2016 Optimization Days

HEC Montréal, Québec, Canada, 2 — 4 May 2016

Schedule Authors My Schedule

TB11 Planning, Optimization and Control in Mobile Robotics

May 3, 2016 03:30 PM – 05:10 PM

Location: Cogeco

Chaired by Jérôme Le Ny

4 Presentations

  • 03:30 PM - 03:55 PM

    Optimal robotic path planning with complex constraints

    • Frank Imeson, presenter, University of Waterloo

    In this talk we present the language SAT-TSP and the powerful solver, callback-TSP. The language allows a user to express single and multiple robot path planning problems with complex logic constraints. Callback-TSP operates by combining the DPLL(T) framework for solving SAT problems with incremental callbacks to a TSP solver.

  • 03:55 PM - 04:20 PM

    Collaborative control for a SmartWheeler

    • Mahmoud Ghorbel, presenter, Polytechnique Montréal

    The SmartWheeler project aims at developing an intelligent wheelchair that
    minimizes the effort required for steering and navigation. We have
    developed a collaborative control system that uses a Partially Observable
    Markov Decision Process to predict the user’s intended destination and
    optimize the control action. Preliminary results are presented from
    several indoor navigation tasks.

  • 04:20 PM - 04:45 PM

    Probabilistic path prediction and planning for 2D target tracking

    • Florian Shkurti, presenter, McGill

    We consider probabilistic pursuit of a moving target in 2D and the particular problem of probabilistic path prediction of the target's future actions, given limited information on their behavior and goals. We frame the tracking problem as a POMDP, approximately solved through sampling-based planning under topological constraints.

  • 04:45 PM - 05:10 PM

    Motion planning strategies for autonomously mapping 3D structures

    • Jérôme Le Ny, presenter, GERAD - Polytechnique Montréal
    • Manikandasriram Srinivasan Ramanagopal, IIT Madras

    We present a system capable of autonomously mapping the visible part of a bounded three dimensional structure using a mobile ground robot equipped with a depth sensor. The emphasis is on accurately reconstructing a 3D model of a structure of moderate size rather than mapping large open environments, with applications for example in architecture, construction and inspection.