research.html

---
layout: default
title: Research
permalink: /research/
---
<h1>
  Research
</h1>


<small> Dissertation </small>
<h4>
  <i>Planning and Control for Hybrid Locomotion of Wheeled-Legged Robots</i>
  <a href=https://youtu.be/b_ey3hXAcfU>(Reuters Press Release)</a>
  <a href=https://www.research-collection.ethz.ch/handle/20.500.11850/515694>(PhD.pdf)</a>

</h4>
<div class="container">
  <iframe src="https://www.youtube.com/embed/39rRhTqcQc0"
          frameborder="0" allowfullscreen class="video"></iframe>
</div>
<div style="padding-top:10px"></div>
<p style="text-align:justify">
  The research community in legged robotics focuses on bio-inspired robots,
  although there are some human inventions that nature could not recreate.
  One of the most significant examples is the wheel that has made our
  transportation system more efficient and faster, especially in urban environments.
  Inspired by this human-made evolution, we developed the wheeled-legged robot
  ANYmal with non-steerable wheels attached to its legs, allowing the robot to
  be efficient on flat as well as agile on challenging terrain.
</p>

<p style="text-align:justify">
  This dissertation describes an optimization-based framework to perform
  complex and dynamic locomotion strategies for robots with legs and wheels.
  The proposed method allows to perform novel maneuvers, which exploit the
  wheeled-legged robot's full capabilities over challenging obstacles. By
  combining innovative techniques in motion control and planning, this work
  reveals the full potential of wheeled-legged robots and their superiority
  compared to their legged counterparts. This novel platform, with powered wheels,
  achieves a speed of 4 m/s on flat terrain, overcomes challenging obstacles
  with 1.5 m/s, and reduces the cost of transport by 83 % compared to
  legged systems. The work in this dissertation is published
  in two conference proceedings and three journal articles.
</p>

<small> March, 2021 &nbsp&middot&nbsp IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) </small>
<h4>
  <i>Whole-Body MPC and Online Gait Sequence Generation for Wheeled-Legged Robots</i>
  <a href=/publications/files/2021_iros_bjelonic.pdf>(IROS2021.pdf)</a>
    <a href=https://youtu.be/tf_twcbF4P4>(ICRA2020-Presentation.mp4)</a>
</h4>
<div class="container">
  <iframe src="https://www.youtube.com/embed/_rPvKlvyw2w"
          frameborder="0" allowfullscreen class="video"></iframe>
</div>
<div style="padding-top:10px"></div>
<p style="text-align:justify">
  The additional degrees of freedom and missing counterparts in nature make
  designing locomotion capabilities for wheeled-legged robots more challenging.
  We propose a whole-body model predictive controller as a single task
  formulation that simultaneously optimizes wheel and torso motions. Due to
  the real-time joint velocity and ground reaction force optimization based on
  a kinodynamic model, our approach accurately captures the real robot's
  dynamics and automatically discovers complex and dynamic motions cumbersome
  to hand-craft through heuristics. Thanks to the single set of parameters for
  all behaviors, whole-body optimization makes online gait sequence adaptation
  possible. Aperiodic gait sequences are automatically found through kinematic
  leg utilities without the need for predefined contact and lift-off timings.
  Also, this enables us to reduce the cost of transport of wheeled-legged
  robots significantly. Our experiments demonstrate highly dynamic motions on a
  quadrupedal robot with non-steerable wheels in challenging indoor and outdoor
  environments. Herewith, we verify that a single task formulation is key to
  reveal the full potential of wheeled-legged robots.
</p>
<div style="padding-top:30px"></div>

<small> March, 2020 &nbsp&middot&nbsp IEEE Robotic and Automation Letters (RA-L) </small>
<h4>
  <i>Rolling in the Deep – Hybrid Locomotion for Wheeled-Legged Robots using Online Trajectory Optimization</i>
  <a href=/publications/files/2020_ral_bjelonic.pdf>(RA-L2020.pdf)</a>
</h4>
<div class="container">
  <iframe src="https://www.youtube.com/embed/ukY0vyM-yfY"
          frameborder="0" allowfullscreen class="video"></iframe>
</div>
<div style="padding-top:10px"></div>
<div class="container">
  <iframe src="https://www.youtube.com/embed/tf_twcbF4P4"
          frameborder="0" allowfullscreen class="video"></iframe>
</div>
<div style="padding-top:10px"></div>
<p style="text-align:justify">
  Wheeled-legged robots have the potential for highly agile and versatile
  locomotion. The combination of legs and wheels might be a solution for any
  real-world application requiring rapid, and long-distance mobility skills on
  challenging terrain. In this paper, we present an online trajectory
  optimization framework for wheeled quadrupedal robots capable of executing
  hybrid walking-driving locomotion strategies. By breaking down the optimization
  problem into a wheel and base trajectory planning, locomotion planning for high
  dimensional wheeled-legged robots becomes more tractable, can be solved in
  real-time on-board in a model predictive control fashion, and becomes robust
  against unpredicted disturbances. The reference motions are tracked by a
  hierarchical whole-body controller that sends torque commands to the robot.
  Our approach is verified on a quadrupedal robot that is fully
  torque-controlled, including the non-steerable wheels attached to its legs. The
  robot performs hybrid locomotion with different gait sequences on flat and
  rough terrain. In addition, we validated the robotic platform at the Defense
  Advanced Research Projects Agency (DARPA) Subterranean Challenge, where the
  robot rapidly maps, navigates, and explores dynamic underground environments.
</p>
<div style="padding-top:30px"></div>

