I have just finished a replica of R2-D2 and now I am working on a full size replica of C-3PO which will be updated as I go.
After some thought, I have finally decided to put Google Ads on the site. They should be contextual so you can click on them if you like, and it might cover my hosting bills 
NEW! There's now an XRobots forum for general discussion, or just to leave a message:
I have some of my previous walking robot and other videos on YouTube:
I sometimes have odd items for sale on eBay, or if you want something else let me know:
If you want to contact me my email address is simply: james 'at' the domain you are reading this.
As a natural progression, after building R2-D2, I have decided to build a replica of C-3PO. This will be life size and completely built from scratch, so I'll essentially be carving him out of wood, and then vacuum forming over all the pieces to make a prop or possibly a wearable suit... depending on how it turns out.
Newest updates will be at the bottom of the article as usual.
This project was last updated on 21st April '08
My new project is underway – to build an unknown amount of Astromech droids based on R2-D2 and friends.
I will be starting with one droid based on R2-D2, but as I develop a process to build parts along the way, I should have various spares and other things I find that ‘aren’t quite right’ to use on another one or two droids.
Check out the build progress pictures below which will be updated, newest at the bottom, as I buy or build parts.
Last updated: 3rd Feb '08 and 99% finished
Since my house burnt down and all the previous projects have been destroyed, I thought I should really start a new project.
I wanted to build something that was a bit more biologically inspired and hade more degrees of freedom than the other previous androids with their twin hinged parallelogram style legs. I also took a lot of influence from TheRobotStudio as you will see through the construction pictures.
What is the aim of the projects on this web site? 
My overall aim is to build a low cost solution using off-the-shelf and easily manufactured parts to solve the problem of mechanical biped gaits (two legs) for machines/robots. As many large companies develop and display very costly and complicated solutions, it seems sensible to develop something that is easy to manufacture, maintain and use.
Recently, Honda have shown us 'ASIMO' along with other large manufacturers such as Toyota. These are all biped solutions but have taken many years and millions to develop. More simple 'non-intelligent' low energy partially passive solutions have been developed by universities and other research institutions, although these are totally the opposite end of the scale and mainly only serve the purpose of demonstrating physics.
As I have mainly finished the mechanical contruction for Android 9 I thought I'd start to look at making it dynamically stable and document some basic testing along with my own thoughts / methods for making it work. 
I have no ankle motors for leaning sideways in the android so mechanically it will rely on end-stops at the ankles to stop it falling too far sideways. However, as it builds up momentum from leaning side-to-side in quick sucsession it will still be able to fall over. I'll be using cheap off the shelf gyro(s) for this project that are designed for use with R/C helecopters and the like. The aim will be to make it lean side to side in order to walk so it will have to take it's weight completely off one foot, but without tipping too far the other way. It will have to do this quickly so that it's gait flows, rather than stopping stationary at each extreme as with Android 3 which locked it's ankles in place to stand on one leg.
Although this isn't the 9th actual android I wanted to keep the numbering as it was, so this will be known as Android 9. I have decided to call him 'Featherweight' as the design will be based around light weight materials and low power actuators. 
I'll be using mainly balsa wood for this biped but I'm still aiming to make the legs about 50cm long so the whole thing comes to about 90/100cm (approx 3ft) tall. It will use some of the high torque linear actuators although I haven't seen them in real life yet so I've no idea if they're up to it. The plan is to move the leg joints as little as possible for walking in order to 'gear down' the actuators and increase power through levering.
I may still use aluminium tube for some of it but the main concept will be a balsa box construction with heavy card glued on with PVA to make the structure rigid. Anything that needs extra strength will be reinforced with 0.5mm aluminium sheet. All the joints will be made from 3mm steel rod rotating inside polythene bushes so I'm hoping that the whole construction will be light but relatively strong.
This robot is really an Idea that I have been thinking about for a while, but I don't know if I'll ever build it.
It's based around the idea of making a robot that's actually useful around the house or office, rather than just making a robot to that walks because it's hard/interesting to build and there aren't that many around.
Mainly I think that if a robot wil actually do anything to assist around any environment it should have some sort of manipulator. Even without any guidance system or even on-board power source, it could still be used for teleoperation in hazardous environments. However, there are many expensive commercial arms available, and also many cheap 'toy' like products, so this robot project will concentrate mainly on building a low cost but reliable arm mechanism on a mobile base.
I will be attempting to explain and draw the ideas for now as the robot only exists in my head so far...
This Article will hold videos of the android testing for all androids as I develop them. 
I have encoded the videos with the DivX codec so it isn't too large, DivX is available for free for many platforms including Windows, Linux, Mac etc. You can get the DivX codec from:
I now have some of my walking robot and other videos on YouTube:
I have had a mini-itx PC kicking around for a while that I use to play mp3s at parties. Now I have space in Mr Stick Legs for it to fit I will be putting a sensible operating system on it, making it fit in the smallest space possible, and attempting to reduce power consumption. 
