Software

TiroKart: the Total Immersion Remotely Operated Go-Kart
This is a specification for what I am trying to achieve with this project:

* Off-road and racing go-kart adapted for computer control.
* Ruggedized on-board PC cluster runs the kart control software and interfaces to the steering, throttle and brake actuators along with the cameras and other sensors.
* Sensors on the kart include stereo cameras, rear view cameras, stereo microphones, GPS, lidar (for collision detection) and telemetry for the kart data (speed etc).
* Wireless data link (802.11n) back to the remote PC cluster.
* Completely separate RC-based remote control for emergency shutdown.
* Remote PC cluster runs the driver interface and performs high-level processing of the data from the kart – especially intelligent collision avoidance.
* Driver interface consists of a 3D stereo display integrating the camera streams from the kart and also a presentation of the data from the kart sensors including audio from the on-board microphones.
* Control is either via a joystick or a steering wheel and pedal setup.
* Semi-autonomous mode implements driver aids – track departure prevention, collision avoidance etc.
* Fully autonomous mode utilizes onboard sensors (GPS, lidar, IMU and vision) to navigate a defined track autonomously.
* Rocamo (the Robotic Camera Operator) is a small mobile robot that autonomously tracks the TiroKart while in operation and provides a video stream from the perspective of a chase vehicle.
* The goal is to be able to drive the kart around as though the driver was sitting in the kart (in a perfect world). The icing on the cake would be to be able to drive a racing go-kart around a track faster than it could go with a person on-board (the TiroKart at least has a big weight advantage!) .

Finally, the software, electronic and mechanical technologies developed during the project will be applied to productized robots to solve selected real world problems.

Follow updates to this project via the development blog url.

Open Hardware Mobile Manipulator is a mobile manipulator robot designed as open hardware.
total build price under $1k
target audiences: (a) univ edu (1 unit/4 students; up to 10 groups/class); (b) maker/hobbyist
This is a design only — users (customize and) build their own from scratch

The project is to build an '''autonomous robot''', which plays football with golf balls on 2m x 3m arena by rules of Robotex . Robotex is an annual open Estonian competition of robotics mainly between three largest informatics related universities: University of Tartu, Tallinn University of Technology and Estonian IT College. The goal of the competition is to raise popularity of exact sciences and computer science. Our team is one of participating teams from University of Tartu. This year we are trying to build a smaller and faster robot than any of robots seen on this competition this far. We will design a robot which fits to cylinder of 18cm diameter with height of 15 cm, so it would also comply to size limits of Robocup SSL .

Features of our new robot include:
omnidirectional movement using brushless motors and self-designed omniwheels
positioning using magnetic hall sensors on wheels and data from image processing
localisation and mapping of target balls, goals and the opponent using self designed hyperbolic mirror based omnidirectional camera system

The objective of this project is to enable/run knoppix-adriane Linux Operating System on PandaBoard.
Adriane knoppix is a debian based linux operating system,with accellerated boot procedure: Independently from the usual SysV bootscripts, multiple tasks of system initialization are run in parallel, so that interactive desktops are reached very fast.

What's Adriane ?
ADRIANE is an easy-to-use, talking desktop system with optional support for braille, which can be used entirely without vision oriented output devices. Especially access to standard Internet services like email, surfing the web, scanning and reading of printed documents and using mobile phone extension services like SMS (over the users own mobile phone) are supported.

ADRIANE uses SBL (Screenreader for Blind Linux Users) from Marco Skambraks and Halim Sahin with Debian-integration by Klaus Knopper, speech-dispatcher/espeak as speech engine. Because of a python bridge, the screenreader also interoperates with orca, which is the graphical screenreader used together with compiz-fusions zoom functions after switching to graphical mode. The easy audio-desktop is being realized by dialog and bash-scripts in a consistent user interface, and is therefore very portable and easily extensible.

