Learning of grasp selection based on shape-templates

The ability to grasp unknown objects still remains an unsolved problem in the robotics community. One of the challenges is to choose an appropriate grasp configuration, i.e., the 6D pose of the hand relative to the object and its finger configuration. In this paper, we introduce an algorithm that is based on the assumption that similarly shaped objects can be grasped in a similar way. It is able to synthesize good grasp poses for unknown objects by finding the best matching object shape templates associated with previously demonstrated grasps. The grasp selection algorithm is able to improve over time by using the information of previous grasp attempts to adapt the ranking of the templates to new situations. We tested our approach on two different platforms, the Willow Garage PR2 and the Barrett WAM robot, which have very different hand kinematics. Furthermore, we compared our algorithm with other grasp planners and demonstrated its superior performance. The results presented in this paper show that the algorithm is able to find good grasp configurations for a large set of unknown objects from a relatively small set of demonstrations, and does improve its performance over time.     

Continuously Changing Information on a Global Scale and its Impact for the Internet-of-Things

This article analyzes the challenges of supporting continual changes of context information in Internet-of-Things applications. These applications require a constant flow of continuously changing information from sensor based sources in order to ensure a high quality-of-experience. However, an uncontrolled flow between sources and sinks on a global scale wastes resources, such as computational power, communication bandwidth, and battery time. In response to these challenges we present a general approach which focuses on four layers where we provide a proposed solution to each layer. We have realized the general model into a proof-of-concept implementation running on devices with limited resources, where we can moderate the information exchange based on relevance and sought after quality-of-experience by the applications. In conclusion, we evaluate our solution and present a summary of our experiences regarding the impact of continuously changing information on the Internet-of-Things.     

Algorithmic probabilistic game semantics

We present a detailed account of a translation from probabilistic call-by-value programs with procedures to Rabin’s probabilistic automata. The translation is fully abstract in that programs exhibit the same computational behaviour if and only if the corresponding automata are language-equivalent. Since probabilistic language equivalence is decidable, we can apply the translation to analyse the behaviour of probabilistic programs and protocols. We illustrate our approach on a number of case studies.