Artificial Intelligence Wiki

The CRAM system consists of an extensive reasoning system and modules for interfacing the ROS system and the PR2 robot. Physics based simulation capabilities enable for foresight into the possible effects of actions. A plan execution layer allows for the execution of rather unspecific robot plans that are filled with information on the fly as necessary. The continuous work on the system is reflected in scientific contributions and papers.

A detailed description how to get CRAM running (tested on Ubuntu 12.04 with ROS Hydro) can be found here.

To automatically install ROS packages necessary for running CRAM components, we created an install script that should handle most of the process. You can find the instructions on how to run it here.

Introductory tutorials for CRAM in a simulated environment:

• Pick and Place Primer showing the basic usage of pick and place plans within CRAM, using a simulated PR2.

A summary of what CRAM is and how is can be used conceptually can be found on the ROS wiki pages:

• CRAM Core
• CRAM Language
• CRAM Reasoning
• CRAM Math
• CRAM Designators
• CRAM Process Modules

• https://github.com/cram-code: GitHub repository holding the latest version of the CRAM source code
• https://github.com/ai-seminar/cram-tutorials: GitHub repository holding tutorial related source code

For developers of CRAM we are assembling a list of useful information here.