Older blog entries for dafyddwalters (starting at number 1)

22 Jul 2003 (updated 5 Aug 2003 at 22:44 UTC) »

Open Automaton Project Update:

Work on the prototype droid continues. I've updated the web site to include a 'prototype' page, detailing some of my design decisions and thought processes in building the prototype, so I won't clutter this diary entry with those details.

On the development front, the Input Module, which is the hardware/firmware component that receives human input, now fully supports IR remote control, so I can now point my TV remote at the robot, and when I press buttons on the remote, the PC motherboard thinks that keys are being pressed on a keyboard. It occurred to me that this part of the project by itself may be of interest to some people even if they're not interested in building a robot. For example, if you're working on a project to make an MP3 jukebox or media device out of a PC, you could use my Input Module to control the device via an TV remote.

I've taken delivery of most of the major hardware components now. One component I was particularly keen to get my hands on was the motherboard, a Nehemiah M10000 EPIA M Mini-ITX 1GHz from VIA. When this arrived, I quickly got it set up on a bench with the DC-to-DC converter and 13AH sealed lead acid battery to take some power measurements. I was pleasantly surprised by what I found: The information published about the Morex DC-to-DC converter mentions that the AC adapter normally used to power the DC-to-DC converter (not used in my project) outputs up to 4.58A, so I was expecting to see a current draw from my battery in the order of 4A to 4.5A. However, what I found was that with a 40GB hard drive and a USB wireless 802.11 adapter plugged into the motherboard, the current draw from the 12V battery was just over 3A. This is good news in that it means that the robot is likely to be able to run for a couple of hours or more between charges.

While the motherboard was on the bench, I also took the opportunity to install the operating system (GNU/Linux) and the driver for the wireless network adapter. I plugged in a conventional ATX power supply to do this, although in fact, the operating system went on so smoothly and quickly, I could have quite honestly done it all on battery power :-)

So far, I'm very impressed with this little motherboard. It packs a lot of punch in a small, lightweight package, and has oodles of robot-friendly I/O ports, including I2C!

11 Jul 2003 (updated 30 Jul 2003 at 18:46 UTC) »

I've been quite busy on the Open Automaton Project lately.

I have just placed orders for most of the expensive off-the-shelf hardware components of the prototype droid, including the base, processor board, sensors (vision, ultrasonic, passive infrared). If you scroll to the bottom of this page, you can see a list of all the major hardware components that make up the robot, including how much I paid and the names of the vendors.

So far, I've developed some of the electronics and firmware that make up the human-robot interface, and I've got a breadboard prototype up and running. The input device on the robot itself is a simple six-button keypad (up, down, left, right, enter and escape), which allows rudimentary control such as program/task selection, and basic human responses to robot prompts (be they spoken or by text displayed on the LCD screen).

I've been spending quite a bit of time looking at existing publicly available open-source software projects of interest to the Open Automaton Project. A couple in particular are of great interest, and I believe will be an excellent fit. One is Player/Stage, and the other is Pyro. At the moment, my plan is to develop sensor and actuator software components that comply with the Player specification, and develop behaviours and high-level task programs using the Pyro framework.

I certified motters as a Master today since I've frequented his web site on numerous occasions in the past couple of years, and have always been impressed with the robotics work he's been doing. I'm particularly impressed with the work he's been doing recently in the area of robot vision, and since he has very generously seen fit to publish this work under the terms of the GNU General Public License, I will very likely be able to re-use some of this work in the Open Automaton Project.

Share this page