Knowledge and Learned Model Exchange Between Robotic Agents using openEASE

<p>In this proposal, I propose a solution for using openEASE as a knowledge bridge between robots. For this, I aim to implement a module in openEASE called openEASEclient (oeClient). This module will make robots connect to openEASE using web sockets and services without a need of web front-end. This module will, also, offer SWI-Prolog predicates as API so that robots that have KnowRob available can easily interface. After the implementation and testing of oeClient, I will try to exchange static ontologies between PR2s. Then, I conductsome grasp learning experimentusing EULisp. The data gained from these experiments will be used to train models using learning modules in KnowRob. After validation of learned models, these models will be transferred to PR2 in Bremen using oeClient. Finally, I will test the learned models with PR2 in Bremen and validate the success of the implemented methodology.</p>