STOMP: Stochastic trajectory optimization for motion planning

By Mrinal Kalakrishnan, Sachin Chitta, Evangelos A. Theodorou, Peter Pastor, Stefan Schaal. IEEE International Conference on Robotics and Automation, 2011

1,015 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Control and Systems Engineering

View Paper: https://www.semanticscholar.org/paper/56ca783385dfa46a3917850bb52179d0bba25903

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Control and Systems Engineering

Citations in the corpus

Finding Locally Optimal, Collision-Free Trajectories with Sequential Convex Optimization (2013, background): “…includes STOMP (Stochastic Trajectory Optimization for Motion Planning), which uses a gradient-free, stochastic scheme for optimization [12]; and ITOMP (Incremental trajectory optimization for real-time re-planning in dynamic environments), which deals with dynamic obstacles and real-time…”
Generating Legible Motion (2013, background): “IV), echoing earlier works in motion planning [9, 21, 22, 27, 29, 33, 35], now with legibility as an optimization criterion.”
Variational Policy Search via Trajectory Optimization (2013, neutral): “Another interesting avenue for future work is to apply model-free trajectory optimization techniques [7], which would avoid the need for a model of the system dynamics, or to learn the dynamics from data, for example by using Gaussian processes [2].”
Robust Control of Mobility and Communications in Autonomous Robot Teams (2013, methodology): “Trajectory optimization methods such as CHOMP [30] and STOMP [31] are inapplicable since our problem statement yields infinite cost and does not permit gradient calculations from inadmissible states.”
Human-robot collaborative manipulation planning using early prediction of human motion (2013, methodology): “The initial algorithm presented in [23] optimizes a combination of two classical criteria, namely obstacle cost and smoothness cost.”
CHOMP: Covariant Hamiltonian optimization for motion planning (2013, methodology): “Researchers have developed algorithms recently that take inspiration from CHOMP, but make alternate design trade-offs [30, 51], and additionally have built this and related approaches into more sophisticated planning techniques.”
Robust and Efficient Quadrotor Trajectory Generation for Fast Autonomous Flight (2019, neutral): “[11] has similar problem formulation, but the optimization is solved by a gradient-free sampling method.”
Motion planning with sequential convex optimization and convex collision checking (2014, background): “On the other hand, spheres and distance ﬁelds are arguably not very well suited to situations where one needs to accurately model geometry, which is why collision-detection methods based on meshes and convex primitives are more prevalent in applications like real-time physics simulation [7] for…”
Guided Policy Search (2013, background): “A promising avenue for future work is to build the guiding distributions with model-free methods that either build trajectory following policies (Ijspeert et al., 2002) or perform stochastic trajectory optimization (Kalakrishnan et al., 2011).”
RAPTOR: Robust and Perception-Aware Trajectory Replanning for Quadrotor Fast Flight (2020, background): “[17] presented a similar formulation, but solves the problem by sampling neighboring candidates iteratively.”
MoveIt! [ROS Topics] (2012, methodology): “A generic ROS interface for geometric planning allowed multiple motion planners to be integrated, including randomized planners from the Open Motion Planning Library (OMPL) [2], search-based planners from the Search-Based Planning Library (SBPL) [3], and trajectory optimization libraries including C”
Optimization-Based Collision Avoidance (2017, neutral): “One way of dealing with obstacle avoidance is to use unconstrained optimization, in which case the objective function is augmented with “artificial potential fields” that represent the obstacles [32, 33, 34, 35, 36].”
EGO-Planner: An ESDF-Free Gradient-Based Local Planner for Quadrotors (2020, neutral): “Nevertheless, optimizing the trajectory in discrete-time [5], [6] is not suitable for drones, because it is much more sensitive to dynamical constraints.”
Recent Advances in Robot Learning from Demonstration (2020, methodology): “Traditional trajectory-optimization methods (5, 107, 108) were originally introduced to generate smooth and efficient trajectories for robots to move between any two points in space.”
A policy-blending formalism for shared control (2013, methodology): “Given such a cost, the robot would use a trajectory optimizer (Todorov and Li, 2005; Ratliff et al., 2009; Toussaint, 2009; Dragan et al., 2011b; Kalakrishnan et al., 2011) to acquire the intended trajectory (Figure 4).”
Real-Time Perception Meets Reactive Motion Generation (2017, methodology): “SDFs provide a way to perform fast collision checking, which has been exploited in many motion optimization works [9], [29], [30].”
Movement primitives via optimization (2015, background): “…a target tracker with a moving target, T (z): τ2ξ̈(t) = K(T (z)− ξ(t))− Dτξ̇(t) (12) with T (z) moving from s to g as a function of z on a straight line constant speed in z plus a deviation f as a function of z, f (z) = f̄ (u): T (z) = s + z(g− s) + f (z) (13) 1Note that here we are overloading M.”
End-to-end dexterous manipulation with deliberate interactive estimation (2012, methodology): “STOMP [23] similarly refines an initial motion through iterative updates based on a weighted sum of perturbed trajectory parameters as opposed to an analytic or numeric gradient.”
A unified framework for coordinated multi-arm motion planning (2018, background): “…ﬁelds have been successful in encoding proximity of obstacles as continuous costs in local trajectory optimization frameworks; by either providing explicit cost gradients (Ratliff et al. 2009; Zucker et al. 2013) or through derivative-free stochastic optimization methods (Kalakrishnan et al. 2011).”