We recommend to use the RoboEarth platform at http://api.roboearth.org. This allows you to participate in a community which shares data about environments, actions and objects.

If you really need a private instance of RoboEarth follow the instructions below but be aware that you need at least some basic knowledge about the key components of the platform.

Preparation

To run your own RoboEarth platform you need at least one GNU/Linux system for productive usage a cluster is recommended.

Further you need to install:

• The Apache Web Server with mod-wsgi on one server

• The Django Web Framework (>= 1.2.3) with following modules on the web server:

  • Django Piston

  • HBase-Thrift module

• The Apache Hadoop Framework for distributed computing with following modules on all servers (Install instruction with pre-compiled Ubuntu, SuSE and RedHat packages: https://ccp.cloudera.com/display/DOC/Documentation/):

  • Hdfs
  • mountable hdfs
  • HBase
  • hbase-thrift.

• Sesame RDF Database on one server

Installing the RoboEarth platform

Once you have a running base system you can set up the RoboEarth platform

Preparing Apache Hadoop

Mount the distributed filesystem (HDFS) to a directory of your choice, e.g.

•  hadoop-fuse-dfs dfs://<name_node_hostname>:<namenode_port> <mount_point> -d 

Now switch to the directory of the HDFS and create the data directory. This is the place where all the binary files like object models will be stored.

•  mkdir data   chmod -R 666 data 

Link the data directory to your Apache web directory

•  ln -s <mount_point> /var/www/data 

To create the HBase tables run this script:

•  python createTables.py <host> <port> 

Preparing Sesame

Create the Sesame repositories "recipes", "objects" and "environment"

Setting up RoboEarth

• Download the RoboEarth Web Interface

• Extract it to /var/django
• Edit /var/django/roboearth/db/roboearth.py and adjust all the configuration parameter to match your domains, directories and computer names.