Topology-based representations for motion planning and generalization in dynamic environments with interactions

@article{Ivan2013TopologybasedRF,
  title={Topology-based representations for motion planning and generalization in dynamic environments with interactions},
  author={Vladimir Ivan and Dmitry Zarubin and Marc Toussaint and Taku Komura and Sethu Vijayakumar},
  journal={The International Journal of Robotics Research},
  year={2013},
  volume={32},
  pages={1151 - 1163}
}
Motion can be described in several alternative representations, including joint configuration or end-effector spaces, but also more complex topology-based representations that imply a change of Voronoi bias, metric or topology of the motion space. Certain types of robot interaction problems, e.g. wrapping around an object, can suitably be described by so-called writhe and interaction mesh representations. However, considering motion synthesis solely in a topology-based space is insufficient… 
Topology-based representations for motion planning and grasping
TLDR
The planning and synthesis of human-like motions remains an open area of robotics research and ideas and methods from topology are employed in order to improve current motion planning algorithms and to enable the synthesis of dexterous manipulations.
Robotic hierarchical motion planning in its representation space
In enormous and complicated robotic applications, the key issue in the field of robot motion planning is to endow the robot with the ability of adapting to various tasks. Although traditional
Homotopic particle motion planning for humanoid robotics
TLDR
This paper exploits the topology of the environment to discover connected components in the problem of motion planning for a humanoid robot and finds the homotopy classes of a single footstep trajectory in an environment.
Task-Space Decomposed Motion Planning Framework for Multi-Robot Loco-Manipulation
TLDR
A dual-resolution motion planning framework which utilizes a convex task region decomposition method, with each resolution tuned to efficient computation for their respective roles, for multi-robot simultaneous locomotion and manipulation.
Real-time motion adaptation using relative distance space representation
TLDR
This paper presents a novel approach that allows the robot to quickly adapt to changes, particularly in the presence of moving targets and dynamic obstacles, and introduces an incremental planning structure that allows us to handle unexpected obstacles that may appear during execution.
Reinforcement Learning in Topology-based Representation for Human Body Movement with Whole Arm Manipulation
TLDR
This paper models the task as a reinforcement learning problem in order to provide a robot behavior that can directly respond to external perturbation and human motion and represents global properties of the robot-human interaction with topology-based coordinates that are computed from arm and torso positions.
Scaling sampling-based motion planning to humanoid robots
TLDR
This work proposes a method that allows existing sampling-based algorithms directly to plan trajectories for humanoids by utilizing a customized state space representation, biased sampling strategies, and a steering function based on a robust inverse kinematics solver.
RMPflow: A Geometric Framework for Generation of Multitask Motion Policies
TLDR
A novel policy synthesis algorithm, Riemannian motion policy (RMP)flow, based on geometrically consistent transformations of RMPs is developed, which can engender natural behavior that adapts instantaneously to changing surroundings with zero planning while performing manipulation tasks.
Space-time area coverage control for robot motion synthesis
TLDR
A novel method for representing the interaction of a robot and an object is proposed, based on computation of electric flux, borrowed from electro dynamics, which makes it suitable as a complementary term in a cost function when constructing a multi-objective problem.
A topological extension of movement primitives for curvature modulation and sampling of robot motion
TLDR
A method to determine how many virtual segments best characterize a trajectory and then find such segments is proposed, which results in a generative model that permits modulating curvature to generate new samples, while still staying within the dataset distribution and being able to adapt to contextual variables.
...
...

References

SHOWING 1-10 OF 34 REFERENCES
Hierarchical Motion Planning in Topological Representations
TLDR
This paper proposes methods to combine and exploit different representations for motion synthesis, with specific emphasis on generalization of motion to novel situations, and illustrates the successful online motion generalization to dynamic environments on challenging, real world problems.
Path planning in 1000+ dimensions using a task-space Voronoi bias
TLDR
It is argued that providing a Voronoi bias in the task space can dramatically improve the performance of randomized motion planners, while still avoiding non-trivial constraints in the configuration (or state) space.
Manipulation Planning for Deformable Linear Objects
TLDR
A new motion planner for manipulating DLOs and tying knots (both self-knots and knots around simple static objects) using two cooperating robotic arms is described using a topologically biased probabilistic roadmap in the configuration space of the DLO.
Controlling humanoid robots in topology coordinates
TLDR
This paper presents an approach to the control of humanoid robot motion, e.g., holding another robot or tangled interactions involving multiple limbs, in a space defined by ‘topology coordinates’, and shows that this approach enables us to synthesize complex motion, such as tangling, very efficiently.
Identification and Representation of Homotopy Classes of Trajectories for Search-based Path Planning in 3D
TLDR
The development of a practical graph-search based planning tool with theoretical guarantees by combining integration theory with search techniques is described, and examples in three-dimensional spaces such as two-dimensional, dynamic environments and three- dimensional static environments are illustrated.
Planning 3-D Collision-Free Dynamic Robotic Motion Through Iterative Reshaping
TLDR
A general and practical planning framework for generating 3-D collision-free motions that take complex robot dynamics into account and demonstrates the effectiveness of the proposed method through examples of a space manipulator with highly nonlinear dynamics and a humanoid robot executing dynamic manipulation and locomotion at the same time.
Latent spaces for dynamic movement primitives
TLDR
This work focuses on learning discrete (point-to-point) movements and proposes a modification of a powerful nonlinear dimensionality reduction technique (Gaussian Process Latent Variable Model), which makes the GPLVM more suitable for the use of DMPs by favouring latent spaces with highly regular structure.
Lazy Reconfiguration Forest (LRF) - An Approach for Motion Planning with Multiple Tasks in Dynamic Environments
TLDR
A novel algorithm for robot motion planning in dynamic environments that dynamically maintains a forest of trees by splitting, growing and merging them on the fly to adapt to moving obstacles and robot motion.
Search-Based Path Planning with Homotopy Class Constraints
TLDR
Using such a representation, it is shown that homotopy class constraints can be seamlessly weaved into graph search techniques for determining optimal path constrained to certain Homotopy classes or forbidden from others, as well as for exploring different homOTopy classes in an environment.
An Approximate Inference Approach to Temporal Optimization in Optimal Control
TLDR
An approximate EM algorithm is derived that efficiently optimizes both the movement duration and control commands offering, for the first time, a practical approach to tackling generic via point problems in a systematic way under the optimal control framework.
...
...