Explaining a Deep Reinforcement Learning Docking Agent Using Linear Model Trees with User Adapted Visualization

Gjærum, Vilde B.; Strümke, Inga; Alsos, Ole Andreas; Lekkas, Anastasios M.

doi:10.3390/jmse9111178

Cited by 12 publications

(4 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Gjaerum et al [14] propose to use decision trees with linear functions in the leaf nodes (linear model trees [87]), showing that they can be valid for generating counterfactual explanations in relatively competitive scenarios with multiple, continuous outputs. Specifically, the challenge is solved using H-LMTs (Heuristic Linear Model Trees) [88]. An H-LMT does not take maximum depth as a restriction but rather the maximum number of leaf nodes allowed.…”

Section: Exploring the Recent Past: Why Did You Do This Instead Of That?mentioning

confidence: 99%

Causal-Based Approaches to Explain and Learn from Self-Extension—A Review

Marfil,

Bustos,

Bandera

2024

Electronics

View full text Add to dashboard Cite

The last decades have seen a revolution in autonomous robotics. Deep learning approaches and their hardware implementations have made it possible to endow robots with extraordinary perceptual capabilities. In addition, they can benefit from advances in Automated Planning, allowing them to autonomously solve complex tasks. However, on many occasions, the robot still acts without internalising and understanding the reasons behind a perception or an action, beyond an immediate response to a current state of the context. This gap results in limitations that affect its performance, reliability, and trustworthiness. Deep learning alone cannot bridge this gap because the reasons behind behaviour, when it emanates from a model in which the world is a black-box, are not accessible. What is really needed is an underlying architecture based on deeper reasoning. Among other issues, this architecture should enable the robot to generate explanations, allowing people to know why the robot is performing, or has performed, a certain action, or the reasons that may have caused a certain plan failure or perceptual anomaly. Furthermore, when these explanations arise from a cognitive process and are shared, and thus validated, with people, the robot should be able to incorporate these explanations into its knowledge base, and thus use this understanding to improve future behaviour. Our article looks at recent advances in the development of self-aware, self-evolving robots. These robots are designed to provide the necessary explanations to their human counterparts, thereby enhancing their functional capabilities in the quest to gain their trust.

show abstract

Section: Exploring the Recent Past: Why Did You Do This Instead Of That?mentioning

confidence: 99%

Causal-Based Approaches to Explain and Learn from Self-Extension—A Review

Marfil,

Bustos,

Bandera

2024

Electronics

View full text Add to dashboard Cite

show abstract

“…A problematic feature of ANNs is their black-box nature, making it difficult to understand the decision-making process. However, there are some attempts to use methods from explainable AI (XAI) to make sense of the decisions made by ANNs employed for the docking problem [159], [160], [164], [209].…”

Section: Ann (Artificial Neural Network)mentioning

confidence: 99%

Automated Docking for Marine Surface Vessels—A Survey

Lexau,

Breivik,

Lekkas

2023

IEEE Access

Self Cite

View full text Add to dashboard Cite

Research on motion control systems for marine surface vessels has generated a vast academic literature and many industrial applications since the early 20th century. The recent focus on autonomous ships has sparked intensive research and innovation activities of dock-to-dock operations, including the final docking phase. Specifically, automated docking involves systems that enable vessels to dock safely, independently, and energy-efficiently at designated locations with a specific heading. Challenges include managing large sideslip angles, static and dynamic obstacles, and navigation in complex port geometries. Despite its complexity, the docking problem has received less attention compared to dynamic positioning or course-keeping systems. This paper therefore provides a thorough overview of the research on automated docking for marine surface vessels, from 1980 to June 2023. The paper introduces the Docking Characteristic Index (DCI) as a metric to identify the scope and approach of systems addressing various docking challenges, not as an evaluation of their overall quality. Using the calculated DCI-Scores, we discuss some of the works in more detail. The survey reveals a rising trend in publications and an increasing emphasis on experimental verification. Despite advancements, current methodologies exhibit trade-offs and limitations, particularly in handling dynamic obstacles, robustness against external forces, and situational awareness. The paper identifies the need for more comprehensive and integrated solutions to these challenges. As the demand for fully autonomous operations grows, the results suggest that future research should focus on developing holistic and robust docking strategies, that are verified experimentally, to achieve safe, efficient, and effective automated docking systems.

show abstract

“…Liessner 2016) with SHAP values to create a new explanation method for controlling an aerial vehicle. Gjaerum et.al evaluated XAI methods like LIME, Anchors, IG, SHAP, and SAGE for DNN explanations, tailored for developers and seafarers/operators, focusing on their need for quick and risk-aware decision-making [ 11]. Apart from these applications of Shapley values to RL, Beechey et al (2023) present the first theoretical analysis of applying Shapley values to explain RL and show that previous uses are incorrect or incomplete [ 4].…”

Section: Introductionmentioning

confidence: 99%

Learning behavior of offline reinforcement learning agents

Shukla,

Dozier,

Henslee

2024

Artificial Intelligence and Machine Learning for Multi-Domain Operations Applications VI

View full text Add to dashboard Cite

Reinforcement learning (RL) agents offer significant value for military applications by effectively navigating complex, dynamic environments typical of mission engineering and operational analysis. Once trained, these agents can be employed to inform mission planners on optimal strategies, tactics, or even innovative utilization of different military platforms within a given scenario. In recent years, RL has become a major research area for automation and solving complex sequential decision-making problems. However, a notable challenge lies in the inherent black-box nature of RL models and their inability to explain their decisions and actions. This limitation serves as a major adoption barrier, especially in Defense. This paper aims to study EXplainable RL (XRL) within an operational context. XRL is a distinct branch of Explainable Artificial Intelligence (XAI) techniques that provides the necessary transparency to make AI models more transparent to address this challenge. This research is an effort to gain insight into the behavior of RL agents in an operational environment and to discuss explainability and interpretability through the lens of different roles within the decision-making pipeline.

show abstract

Explaining a Deep Reinforcement Learning Docking Agent Using Linear Model Trees with User Adapted Visualization

Cited by 12 publications

References 30 publications

Causal-Based Approaches to Explain and Learn from Self-Extension—A Review

Causal-Based Approaches to Explain and Learn from Self-Extension—A Review

Automated Docking for Marine Surface Vessels—A Survey

Learning behavior of offline reinforcement learning agents

Contact Info

Product

Resources

About