<small> April, 2020 &nbsp&middot&nbsp IEEE Robotic and Automation Letters (RA-L) </small>
<h4>
  <i>Trajectory Optimization for Wheeled-Legged Quadrupedal Robots Driving in Challenging Terrain</i>
  <a href=/publications/files/2020_ral_medeiros.pdf>(RA-L2020.pdf)</a>
</h4>
<div class="container">
  <iframe src="https://www.youtube.com/embed/DlJGFhGS3HM"
          frameborder="0" allowfullscreen class="video"></iframe>
</div>
<div style="padding-top:10px"></div>
<p style="text-align:justify">
  Wheeled-legged robots are an attractive solution for versatile locomotion in
  challenging terrain. They combine the speed and efficiency of wheels with the
  ability of legs to traverse challenging terrain. In this paper, we present a
  trajectory optimization formulation for wheeled-legged robots that optimizes
  over the base and wheels' positions and forces and takes into account the
  terrain information while computing the plans. This enables us to find optimal
  driving motions over challenging terrain. The robot is modeled as a single
  rigid-body, which allows us to plan complex motions and still keep a low
  computational complexity to solve the optimization quickly. The terrain map,
  together with the use of a stability constraint, allows the optimizer to
  generate feasible motions that cannot be discovered without taking the terrain
  information into account. The optimization is formulated as a Nonlinear Programming problem
  and the reference motions are tracked by a hierarchical whole-body controller
  that computes the torque actuation commands for the robot. The trajectories
  have been experimentally verified on quadrupedal robot ANYmal equipped with
  non-steerable torque-controlled wheels. Our trajectory optimization framework
  enables wheeled quadrupedal robots to drive over challenging terrain, e.g.,
  steps, slopes, stairs, while negotiating these obstacles with dynamic motions.
</p>
<div style="padding-top:30px"></div>

<small> January, 2019 &nbsp&middot&nbsp IEEE Robotic and Automation Letters (RA-L) </small>
<h4>
  <i>Keep Rollin’ – Whole-Body Motion Control and Planning for Wheeled Quadrupedal Robots</i>
  <a href=/publications/files/2019_ral_bjelonic.pdf>(RA-L2019.pdf)</a>
  <a href=/publications/files/2019_ral_bjelonic_poster.pdf>(Poster.pdf)</a>
</h4>
<div class="container">
  <iframe src="https://www.youtube.com/embed/nGLUsyx9Vvc"
          frameborder="0" allowfullscreen class="video"></iframe>
</div>
<div style="padding-top:10px"></div>
<p style="text-align:justify">
  We show dynamic locomotion strategies for wheeled quadrupedal robots which
  combine the advantages of walking and driving. The developed optimization
  framework tightly integrates the additional degrees of freedom introduced by
  the wheels. Our approach relies on a zero-moment point based motion optimization
  which continuously updates reference trajectories. The reference motions are
  tracked by a hierarchical whole-body controller which optimizes the generalized
  accelerations and contact forces by solving a sequence of prioritized tasks
  including the nonholonomic rolling constraints. Our approach has been tested on
  the torque-controlled robot ANYmal equipped with non-steerable,
  torque-controlled wheels. We conducted experiments on flat, inclined and rough
  terrain, whereby we show that integrating the wheels into the motion control
  and planning framework results in intuitive motion trajectories, which enable
  more robust and dynamic locomotion compared to other wheeled-legged robots.
  Moreover, with a speed of 4 m/s and a reduction of the cost of transport
  by 83 % we prove the superiority of wheeled-legged robots compared to their
  legged counterparts.
</p>
<div style="padding-top:30px"></div>

