IRIM News & Events
Robotics News Highlights
IROS 2020 Paperwatch!
October 25 - November 25 | On Demand - Online
Professor Frank Dellaert will present the Keynote Lecture for the "Air, Space and Sea Robotics" Section titled Perception in Aerial, Marine & Space Robotics: a Biased Outlook
Tutorials & Workshops
- Professors Jaydev Desai & Jun Ueda will present at the IROS Medical Robotics Workshop
- Professor Kyriakos Vamvoudakis will present the tutorial Using Reinforcement Learning to Solve Control Problems for Robotics
Georgia Tech Atlanta Submissions
- R. Connor Lawson, Linda Wills and Panagiotis Tsiotras, GPU Parallelization of Policy Iterationn RRT#. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020.
- Erickson Z., Xing E., Srirangam B., Chernova S., and Charles C. Kemp, Multimodal Material Classification for Robots using Spectroscopy and High Resolution Texture Imaging. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020.
- V. Aladele and S. Hutchinson, Collision reaction through internal stress loading in cooperative manipulation, IEEE/RSJ Inter-national Conference on Intelligent Robots and Systems (IROS), 2020.
Georgia Tech Lorraine Submissions
- Pradalier C., Ouabi O., Pomarede P., Steckel J. On-plate localization and mapping for an inspection robot using guided waves: a proof of concept. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020.
- Wu X., Vela P., Pradalier C. Robust Monocular Edge Visual Odometry through Coarse-to-Fine Data Association. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020.
IROS 2020 Is FREE and On Demand!
Register for Access Here
Speaker 1 | Topic: Coverage and Inspection Planning for Unmanned Aerial Vehicles
Kevin Yu | Virginia Polytechnic Institute and State University
Abstract: In this presentation, we investigate how to plan paths for Unmanned Aerial Vehicles (UAV) for coverage of an environment. I will present three increasingly complex coverage problems based on the environment. We start with a 2D point coverage problem where the UAV needs to visit a set of sites on the ground plane by flying at a fixed altitude above the ground. The UAV has limited battery capacity which may make it infeasible to visit all the points. We propose a novel symbiotic UAV and Unmanned Ground Vehicle (UGV) system where the UGV acts as a mobile recharging station. We present a practical, efficient algorithm for solving this problem using Generalized Traveling Salesperson Problem (GTSP) solver. We then extend this algorithm to covering 2D regions on the ground with UAVs that can operate in fixed-wing or multi-rotor modes. Finally, we propose to investigate a general version of the problem where the UAV is allowed to fly in full 3D space and the environment to be covered is in 3D as well. We propose an algorithm that clusters points in the free space to have a UAV autonomously plan online paths for bridge inspection. These online paths can be re-planned in real-time such that the UAV strives to obtain an optimal 3D coverage path.
Speaker 2 | Topic: Desensitization for Safe Planning under Parametric Uncertainties
Venkata Ramana Makkapati | Georgia Institute of Technology
Abstract: The tension between optimality and safety is often evident in robotics---particularly for applications that have stringent performance requirements---under conditions for which uncertainties in sensing, environment models, and control effectiveness are unavoidable. For all but the simplest applications, optimal solutions tend to bring the robot dangerously close to the operational safety margins. For example, it is well known that the shortest path for a mobile robot in a polygonal environment lies in the visibility graph which implies that the optimal path would contact the obstacles while traversing the path. While in practice it is typical to perturb paths slightly such that they do not reach the constraint boundaries, this safety strategy raises a number of significant questions: How should one perform these perturbations? How should one balance the cost of violating constraints against reduced performance? And, perhaps most importantly, how can one provide a principled evaluation of the effects of uncertainty with respect to the trade-offs between optimality and safety, and adjust the path to optimally balance between the two? It is this latter question that is addressed in the work.
The issue of safe optimal path planning under parametric uncertainties is addressed using a novel regularizer that allows trading off optimality with safety. The proposed regularizer leverages the notion that collisions may be modeled as constraint violations in an optimal control setting in order to produce open-loop trajectories with reduced risk of collisions. The risk of constraint violation is evaluated using a state-dependent relevance function and first-order variations in the constraint function with respect to parametric variations. The approach is generic and can be adapted to any optimal control formulation that deals with constraints under parametric uncertainty.