The board is a mini-itx EPIA 800 which is only 17cm square, it has a fan for cooling but is very low power. It's only second from bottom of range - the lowest spec is the EPIA 5000 board which has no fan for cooling and runs at 500Mhz.
I have decided to build a new head project encompassing a pan and tilt camera along with stereo microphones for directional hearing. The whole thing will also be wireless so it can be attached to any android project in the future. 
I will be addressing the issue of getting the video and audio away to a remote computer for processing and also getting the control signal back to the device to move the camera etc. The remote computer could be replacd by a person for teleoperation, but I am also going to attempt to write some machine vision software to autonomously control the camera to follow movement or different coloured objects.
Oh well, here's another android. I shall still be working on number 6, but I've had so many thoughts about doing it differently again that I'm making another prototype. 
This android is kind of a cross between 3 and 6. It has quite similar legs to 3 but will use wiper motors in the body pulling cables to the leg joints so it can be more spindly than 6. This is why it's called 'Mr Stick Legs'.
I am trying to reduce the motor count from five to three for walking and also build in the ability to rotate the legs so it can walk around corners and turn on the spot. Because the legs don't have any motors mounted in them they are also able to bend further and so it should be able to go up steps/stairs. The body will also hinge at the waist so it can shift its centre of gravity forwards and backwards.
This is the sixth and hopefuly the most final android I will build. It will use the hip, foot and mostly the body structure from android 4, the control system will come from android 5. 
The legs will use wiper motors, while the arms will use the linear actuators.
I will also take parts from android 3 as it has some nice motors that can be used for the wrists and some control electronics.
This my fifth android project, although the forth android, Killertron is still a current project. This will leave me two current projects to work on, Killertron is pneumatic, while this project will be driven by electric actuators.
This android will be similar to Discotron and Killertron in that again, it will have legs with two parallel sections. However this time is will be more 'boxy' so that the mechanics are inside the legs.
The last Android, 'Discotron' works ok - it walks in a 'fairly' stable manner but was very slow. It is also 'statically stable' which means it puts all it's weight over one leg very slowly and can lift the other leg off the ground - then stop like that. When a person walks they don't lean over nearly as much, putting the opposite foot back down on the ground before falling. 
This is called being 'dynamically stable', or a 'dynamic gait'. In order to realise this in a walking machine, it must be able to move a lot quicker but still be strong enough to support it's weight on one ankle.
For this reason I have used pneumatics for this android, it will swing it's weight onto one side and step with the opposite leg a that exact time. The 'natural swing' of the android should enable it to walk like a human without leaning over nearly as far as it would to balance on one leg and stay there.
So far, the legs are the same length as mine...
Thought I would add this article which I will update when I find useful items at low cost.
With a wheeled robot the motors can simply be switched off and on to drive it around. With a legged robot it is slightly more difficult as each joint needs to be positoned accurately. 
This article is a brief description of how I am building a control system for Android 5, it will allow accurate positioning of each linear actuator from commands sent on a serial bus.
I needed to attach some postion sensors to the linear actuators I had made for Android 5 - Combotron.
These are sliding potentiometers which are in the style of faders on a mixing desk. the only way to mount them was with their PCB pins, so I decided to make some boards to fit them onto.
As with the android head project, when androids begin to interact with society they wil need some form of manipulators to carry out physical actions and to perform artificial intellegence testing. 
So, I decided to make some farily human looking hands (but with only three fingers) which will allow as much freedom as possible for a resonable / limited cost.
This article shows the steps I took to make a low cost linear actuator for Android 5 - 'Combotron'. Although it could have many other uses in robotic arms, legs or anywhere esle where 'piston' style movement is requierd at high torque and relatively low speed. 
When androids start to gain intellegence and interact with mainstream society, facial appearence and expression will become important. 
For this reason, as well as the blank and rather scary expression from Android 1, I have decided to make a head with capability for facial expression.
After much thought about making my last android stronger and more stable (it did just about hobble along but was a bit bendy), I decided to rethink the leg design and start yet again.
The concept of this android will be rather like those tool boxes with fold out trays that always stay level (a 'cantilever tool box'). The knee, hip and foot will all be held parallel by two shin and thigh parts hinged twice at each point. This means that only two motors are required per leg to move all three joints, one above and one below the knee. The android's upper half will also be held bolt upright, although the body may hinge in turn above the waist to allow it to balance backwards and forwards.
I've spent some time attempting to justify continuing with Android1 - 'Pinocchio', named due to it'd largely wooden appearance. However, it's just too big and heavy, it took one little step but then the other leg broke. 
So now I've decided to build a much lighter android with more joints like a human. I will be taking parts from the first android as I go to save money, but this one will be made from Aluminium, It should turn out around 5'5'' instead of 6'7'' like the last one, then it can ride in my car and fit through door ways.
I decided to make my own life size android, this was my first attempt...
So, my plan became to create something with the requirements as below:
To be near human sized.
To walk on two legs, be able to turn corners, maybe navigate stairs.
To have human features such as two arms and a head in the right place.
To speak in English.
To see with 'freaky eyes'.
Have a degree of intelligence in the future depending on how my software skills go.