Determining possible failure scenarios is a critical step in the evaluation of autonomous vehicle systems. Real world vehicle testing is commonly employed for autonomous vehicle validation, but the costs and time requirements are high. Consequently, simulation driven methods such as Adaptive Stress Testing (AST) have been proposed to aid in validation. AST formulates the problem of finding the most likely failure scenarios as a Markov decision process, which can be solved using reinforcement learning. In practice, AST tends to find scenarios where failure is unavoidable and tends to repeatedly discover the same types of failures of a system. This work addresses these issues by encoding domain relevant information into the search procedure. With this modification, the AST method discovers a larger and more expressive subset of the failure space when compared to the original AST formulation. We show that our approach is able to identify useful failure scenarios of an autonomous vehicle policy.
<div>Identifying and eliminating failure scenarios is critical in the development of autonomous vehicle (AV) systems. However, finding such failures through real-world vehicle-level testing is a difficult task as system disengagements and accidents are rare occurrences. Simulation approaches have been proposed to supplement vehicle-level testing and reduce the costs associated with operating large fleets of autonomous test vehicles. While one can run more vehicles in simulation than in the real world, applying traditional Monte Carlo sampling techniques to find failures still yields an unguided search and a large waste of computing resources. A more directed method than random sampling is needed to identify failure scenarios in a computationally efficient manner. Adaptive Stress Testing (AST) is a method that uses reinforcement learning (RL) paradigms to efficiently find failure scenarios in stochastic sequential decision-making systems. Through iteratively exploring the action space and collecting rewards, AST aims to establish an optimal policy that generates a set of high-probability failure trajectories. However, the trajectories obtained through AST tend to lack diversity and converge to similar failure states. Due to the range of possible accident scenarios an AV can face, such homogeneous failures are not very beneficial in validating a system’s roadworthiness. In this article, we present a method to enhance the expressiveness of the failure scenarios found using AST. By augmenting the reward function used by AST with domain relevant information, we guide the solver to discover more diverse sets of trajectories. We present an implementation using Monte Carlo Tree Search (MCTS). To show the efficacy of our approach, we evaluate the failure trajectories obtained for a vehicle and pedestrian crosswalk scenario. We show that our implementation is able to find more diverse and domain-relevant failures when compared with baseline AST.</div>
The nervous system plays an increasingly appreciated role in the regulation of cancer. In malignant gliomas, neuronal activity drives tumor progression not only through paracrine signaling factors such as neuroligin-3 and brain-derived neurotrophic factor (BDNF), but also through electrophysiologically functional neuron-to-glioma synapses. Malignant synapses are mediated by calcium-permeable AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid) receptors in both pediatric and adult high-grade gliomas, and consequent depolarization of the glioma cell membrane drives tumor proliferation. The nervous system exhibits plasticity of both synaptic connectivity and synaptic strength, contributing to neural circuit form and functions. In health, one factor that promotes plasticity of synaptic connectivity and strength is activity-regulated secretion of the neurotrophin BDNF. Here, we show that malignant synapses exhibit similar plasticity regulated by BDNF-TrkB (tropomyosin receptor kinase B) signaling. Signaling through the receptor TrkB, BDNF promotes AMPA receptor trafficking to the glioma cell membrane, resulting in increased amplitude of glutamate-evoked currents in the malignant cells. This potentiation of malignant synaptic strength shares mechanistic features with the long-term potentiation (LTP) that is thought to contribute to memory and learning in the healthy brain. BDNF-TrkB signaling also regulates the number of neuron-to-glioma synapses. Abrogation of activity-regulated BDNF secretion from the brain microenvironment or loss of TrkB in human glioma cells exerts growth inhibitory effects in vivo and in neuron:glioma co-cultures that cannot be explained by classical growth factor signaling alone. Blocking TrkB genetically or pharmacologically abrogates these effects of BDNF on glioma synapses and substantially prolongs survival in xenograft models of pediatric glioblastoma and diffuse intrinsic pontine glioma (DIPG). Taken together, these findings indicate that BDNF-TrkB signaling promotes malignant synaptic plasticity and augments tumor progression. While targeting Trk signaling is presently being explored for NTRK-fusion brain tumors, these findings indicate that BDNF-TrkB signaling may also represent an important therapeutic target for NTRK2 wildtype gliomas.
SPEThis pafw was prepared for presenlalicm at the 56th Calitomia Reg!onalMeeting of the Soc!etyof Petroleum Engineers held m Oakland, CA, April2-4, 19ss,This paper was selected for presentation by an SPE Program Committee followingreview of informationoon!ainad in an abatract submittedby the author(a).Contents of the papar, ae preaentad, have no! been reviewed by the Society of Petroleum Engineers and are subjad to correctionby the author(e),The matarial, aa presented, doea not necessarily reflect any positionof the Society of Petroleum Engineers, its officers,or members. Papere presented al SPE maalinga are subject to publication review by Editorial Committee of the Society of Petroleum Engineers, Permission to copy la reatricmdto an abatractof not more than 300 words, Iilustralionamay notbe copied The abstract shouldcontainconspicuousacknowledgmentof where and by whom the paper is presented, Write Publicaliona Manager, SPE, P.0, Box S33S3S, Richardson, TX 750S3-3S3S, Telex, 7309s9, SPEDAL.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.