Thesis

Abstract

The thesis deals with the integration of conditional planning and robust execution in human-robot interaction. In this work I have developed a framework for the generation of plans that are robust and reliable with respect to the context, i.e. social interaction between human and robots. The architecture proposed operates on plans with an additional layer of representation that adds the complexity necessary to produce robust plans. With the proposed approach, we are able to provide plans that are more complex that the ones provided by any formalism/planner. The work developed in this thesis was tested during the Maker Faire 2016 (Rome) event, where the proposed approach has proven to be flexible and robust. Furthermore, this work is part of an ongoing collaboration with the Imperial College of London to develop a framework for general and robust planning in ROS.

The contribution produced in this thesis was submitted to the ICAPS 2017 conference and is actually under reviewing. The framework is released as Open Source Software and available on-line.

[UPDATE 2017.01.26] The work has been accepted for publication as a full paper at the International Conference on Automated Planning and Scheduling (ICAPS 2017).