{"id":13150,"date":"2013-03-12T12:42:15","date_gmt":"2013-03-12T16:42:15","guid":{"rendered":"http:\/\/sciencebusiness.technewslit.com\/?p=13150"},"modified":"2013-03-12T12:42:15","modified_gmt":"2013-03-12T16:42:15","slug":"robotics-cloud-software-platform-being-demonstrated","status":"publish","type":"post","link":"https:\/\/technewslit.com\/sciencebusiness\/?p=13150","title":{"rendered":"Robotics Cloud Software Platform Being Demonstrated"},"content":{"rendered":"
\"Robotic<\/a>
(NASA)<\/figcaption><\/figure>\n

Computer scientists and engineers at five European universities and the Dutch electronics company Phillips designed and are implementing an Internet-based robotics software repository<\/a> to relieve individual robots of heavy processing tasks. Workshops on cloud robotics, featuring hands-on practice with the RoboEarth system, are being offered next week at the euRobotics Forum<\/a> in Lyon, France.<\/p>\n

RoboEarth<\/a> offers open-source access to software for complex functions such as mapping, navigation, or processing of human voice commands that would otherwise need to be programmed into individual robotic devices. The platform also enables cloud-connected robots to access widely used database services such as those provided by Google and Amazon. Moving these shared heavy-duty processing tasks to the cloud, say the developers, makes it possible to develop simpler and less expensive robotic devices and autonomous vehicles, although client devices need high-bandwidth full duplex connections.<\/p>\n

The software components<\/a> in RoboEarth include a knowledge processing system called KnowRob<\/a> that provides a common semantic framework for integrating information from different sources. RoboEarth’s cloud engine is called Rapyuta<\/a>, from a robotic world portrayed in a Japanese film, that offers an open-source platform-as-a-service that connects client devices to the RoboEarth knowledge base. The software has as well an object detector<\/a> to capture and store three-dimensional object models with a Kinect camera to aid in object detection.<\/p>\n

The RoboEarth knowledge base<\/a> stores software components, locations and map models for navigation, task knowledge in the form of action recipes and manipulation strategies, and object recognition models. In addition to off-loading complex processing, the knowledge base makes it possible to share the experiences of client devices, enriching the store of knowledge over time.<\/p>\n

Four participants in RoboEarth from the Swiss Federal Institute of Technology<\/a> in Zurich will present a paper<\/a> on the Rapyuta platform at the IEEE International Conference on Robotics and Automation<\/a> in Karlsruhe, Germany, 6-10 May 2013.<\/p>\n

The following video gives an animated overview of Rapyuta and RoboEarth.<\/p>\n