are triggered by the grammer of the word it is parsing. This method is quite fiddley and takes a lot of fine tuning, but you can usually make sense of the sentances spoken.

At this time Ive gotten as far as the ATN section. I'll get around to the rest of it when I get some time.

Neural Networks

2006-11-12T05:44:00.000-08:00

As well as working on my android I am also developing a general purpose vision system in my spare time. I wanted it to recognise objects by the way they look. I intend to use this mechanism in my robot for determining who people are by looking at thier faces.

I decided that I would try a nueral network just as an excersise to see how well suited neural networks are to recognising visual objects. I can tell you - they are pretty impressive!

I trained this network to recognise the numbers 0-9 visually by looking at the content of some bitmap files. All the bitmaps were the same size 80x60 pixels so I needed to give the network to have 4800 inputs and 10 outputs (one for each number). After a few nights of messing around with learning coefficients and training epochs I got the network to generate a really good set of outputs. I am seriously considering not programming my robot to walk and letting a neural net do the work for me.

Neural networks can do amazing things if you have the patience to sit there and train them!

Data Recovered

2006-10-29T14:33:00.000-08:00

I wrote a vision system making use of the VFW system to capture live images in windows before I learned to program Direct X. The hard disk the data was on decided to lose its FAT so I lost all my data. I had to leave my computer on for three days to recover all teh stuff on the broken hard disk but I got the bits I wanted back!

Here is a screenshot of the basic version 2 vision system just doing basic image processing. Im going to expand the basic engine to include Neural nets for face and object detection. I also intend to whack in some program code to deal with stereoscopic vision and also 2d to 3d conversion in the style of David Marr.

I also intend to add in the functionality to ascertain the colour of an given pixel so the name of the colour can be reported back. Ive got an algorithm that tells you the name of the closest colour match i.e. "dark blue" to any given RGB value. The table has over 700 colour definitions in it. Its has so many colour names that I had to resort to using excel to write the C program code for me!

A New Hope

2006-10-23T12:30:00.000-07:00

Maybe its too soon to tell but I got hold of some data recovery software and I think theres a chance that I will be able to get back all of the work I lost on my hard disk. If thats the case then building my robot vision AI and navigation systems will be a breeze. All I need to do is leave my PC on for the next 7 hours...

PCBs

2006-10-13T01:47:00.000-07:00

Here are the PCB layouts for the boards Im having made. Im expecting that the robot will have somewhere in the order of 30 servo motors and 5 to 6 controller units.

I mentioned before that the electronic design of my robot will be generic. Most robot parts (limbs and body) will simply consist of multiple servomotors and sensors connected to a single SSC and input sensor reader. Im going to need a USB hub or two to support that many peripheral devices.

I'll bet the guys that will assemble the boards for me must be rubbing their hands together!

New Designs

2006-10-11T05:33:00.000-07:00

Here are the designs Im having made for me. One is the Multiplexer/SSC and the other is a revision of my older dual servo. In theory they will work with almost any radio control equipment so possibly if people want them I will sell units on.

Current Status

2006-10-11T05:10:00.000-07:00

Ive been very busy recently! Ive let my projects slip a bit. Still thats not such a bad thing, Since my last blog Ive learned PIC Assembly langauge, Network Programming and the basics of Direct X 9.0 programming for C++. I already knew assembly language but didnt use it on my projects bacause it was so much cheaper and easier to use Picaxe Chips instead. Thats probably going to change because soon I will be ditching RS232 and using USB as the communication system for most of my projects.

Anyway I digress.. my Android project is back on and running at full steam! Whilst Im waiting for my electronic systems to be made for me im going to have another look at robot vision. This time Im going to go the whole hog and do it full colour stereoscopic. Ive seen program code for acquiring images with two Webcams simultaneously but im not sure If I want to do it that way at all. I waiting for two colour analogue cameras and I think I'll do a bit of hardware multiplexing with that..

Until then watch this space...

Modular design

2006-07-02T08:46:00.000-07:00

Ive opted for a completely modular design for this robot. To this end nearly every single component (mechanica,l software or electronic) is modular, generic and can also be used somewhere else in the robot.