This is about a portable media player with the best balance between price and specs:
-Android 2.3.3 or better
-The device should be social-oriented ( custom and intuitive UI with facebook, twitter, myspace, etc) (something like the INQ smartphones but not facebook exclusive)
-Sound quality must be highest possible (licensing DTS-codecs?)
- Must have a "good" screen (640x360 OLED?)
- Minimalist-industrial design
- Gaming would be an strengh taking in account the 4430 SOC
- Other possibilities: PS certified, android market available or amazon android market, look for partners (amazon, spotify?)
-3G

The device will be oriented to the Latin-American Market and with some carrier partner (Claro and Movistar work on the whole region) with 3G-special plans (social-music streaming-instant messaging exclusive plans -VOIP, streaming videos and "complete" internet would be an extra or wifi only capabilities)

The ipod its the only media player sold in the region with quite a good a acceptance and here its sold at least with 30% over the US price.

Getting an aggressive price with a carrier partner would be a deal breaker on the region.
*Maybe prices like 150,165,180 and 270 USD or eq for 8,16,32 and 64GB but thats full speculation

The Alex platform is a small robot composed of the following parts:

iRobot Create base
2 arms made of Bioloid parts
2 eyes on pan tilt mounts made of Nexus One phones and Bioloid parts
A PandaBoard for central control

Alex is part of a larger project to advance artificial intelligence through embodied ai. Check out the project website for details.

Why Alex is awesome:

Willow Garage had the right idea with the PR2, but the wrong pricepoint. Alex will be the PR2 for the rest of us. We're targeting a pricepoint just north of 1k, as opposed to 400k + for the PR2.

There are thousands of robotics research labs reinventing robots that can move, see, hear, pick up things, make noise, and generally interact with their environment in useful ways. With Alex we plan to stop that cycle of reinvention, and start a cycle of innovation in software.

Why Alex will get done:

Over the last three months, I quit my job, completed an robotics internship at Anybots, and landed a new job in Artificial Intelligence powerhouse. I'm committed to spending significant personal resources on seeing Alex get to production quality and beyond. I've already committed thousands of dollars to support similar projects like the bilibot (bilibot.com). Eventually the Alex project will be a large consumer of PandaBoards, unfortunately you just can't buy one right now!

Alex and Panda:

The mobile robotics space is constrained by space, power, weight, and cpu. We can make use of the significant power available in mobile phones (see the cellbots project) but to provide a powerful central processing system that has room to grow you need a board with the feature set and efficiency of the PandaBoard. If we want to achieve untethered operation there's no better solution than Panda.

(This is the Pandacomb project submission with the official Pandacomb website link.)
Android, as we all know, is everywhere nowadays! Its on our smartphones, media players, netbooks, and tablets. But the problem is that these devices are overly priced and therefore not many people can enjoy the features of the amazing Operating System that is Android. The creators of the Pandaboard have had one goal since its beginning, to bring a fast platform with an affordable price point to the general public. This is where this project comes in. My team and I will try to bring Android 3.0 Honeycomb to the Pandaboard with a working multi-touch screen, drivers, and sensors. In turn our goal is to make a tablet with Android Honeycomb for as cheap as possible but to also use all the great hardware of the Pandaboard to it's full potential.

Kozio, Inc. (www.kozio.com) has made available free use of their kDiagnostics® Suite, for owners of PandaBoard. The Kozio download gives users an interactive guided tour of the inner workings of PandaBoard, along with a full design verification and hardware validation solution. The download is immediately available directly from the Kozio website http://www.kozio.com/services-and-supports/downloads. Kozio also provides the same expert solution set for customers designing custom devices.

This project is to port honeycomb onto the panda, use hardware acceleration, have a capacitive touch screen, and make the whole project run on recharcgable batterie. What do you get? A powerful dual core tablet. Doesn't it just sound nice.
Email: zaidhorse@gmail.com

We are going to attempt to connect and control various sensors and components to the PandaBoard. The component sensors will be a two-directional gyroscope (ADG-500) along with its on-board temperature sensor, BlinkM - I2C Controlled RGB LED(s), an accelerometer device (ADXL-300). We will be using an ADS1100 A/D converter for for our sensor signal conversion.

All these devices will be inserted into a custom made device fabricated using an additive manufacturing process called stereolithography which will contain our circuit. This project is designed to familiarize us with the process of controlling alternate devices using the PandaBoard. Any feedback or suggestions are welcome.