SABER: Data-Driven Motion Planner for Autonomously Navigating Heterogeneous Robots

@article{Schperberg2021SABERDM,
  title={SABER: Data-Driven Motion Planner for Autonomously Navigating Heterogeneous Robots},
  author={Alexander Schperberg and Stephanie Tsuei and Stefano Soatto and Dennis W. Hong},
  journal={IEEE Robotics and Automation Letters},
  year={2021},
  volume={6},
  pages={8086-8093}
}
We present an end-to-end online motion planning framework that uses a data-driven approach to navigate a heterogeneous robot team towards a global goal while avoiding obstacles in uncertain environments. First, we use stochastic model predictive control (SMPC) to calculate control inputs that satisfy robot dynamics, and consider uncertainty during obstacle avoidance with chance constraints. Second, recurrent neural networks are used to provide a quick estimate of future state uncertainty… 

Figures and Tables from this paper

COP: Control & Observability-aware Planning

This proposed novel Control & Observability-aware Planning (COP) framework is the first that uses these possibly opposing objectives in a Single-Objective Optimization problem based on the Augmented Weighted Tchebycheff method to perform the balancing of them and generation of Bézier curve-based trajectories.

Auto-Tuning of Controller and Online Trajectory Planner for Legged Robots

This letter presents an approach for auto-tuning feedback controllers and online trajectory planners to achieve robust locomotion of a legged robot. The auto-tuning approach uses an Unscented Kalman

References

SHOWING 1-10 OF 25 REFERENCES

Risk-Averse MPC via Visual-Inertial Input and Recurrent Networks for Online Collision Avoidance

This paper combines an object detection pipeline with a recurrent neural network (RNN) which infers the covariance of state estimates through each step of the model predictive control (MPC)’s finite time horizon to consider future location uncertainties for safer navigation through cluttered environments.

Model-based Reinforcement Learning for Decentralized Multiagent Rendezvous

This work proposes hierarchical predictive planning (HPP), a model-based reinforcement learning method for decentralized multiagent rendezvous that removes the need for a centralized operator in multiagent systems by combining model- based RL and inference methods, enabling agents to dynamically align plans.

Chance-Constrained Optimal Path Planning With Obstacles

A chance-constrained approach that plans the future probabilistic distribution of the vehicle state so that the probability of failure is below a specified threshold, and introduces a customized solution method that returns almost-optimal solutions along with a hard bound on the level of suboptimality.

A Scalable Framework For Real-Time Multi-Robot, Multi-Human Collision Avoidance

A novel framework for robot navigation is introduced that accounts for high-order system dynamics and maintains safety in the presence of external disturbances, other robots, and humans, effectively enabling scalable safe trajectory planning and execution.

Risk-Aware Motion Planning for a Limbed Robot with Stochastic Gripping Forces Using Nonlinear Programming

A motion planning algorithm with probabilistic guarantees for limbed robots with stochastic gripping forces is presented, formulated as a nonlinear programming problem with chance constraints, which allows the robot to generate a variety of motions based on different risk bounds.

Coordinating heterogeneous teams of robots using temporal symbolic planning

The results show that this approach outperforms ad-hoc extensions of state-of-the-art cost-based coordination methods and that the approach is able to efficiently coordinate teams of heterogeneous robots and to consider symbolic actions.

FASTER: Fast and Safe Trajectory Planner for Flights in Unknown Environments

FASTER (Fast and Safe Trajectory Planner) obtains high-speed trajectories by enabling the local planner to optimize in both the free-known and unknown spaces by way of a Mixed Integer Quadratic Program formulation.

Avoidance Critical Probabilistic Roadmaps for Motion Planning in Dynamic Environments

This work proposes a self-supervised methodology for learning to identify regions frequently used for obstacle avoidance from local environment features and leverages a neural network to generate hierarchical probabilistic roadmaps termed Avoidance Critical Probabilistic Roadmaps (ACPRM).

Neural Collision Clearance Estimator for Batched Motion Planning

A neural network collision checking heuristic, ClearanceNet, and a planning algorithm, CN-RRT, that accelerates motion planning by up to 42% over a baseline and finds paths up to 36% more efficient in configuration spaces with up to 30 degrees of freedom.