Ive decided that electrically, every single limb on the robot should be completely self contained so that its only requirements are power and data. Mechanically most of the limbs will be the same because they are all made of the same basic components (when I get to the hands and head this may well change).
A block diagram for the system roughly follows that shown below:

The software to control the robot is already modular. Ive used C++ and written a class header with generic functions for robot limbs. OO languages are excellent for this sort of thing.

Robot Calibration

2006-07-02T06:57:00.000-07:00

I have just put the first servo onto my robot. I am literally starting fromthe bottom up so the first motors I wired up and the ankle Y and Z roll motors. I have designed and breadboarded my own 8-channel Serial servo Controller unit so up to 8 servos can be controlled at any one time. This will eventually be on the same board as the 16 channel multiplexer unit.

I built a hand calibration unit that can output a signal to a single servo motor. The point of this is that you can set a demand angle for the servo and read back (using a protactor) the angle that the servo shaft actually went to. This allows you to adjust feedback potentiometer alignments or gearboxes etc if need be.

I used this to correctly wire up the Y and Z roll joints on my robot ankle. Moving the hand calibration unit will now move the robot ankle.

Bilateral Servo control

2006-06-11T05:34:00.000-07:00

It is often easiest to implement a robotics project using open loop control systems. In these cases you can quickly make systems but once you have issued a command to the system, you have to assume that it has done its job because there is no way of monitoring it to find out for sure.

Using off the shelf serial servo controllers (SSC) is a bit like this. They are very good but how do you know the servo motor shaft is where you instructed it to go? Although servos themselves are closed loop feedback systems the SSC is not part of that feedback system, hence the overall system is open loop.

A bipedal robot is quite complex and Ive decided to make it completely closed loop so I can keep tabs on all real world variables at all times. To this end Ive made a 16 channel multiplexed servo position monitor. At any time you can know exactly what angle each servomotor shaft is at.

This system is truly bilateral since the monitor transmits the same data packets as an SSC would receive. This means that connecting the SSC input to the Monitor output for the same servo would actually result in either a positive or a negative closed feedback loop depending on which way you wired up the potentiometer.

My nice new house

2006-06-11T05:09:00.000-07:00

I havent bloged for a while because I've recently moved into a nice new house with my Girlfriend.

After settling in the first thing I did was take over ther larder and convert it into a cybernetics laboratory. Not sure if my girlfriend likes that or not - but tough!
Anyway here are a few pics of the new setup Ive got in my larder.

Ive set up a work bench with all my tools and made up a special rack for holding up my Bipedal robot. This stops it collapsing under its own weight when the motors are not engaged. It also allows me to test different walking gaits very quickly.

Lead by Nose Programming

2006-03-14T00:56:00.000-08:00

Here are a pair of 3-DOF legs. All I need to do now is add the servo controllers. I've made the limbs in such a way that it is possible to dissengage the motor and potentiometer just by removing two screws. Using this method it is possible to program the robot "Lead By Nose style". So in order to program my robot to perform a specific repeatable motion all I have to do is dissengage the motor and pull the joints to whatever position I like. The potentiometer is still engaged so it will track where I pull the joints to. I can then simply record which positions I like, re-engage the motor and play back the joint trajectories and Bob's your uncle!

Here are some pics of the two legs together..

High Power Servo Electronics

2006-03-14T00:24:00.000-08:00

Now Ive made a pair of 3-DOF legs I have only to attach them to each other. I am literally building this robot from the bottom up. I intend to use standard Picaxe 18 chips as positioning servos for each of the robot joints.

Here is the schematic in jpeg form. IC1 is a standard Picaxe 18 chip. I will use it to read in two streams of standard 1-18ms PWM from a servo controller and convert that into positional output for two motors.

Ive used this circuit on older robots and coupled with the Como 810:1 motors you have quite a powerful servo. In theory each servo will have a torque in the region of 3.3Nm at 12v over a travel range of about 180 degrees.

The H-bridge is only a bipolar one but should be good for about 60 watts. I might use MOSFETs later on depending on how well it performs.

Static Equilibrium On One Leg

2006-02-20T02:29:00.000-08:00

Here are some pics of the right robot leg (I havent made a left leg yet). It is pretty stable thanks to its oversized foot. The motors a powerfull enough to drive the joints to all positions without stalling (and that was only using 3volts!).

Its so stable that I'm pretty sure I could get something like this to take simple steps in an open loop feedback system without any proper controller. - I dont really intend to do this but I think it could work.

