Talks from IROS 2018
Presenters: Kenechukwu C. Mbanisi and Nathalie Majcherczyk
Kenechukwu C. Mbanisi – RBE PhD Student
Advisor: Jane Li
Learning Coordinated Vehicle Maneuver Motion
Primitives from Human Demonstration
Abstract: High-fidelity computational human models provide a safe and cost-efficient method for the study of driver experience in vehicle maneuvers and for the validation of vehicle design. Compared to passive human model, an active human model that can reproduce the decision-making, as well as vehicle maneuver motion planning and control will be able to support more realistic simulation of human-vehicle interaction. In this paper, we propose a integrated human-vehicle interaction simulation framework that can learn the motion primitives of vehicle maneuver motions from human drivers, and use them to compose natural and contextual driving motions in simulation. Specifically, we recruit seven experienced drivers and record their vehicle maneuver motions on fixed-base driving simulation testbed. We further segmented the collected data and classified them based on their similarity in joint coordination. Using a combination of imitation learning methods, we extracted the regularity and variability of vehicle maneuver motions across subjects, and learned the dynamic motion primitives that can be used for motion reproduction in simulation. Our research efforts lead to a motion primitive library that can be used for planning natural and contextual driver motion, and will be integrated with the driving decision-making, motion control, and vehicle dynamics in the proposed framework for simulating human-vehicle interaction.
Nathalie Majcherczyk – RBE PhD Student
Advisor: Carlo Pinciroli
Decentralized Connectivity-Preserving Deployment of Large-Scale Robot Swarms
Abstract: We present a decentralized and scalable approach for deployment of a robot swarm. Our approach tackles scenarios in which the swarm must reach multiple spatially distributed targets, and enforce the constraint that the robot network cannot be split. The basic idea behind our work is to construct a logical tree topology over the physical network formed by the robots. The logical tree acts as a backbone used by robots to enforce connectivity constraints. We study and compare two algorithms to form the logical tree: outwards and inwards. These algorithms differ in the order in which the robots join the tree: the outwards algorithm starts at the tree root and grows towards the targets, while the inwards algorithm proceeds in the opposite manner. Both algorithms perform periodic reconfiguration, to prevent suboptimal topologies from halting the growth of the tree. Our contributions are (i) The formulation of the two algorithms; (ii) A comparison of the algorithms in extensive physics-based simulations; (iii) A validation of our findings through real-robot experiments.
Friday, September 21, 2018
11:00 a.m. - 12:00 p.m.
60 Gateway Park, GP 1002