Special issue on “Robotics: Science and Systems”, 2016

in keeping with the precedent set by the previous eight meetings, attracted a diverse range of outstanding papers from across the major disciplines in robotics. In this issue, you will find seven papers carefully selected from the RSS 2013 proceedings. These papers cover topics from semantic language processing and resource constrained communications to planning and optimization in control. These papers address a variety of challenging robotic tasks and show-case recent advances in the field. If you like what you read here then consider following up with the sibling RSS 2013 special issue being published by Autonomous Robots. In the first paper, Walter, Hemachandra, Homberg, Tellex and Teller examine the integration of natural language information with metric and topological maps to improve mapping performance. The authors describe their mapping approach and the techniques used for exploiting place descriptions and relationships provided by users. The method relies on semantic information provided by the user in order to identify loop closures and maintains a joint estimate over metric and semantic information. The technique can also make use of semantic information for areas of the map that are not currently in view and can even extend to areas of the map not yet visited by the robot. The authors demonstrate the utility of this approach on scenarios featuring increasingly complex interactions between the robot and users. In contrast to communication between robots and users, Walls and Eustice develop an approach to cooperative localization suitable for use with an unreliable and bandwidth-limited communication channel between a team of robots. In the proposed operating scenario, a single, well-instrumented server vehicle broadcasts navigation information to multiple client vehicles. The pose-graph information received by the clients is shown to allow all of the clients to reconstruct the centralized server–client estimate even in the presence of lost information. The method is demonstrated using deployments of multiple Autonomous Underwater Vehicles (AUVs) and a surface vessel. Majumdar, Vasudevan, Tobenkin and Tedrake present the first of the special issue papers on optimization in control. They explore feedback control design suitable for under-actuated, non-linear systems subject to constraints such as saturation limits. The method proposes the use of occupation measures to pose the problem as an infinite-dimensional linear program that can be approximated in finite dimensions using semi-definite programs. The work is contrasted against more traditional feedback lineariza-tion techniques and methods based on Lyapunov's criteria for stability. Validation of the proposed method is …