As you can see it is pretty close to my original drawing of the robot. I havent decided yet whats going to go above the waist but Im sure it will be a similar style to whats below it...

Thats just balancing statically and most of the joints are set to about 45' to the normal. The weight is noticably at the front of the foot (Where a big toe would be).

Robot Leg Subassemblies

2006-02-20T02:13:00.000-08:00

I decided to make this robot as easy as possible to construct and duplicate so I built it from easy subassemblies which all use common parts.

Each leg consists of three parts each containing two servomotors. These are shown below...

UPPER LEG (X-Z ROLL +-45')

MIDLEG (Y-Y ROLL +-45')

LOWER LEG (Y-Z ROLL +-45')

As you can see it really is simple to construct. All I need to do is put thses assemblies together and drop in a servocontroller circuit board...

New 3 DOF fully articulated leg

2006-02-20T01:01:00.000-08:00

This is my initial plan for the biped section of my new android. The construction of the real thing looks strikingly similar to my original idea.

I was thinking about using standard off the shelf parts like easily accessible gearmotors and boxes so the robot could be constructed easily cheaply and quickly.

The robot assembly has 3 degrees of freedom in each leg which means it can perform similar movements to a human leg.

Each joint uses a Como Drill 810:1 ratio motor for maximum torque for at low weight. The motors are coupled to potentiometers giving positional feedback. The whole thing is controlled by a Picaxe microcontroller run as a servocontroller.

Im planning to create a 3D computer model of the robot using the Direct X API so when I write the rest of the control software the robot will have a spatial model of itself. I plan to use Denavit Hartenburg 3Dimensional coordinates do the robot will know exactly (by measurement) where all of its joints are at any given time. See below

2006-02-09T05:18:00.000-08:00

I abandoned that project ages ago but recently I had the idea of building another android, but this time doing the legs differently. Previously I had used a double hinged mechanism with only 6 motors for both legs but this led to an underactuated design. It was also a bit to flexible too. Ive come up with a new simpler design using 6 motors per leg. And thats what I've come up with.... Im going to put in the other three motors on saturday.

2006-02-09T05:12:00.000-08:00

There he is sitting in my window sill waiting for some more rewiring..

2006-02-09T05:09:00.000-08:00

Once upon a time I thought you could get away without using microcontrollers in these sort of robots. I built this TTL multiplexer to control my robot. It interfaced to a PC using the parallel port. nice idea but you soon realise that serial/usb comms are far superior. Still it was pretty cool to see those LEDS flashing away inside the robots chest as it moved about.

2006-02-09T05:07:00.000-08:00

I went to Rajani's (a bargain basement in Bristol England) and they had these weird gyroscopic gamepads that they obviously couldnt sell. i bought one and took out the gyroscope (thats what that thing in the middle is). I was going to write software to use it as a roll/pitch attitude stabiliser to stop my robot from falling over.

2006-02-09T04:54:00.000-08:00

Then I realised that in order to walk usually you really need to add motors for shifting the balance left and right.

2006-02-09T04:53:00.000-08:00

Now I dont know about you guys but I thought it would be easy to just add the legs I'd made to the torso of my robot.

2006-02-09T04:50:00.000-08:00

I built clusters of servomotors together on one board. They are controlled by a standard Milford 8-channel serial servo controller. The servos used a PICAXE 18 (standard) microcontroller using a BASIC program I wrote.

2006-02-09T04:49:00.000-08:00

I built some legs for the robot. (These are the earliest biped legs I built). I built my own remote positioning servomotors using Como 810:1 drill motors and potentiometers. They are completely compatible with RC model servos but use a higher voltage.

2006-02-09T04:42:00.000-08:00

Theres the robot tired after a hard days work.Sitting on top of its charger.

2006-02-06T04:39:00.000-08:00

It all started with this guy.I thought it would be cool to build a robotic head that could see, hear, speak and perceive distance. So I built MR_HEAD.

Android

2006-02-06T01:58:00.000-08:00

A while ago I was watching TV and thinking about all the old fashioned portrayals of the future, they always had robots running around doing stuff. I got to thinking "Well its the 21st Century so where are all the robots?"

Then it struck me.. There were none! So I thought I'd better do something about the situation and took it upon myself to build some funky robots.

So here we are, a couple of robotics and AI projects Ive been tinkering with..