Fast discovery of influential outcomes for risk-aware MPDM

Mehta, Dhanvin; Ferrer, Gonzalo; Olson, Edwin

doi:10.1109/icra.2017.7989736

Cited by 8 publications

(5 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Model-Based. We can also cite model-based works, these methods use several models / behavior depending on the situation, such as Multi-Policy Decision-Making [23,24] which uses the policy with the best expected utility among a set of different policies. Sebastian et al [25] proposed a Gaussian Mixture Model in order to differentiate people's behavior and choose the most relevant and respectful path towards people.…”

Section: Supervising Learningmentioning

confidence: 99%

Benchmarking Off-the-Shelf Human-Aware Robot Navigation Solutions

Gouguet,

Karami,

Lozenguez

et al. 2024

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

In recent years, the number of robotic applications in public spaces has been growing. Decades of research have given rise to various methods of human-aware robotic navigation. There are a lot of different navigation solutions to guide a robot in presence of humans. Despite multiple surveys comparing existing navigation solutions, few of them take social criteria into account. In this sense, it is difficult to evaluate existing methods and select the one that performs better in a given context. In this article, we first provide a thorough classification of state-of-the-art solutions regarding human-aware robotic navigation solutions. Then, we select a set of measurable criteria to evaluate both the efficiency and the social-compliance of navigation solutions. Using these criteria, we finally compare representative off-the-shelf navigation solutions using the SEAN Simulator to identify the most suitable for human-aware navigation.

show abstract

Section: Supervising Learningmentioning

confidence: 99%

Benchmarking Off-the-Shelf Human-Aware Robot Navigation Solutions

Gouguet,

Karami,

Lozenguez

et al. 2024

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

show abstract

“…In each Monte Carlo sample, all the agents are forward simulated together to capture closed-loop interactions [29] and the best ego agent policy is elected from the cumulative results. Further work in the MPDM framework focuses on discovering risky configurations through stochastic gradient descent of a heuristic cost function [30] and back-propagation techniques [31].…”

Section: F Multi-policy Approachesmentioning

confidence: 99%

Monte-Carlo Policy-Tree Decision Making

Haggenmiller¹,

Olson²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

In this paper, we propose Monte-Carlo Policy-Tree Decision Making (MCPTDM), an uncertainty-aware framework for high-variance planning problems with multiple dynamic agents. Planning when surrounded by multiple uncertain dynamic agents is hard because we cannot be certain of either the initial states or the future actions of those agents, leading to an exponential explosion in possible futures. Many important real-world problems, such as autonomous driving, fit this model. To address these difficulties, we combine Multi-policy Decision Making (MPDM) and Monte Carlo tree search (MCTS) and perform policy tree search with marginal action cost (MAC) estimation and repeated belief particles. We first design a synthetic experiment to evaluate these novel improvements in isolation. Then we evaluate the complete framework in a self-driving car simulation experiment and compare it against MPDM and Efficient Uncertainty-aware Decision Making (EUDM) methods. We release our complete source code for replicating our experiments and results.

show abstract

“…Moreover, the hidden intentions (driving policy of other agents) are sampled according to initial behavioral prediction (initial belief) and will not be updated during the simulation. As a result, risky outcomes may not be reflected in policy evaluation due to inaccurate initial behavior prediction or insufficient intention samples [17].…”

Section: Related Workmentioning

confidence: 99%

“…In the case of MPDM, the intention of the nearby vehicles is fixed for the whole planning horizon, and the initial intention is sampled according to a behavior prediction algorithm. The limitation of MPDM is that, with a limited number of samples, influential risky outcomes may not be rolled out, especially when the initial intention prediction is inaccurate [17].…”

Section: Conditional Focused Branchingmentioning

confidence: 99%

See 1 more Smart Citation

Efficient Uncertainty-aware Decision-making for Automated Driving Using Guided Branching

Zhang

Ding

Chen

et al. 2020

Preprint

View full text Add to dashboard Cite

Decision-making in dense traffic scenarios is challenging for automated vehicles (AVs) due to potentially stochastic behaviors of other traffic participants and perception uncertainties (e.g., tracking noise and prediction errors, etc.). Although the partially observable Markov decision process (POMDP) provides a systematic way to incorporate these uncertainties, it quickly becomes computationally intractable when scaled to the real-world large-size problem. In this paper, we present an efficient uncertainty-aware decision-making (EUDM) framework, which generates long-term lateral and longitudinal behaviors in complex driving environments in realtime. The computation complexity is controlled to an appropriate level by two novel techniques, namely, the domain-specific closed-loop policy tree (DCP-Tree) structure and conditional focused branching (CFB) mechanism. The key idea is utilizing domain-specific expert knowledge to guide the branching in both action and intention space. The proposed framework is validated using both onboard sensing data captured by a real vehicle and an interactive multi-agent simulation platform. We also release the code of our framework to accommodate benchmarking.

show abstract

Fast discovery of influential outcomes for risk-aware MPDM

Cited by 8 publications

References 22 publications

Benchmarking Off-the-Shelf Human-Aware Robot Navigation Solutions

Benchmarking Off-the-Shelf Human-Aware Robot Navigation Solutions

Monte-Carlo Policy-Tree Decision Making

Efficient Uncertainty-aware Decision-making for Automated Driving Using Guided Branching

Contact Info

Product

Resources

About