Getting Robots Unfrozen and Unlost in Dense Pedestrian Crowds

@article{Fan2019GettingRU,
  title={Getting Robots Unfrozen and Unlost in Dense Pedestrian Crowds},
  author={Tingxiang Fan and Xinjing Cheng and Jia Pan and Pinxin Long and Wenxi Liu and Ruigang Yang and Dinesh Manocha},
  journal={IEEE Robotics and Automation Letters},
  year={2019},
  volume={4},
  pages={1178-1185}
}
Our goal is to navigate a mobile robot to navigate through environments with dense crowds, e.g., shopping malls, canteens, train stations, or airport terminals. In these challenging environments, existing approaches suffer from two common problems: the robot may get frozen and cannot make any progress toward its goal, or it may get lost due to severe occlusions inside a crowd. Here, we propose a navigation framework that handles the robot freezing and the navigation lost problems simultaneously… Expand
Frozone: Freezing-Free, Pedestrian-Friendly Navigation in Human Crowds
TLDR
Frozone is a novel algorithm that senses and explicitly predicts the trajectories of pedestrians and constructs a Potential Freezing Zone (PFZ) where the robot could freeze or be obtrusive to humans, which results in smooth and collision-free navigation in dense environments. Expand
Applying Deep Reinforcement Learning in the Navigation of Mobile Robots in Static and Dynamic Environments
Nowadays mobile robots operate reliably in clean static environments like industry setups, but rather fail in complex dynamic environments, like e.g. airports or shopping centers. Those environmentsExpand
DenseCAvoid: Real-time Navigation in Dense Crowds using Anticipatory Behaviors
TLDR
This work presents DenseCAvoid, a novel algorithm for navigating a robot through dense crowds and avoiding collisions by anticipating pedestrian behaviors that reduces the occurrence of the robot freezing in a crowd by up to 48%, and performs comparably with respect to trajectory lengths and mean arrival times to goal. Expand
Realtime Collision Avoidance for Mobile Robots in Dense Crowds using Implicit Multi-sensor Fusion and Deep Reinforcement Learning
TLDR
This work presents a novel learning-based collision avoidance algorithm, CrowdSteer, for mobile robots operating in dense and crowded environments and shows that the learned navigation model is directly transferable to previously unseen virtual and dense real-world environments. Expand
IAN: Multi-Behavior Navigation Planning for Robots in Real, Crowded Environments
TLDR
This work addresses the question of whether it can be beneficial to use interaction actions, such as saying, touching, and gesturing, for the sake of allowing robots to navigate in unstructured, crowded environments, and proposes a high level planning algorithm for multi-behavior navigation, named Interaction Actions for Navigation (IAN). Expand
Long-Range Indoor Navigation With PRM-RL
TLDR
This article uses probabilistic roadmaps (PRMs) as the sampling-based planner, and AutoRL as the RL method in the indoor navigation context, and shows that PRM-RL with AutoRL is more successful than several baselines, is robust to noise, and can guide robots over hundreds of meters in the face of noise and obstacles in both simulation and on robots. Expand
Towards Multi-Modal Perception-Based Navigation: A Deep Reinforcement Learning Method
TLDR
A novel navigation system of unmanned ground vehicle (UGV) for local path planning based on deep reinforcement learning that decouples perception from control and takes advantage of multi-modal perception for a reliable online interaction with the surrounding environment of the UGV, exhibiting the superiority over state-of-the-art approaches. Expand
Move Beyond Trajectories: Distribution Space Coupling for Crowd Navigation
TLDR
A general objective for coupled prediction and planning in distribution space is introduced, and an iterative best response optimization method based on variational analysis with guaranteed sufficient decrease is proposed that is competitive with human safety and efficiency performance. Expand
From Crowd Simulation to Robot Navigation in Crowds
TLDR
It turns out that Crowd Simulation relies on two assumptions that would not hold for a real mobile robot, so a test protocol has been designed in order to thoroughly evaluate how three representative Crowd Simulation techniques would perform when said assumptions are relaxed. Expand
Efficient re-localization of mobile robot using strategy of finding a missing person
TLDR
The proposed SFMP-based re-localization ensures the existence of particles in where the robot appears again, which improves the robustness of MCL. Expand
...
1
2
3
4
...

References

SHOWING 1-10 OF 36 REFERENCES
Unfreezing the robot: Navigation in dense, interacting crowds
TLDR
IGP is developed, a nonparametric statistical model based on dependent output Gaussian processes that can estimate crowd interaction from data that naturally captures the non-Markov nature of agent trajectories, as well as their goal-driven navigation. Expand
Robot navigation in dense human crowds: Statistical models and experimental studies of human–robot cooperation
TLDR
It is concluded that a cooperation model is critical for safe and efficient robot navigation in dense human crowds and the salient characteristics of nearly any dynamic navigation algorithm. Expand
Robot navigation in dense human crowds: the case for cooperation
TLDR
A cooperation model is important for safe and efficient robot navigation in dense human crowds by developing a probabilistic predictive model of cooperative collision avoidance and goal-oriented behavior by extending the interacting Gaussian processes approach to include multiple goals and stochastic movement duration. Expand
Socially aware motion planning with deep reinforcement learning
TLDR
Using deep reinforcement learning, this work develops a time-efficient navigation policy that respects common social norms and is shown to enable fully autonomous navigation of a robotic vehicle moving at human walking speed in an environment with many pedestrians. Expand
Towards Optimally Decentralized Multi-Robot Collision Avoidance via Deep Reinforcement Learning
TLDR
This work presents a decentralized sensor-level collision avoidance policy for multi-robot systems, which directly maps raw sensor measurements to an agent's steering commands in terms of movement velocity and demonstrates that the learned policy can be well generalized to new scenarios that do not appear in the entire training period. Expand
Lifelong localization in changing environments
TLDR
This paper presents a sound probabilistic approach to lifelong localization in changing environments using a combination of a Rao-Blackwellized particle filter with a hidden Markov model, and obtains a highly efficient map management approach for dynamic environments. Expand
Active localization with dynamic obstacles
TLDR
Experimental results from multiple simulations and from real robot deployments validate the localization improvements achieved by the proposed method. Expand
Fully Distributed Multi-Robot Collision Avoidance via Deep Reinforcement Learning for Safe and Efficient Navigation in Complex Scenarios
TLDR
This paper presents a decentralized sensor-level collision avoidance policy for multi-robot systems, which shows promising results in practical applications and directly maps raw sensor measurements to an agent's steering commands in terms of the movement velocity. Expand
Markov Localization for Mobile Robots in Dynamic Environments
TLDR
A version of Markov localization which provides accurate position estimates and which is tailored towards dynamic environments, and includes a filtering technique which allows a mobile robot to reliably estimate its position even in densely populated environments in which crowds of people block the robot's sensors for extended periods of time. Expand
BRVO: Predicting pedestrian trajectories using velocity-space reasoning
We introduce a novel, online method to predict pedestrian trajectories using agent-based velocity-space reasoning for improved human–robot interaction and collision-free navigation. Our formulationExpand
...
1
2
3
4
...