<small> January, 2019 &nbsp&middot&nbsp IEEE Robotic and Automation Letters (RA-L) </small>
<h4>
  <i>Trajectory Optimization for Wheeled-Legged Quadrupedal Robots using Linearized ZMP Constraints</i>
  <a href=/publications/files/2019_ral_de_viragh.pdf>(RA-L2019.pdf)</a>
</h4>
<div class="container">
  <iframe src="https://www.youtube.com/embed/I1aTCTc0J4U"
          frameborder="0" allowfullscreen class="video"></iframe>
</div>
<div style="padding-top:10px"></div>
<p style="text-align:justify">
  We present a trajectory optimizer for quadrupedal
  robots with actuated wheels. By solving for angular, vertical, and
  planar components of the base and feet trajectories in a cascaded
  fashion and by introducing a novel linear formulation of the zero-
  moment point (ZMP) balance criterion, we rely on quadratic
  programming only, thereby eliminating the need for nonlinear
  optimization routines. Yet, even for gaits containing full flight
  phases, we are able to generate trajectories for executing complex
  motions that involve simultaneous driving, walking, and turning.
  We verified our approach in simulations of the quadrupedal
  robot ANYmal equipped with wheels, where we are able to run
  the proposed trajectory optimizer at 50 Hz. To the best of our
  knowledge, this is the first time that such dynamic motions are
  demonstrated for wheeled-legged quadrupedal robots using an
  online motion planner.
</p>
<div style="padding-top:30px"></div>

<small> March, 2018 &nbsp&middot&nbsp IEEE/RSJ International Conference on Intelligent Robots and Systems
  (IROS) </small>
<h4>
  <i>Skating with a force controlled quadrupedal robot</i>
  <a href=/publications/files/2018_iros_bjelonic.pdf>(IROS2018.pdf)</a>
</h4>
<div class="container">
  <iframe src="https://www.youtube.com/embed/fJfAWiylpxw"
          frameborder="0" allowfullscreen class="video"></iframe>
</div>
<div style="padding-top:10px"></div>
<p style="text-align:justify">
  Traditional legged robots are capable of traversing challenging terrain, but
  lack of energy efficiency when compared to wheeled systems operating on flat
  environments. The combination of both locomotion domains overcomes the trade-off
  between mobility and efficiency. Therefore, this paper presents a novel motion
  planner and controller which together enable a legged robot equipped with skates
  to perform skating maneuvers. These are achieved by an appropriate combination
  of planned reaction forces and gliding motions. Our novel motion controller
  formulates a Virtual Model Controller and an optimal contact force distribution
  which takes into account the nonholonomic constraints introduced by the skates.
  This approach has been tested on the torque-controllable robot ANYmal equipped
  with passive wheels and ice skates as end-effectors. We conducted experiments
  on flat and inclined terrain, whereby we show that skating motions reduces the
  cost of transport by up to 80 % with respect to traditional walking gaits.
</p>
<div style="padding-top:30px"></div>

<small> November, 2015 - September, 2016 &nbsp&middot&nbsp Master thesis </small>
<h4>
  <i>Weaver: Hexapod robot for autonomous navigation on unstructured terrain</i>
  <a href=/publications/files/2019_ral_buchanan.pdf>(RA-L2019.pdf)</a>
  <a href=/publications/files/2018_jfr_bjelonic.pdf>(JFR2018.pdf)</a>
  <a href=/publications/files/2017_icra_bjelonic.pdf>(ICRA2017.pdf)</a>
  <a href=/publications/files/2016_iser_homberger.pdf>(ISER2016.pdf)</a>
  <a href=/publications/files/2016_iros_bjelonic.pdf>(IROS2016.pdf)</a>
</h4>
<div class="container">
  <iframe src="https://www.youtube.com/embed/eLMUiX96En0"
          frameborder="0" allowfullscreen class="video"></iframe>
</div>
<div style="padding-top:10px"></div>
<p style="text-align:justify">
  Legged robots are an efficient alternative for navigation in challenging
  terrain. In this paper we describe Weaver, a six‐legged robot that is designed
  to perform autonomous navigation in unstructured terrain. It uses stereo vision
  and proprioceptive sensing based terrain perception for adaptive control while
  using visual‐inertial odometry for autonomous waypoint‐based navigation.
  Terrain perception generates a minimal representation of the traversed
  environment in terms of roughness and step height. This reduces the complexity
  of the terrain model significantly, enabling the robot to feed back information
  about the environment into its controller. Furthermore, we combine exteroceptive
  and proprioceptive sensing to enhance the terrain perception capabilities,
  especially in situations in which the stereo camera is not able to generate an
  accurate representation of the environment. The adaptation approach described
  also exploits the unique properties of legged robots by adapting the virtual
  stiffness, stride frequency, and stride height. Weaver's unique leg design with
  five joints per leg improves locomotion on high gradient slopes, and this novel
  configuration is further analyzed. Using these approaches, we present an
  experimental evaluation of this fully self‐contained hexapod performing
  autonomous navigation on a multiterrain testbed and in outdoor terrain.
</p>
<div style="padding-top:30px"></div>