|Home Documentation Project Proposals Student Projects|
Download this document in Word Doc format
RESPONSPHERE YEAR 2 Progress Report
Current Reporting Year: October 1, 2005 - September 30, 2006
Project & Personnel Information
Please provide a list of the following EIGHT items:
1. Project Title:
2. Names of team members:
Javier Rodriguez Molina, Brian Braunstein, Anthony Nwokafor, Don Kimball, Boz Kamyabi.
3. List of Collaborators on Project:
Javier: hardware design, system integration.
Anthony: user interface, software
Brian: network, software
Don and Boz: Advisor and technical support
None at the moment.
5. List of Products created from this project:
An RC truck build from scratch with off the shelf components: motor, servos, speed controller, 3-channel transmitter and receiver, Basic stamp, distant sensor, GPS, micro serial servo controller, 12v 10Ah battery supply, 5 Amp RC switch, Ethernet webcam and a Meshnode (calmesh project) board configure to be a client.
-Gizmo User Interface:
A software which allows communication and control of Gizmo Truck via a webpage.
6. Significant Contribution to the body of Scientific Knowledge
Gizmo Truck project description:
Gizmo Truck is our solution to an easy and autonomous way to deploy the calMesh network in outdoor environments. The main goal for this project is to be able to send GPS final destination info to several Gizmo Trucks over the net, have them move to those locations autonomously and create the wireless network (calMesh project). The number and/or the type of features which can be added to these Gizmo Trucks depend purely on the applications they will be used for. In fact, Gizmo Truck can be seen as a platform on wheels to deploy different technologies, not only a wireless network. The basic features installed in a Gizmo Truck are:
-Full motor capabilities, forward, reverse, and break.
-Software and web user interface.
-Override circuit for manual remote control.
Where are we now:
We build the first Gizmo truck with all the basic features mentioned above except the Distance/Collision system. This feature has been designed and it is fully functional but it hasn't been integrated in the system (we want to upgrade the truck chassis before). Also, the Distance/Collision software needs to be developed further to be able to respond to different obstacles and environment situations. The main problem that we are facing at the moment is noise interference which causes malfunctions in the manual override system. We tried to reduce the noise with some simple circuitry but it showed no improvement. Next step will be to upgrade the transmitter and receiver (which we believe are the main cause of the noise) and to install filters to get rid of the noise component in the signals.
Where do we want to be in 3 months:
In three month we want to have:
- 5 functional and chassis-upgraded Gizmo Trucks.
- The web user interface upgraded, revised and supporting more features such as new ways to remotely control the truck's motor and steering, and more reliable video stream capabilities.
- Distance/Collision software adaptable to different environments.
Future ideas, applications, and projects:
Give these truck to people to use them to deploy their technologies:
-Computer Vision project:
Inside of buildings it can be difficult to use GPS to determine precise position for a robotic agent. Taking advantage of simple, user-provided, visual tags we propose to use computer vision to determine position information for the Gizmo truck. These tags can also contain other tags, information or instruction in them so that the truck can more robustly navigate its space.
In both inside or outside scenarios, the deployment of an homogenous and stable wireless network is crucial in emergency situations. Providing Gizmo Truck with netstumbler capabilities, a set of five trucks (with the appropriate learning and adaptive algorithms) could sense where the best location to deploy the network node is. Also, trucks could create signal strength mapping data that can help us to learn how to distribute wireless networks in a more efficient and successful way.