Team members: Freyr Magnússon
Team name: Team ROBOMASS
Project name: Skundar Hermar
Project tag line: A full spectrum simulation of a robot and it's environment

• Building a robot takes a lot of effort and is very dependent upon the hardware its built with. To try to make development easier we could use a broad spectrum simulation of the robot hardware and it's immediate environment. This will create a powerful testbed for modular development of utilities that help us make smarter, better and more productive robots.
• The simulated environment provides us with an interface that mimics the low-level controls we expect from the robot. It also supplies the means for our robot to sense it's surroundings so that it can take informed decisions.
• Realization of this simulation environment will enable future students to write modules that control and interact with the robot in a distributed manner. A wide array of features can then be assembled with our having to rely on the physical robot and can be replicated with ease (bytes are made for copying by the way).
• The ability to setup scenarios where all factors are controllable is very important. Also the ability to interact with and monitor sensors, actuators, postures other informations that has spatio-temporal meaning.

• By simulating motion physics and collisions we can move the robot in a realistic manner through our virtual world. The motion will be governed by the forces applied to actuators such as motorized wheels and joints. And rigid body physics will dictate the consequences of bumping into things.
• Ultrasonic sonars are simulated with geometric collision systems as well. This will enable collision avoidance algorithms to be developed along with environmental mapping.
• Visual imaging is possible using virtual cameras and rendering straight to memory.
• 3d audio may be used to simulate microphone input for sound detection and processing.
• An interface to write programs will be exposed to users. This will enable people to write their own modules that analyzes sensor output and provide control commands for the robot actuators.
• The simulation environment will provide display facilities so the user can observe the robot behavior with in the environment and provide means for visualization for sonar frustums and mapping overlays along with various other indicators.
• Provide a data driven framework for setting up environments and scenarios along with other simulated entities
• The simulation is connected to a distributes system that will provide the robot itself with scalability, modularity and extensibility

• Uses Ogre3d graphics engine to power the visualization along with physics and collisions through extensions.
• Will recreate the CADIA lab to use as a simulation environment.
• Will implement the lowest layers of the Skundar robot interface.
• Simulate sensors
  • Sonars
  • Position tracking (speed, direction, displacement)
  • Ranger camera
  • Joint positions from head unit
• Publishes sensor information using Psyclone whitebords

• Open Graphics Rendering Engine
• Open Dynamics Engine
• Presentation