Towards Verification of Learning-Based UAS Trajectory Generation in Uncertain Environments

Reinforcement learning (RL), a promising framework for data-driven decision making in an uncertain environment, has successfully been applied in many real-world operation and control problems. However, the application of RL in a large-scale decentralized multi-agent environment remains a challenging problem due to the partial observability and limited communications between agents. In this paper, we develop a model-based kernel RL approach and a model-free deep RL approach for learning a decentralized, shared policy among homogeneous agents. By leveraging the strengths of both these methods, we further propose a novel deep ensemble multi-agent reinforcement learning (MARL) method that efficiently learns to arbitrate between the decisions of the local kernel-based RL model and the wider-reaching deep RL model. We validate the proposed deep ensemble method on a highly challenging real-world air traffic control problem, where the goal is to provide effective guidance to aircraft to avoid air traffic congestion, conflicting situations, and to improve arrival timeliness, by dynamically recommending adjustments of aircraft speeds in real-time. Extensive empirical results from an open-source air traffic management simulation model, developed by Eurocontrol and built on a real-world data set including thousands of aircrafts, demonstrate that our proposed deep ensemble MARL method significantly outperforms three state-of-the-art benchmark approaches.

Introduction Reinforcement learning formalizes the concept of learning from interactions.1 Broadly, reinforcement learning focuses on a setting in which an agent (decision maker) sequentially interacts with an environment that is partially unknown to them. At each stage, the agent takes an action and receives a reward. The objective of the agent is to maximize rewards accumulated in the long run. There are many situations in health care where decisions are made sequentially for which reinforcement learning approaches could prove useful for decision making. Throughout this article, we consider treatment prescription as an archetypical example to connect reinforcement learning concepts to a health care setting. In this setting, the care provider, the prescribed treatment, and the patients can be viewed as the agent, the action, and the environment, respectively, as depicted in Figure 1.Figure 1: Sequential treatment of AKI or CKD complications modeled as a reinforcement learning problem.Background In this section, with the objective of making reinforcement learning literature more accessible to a clinical audience, we briefly introduce related fundamental concepts and approaches. We refer the interested reader to Sutton and Barto1 for a comprehensive introduction to reinforcement learning. Markov Decision Processes Markov decision processes (MDPs) are a formalism of the sequential decision-making problem that has been central to the theoretical and practical advancements of reinforcement learning. In each stage of an MDP, the agent observes the state of the environment and takes an action, which, in turn, results in a change of the state. This change of state is assumed to be probabilistic with the next state being determined only by the preceding state, the chosen action, and the transition probability. The agent also receives a reward that is a function of the taken action, the preceding state, and the subsequent state. In an MDP, the objective of the agent is to maximize the return defined as the reward accumulated over a time horizon. In some applications, it is common to consider the horizon to be infinite, in which case the future rewards are discounted by a factor smaller than one. The selection of action by the agent on the basis of the observed state is known as the policy. More formally, a policy is a probabilistic mapping from states to each possible action. Because the policy and the reward are a function of the state, it is critical to estimate the utility of being in a certain state. More specifically, the value function is defined as the expected return starting from a given state under the chosen policy. Under this formalism, the objective of the agent is to find the optimal policy that maximizes the value function for all states. Reinforcement Learning Methods Action-value methods are a class of reinforcement learning methods in which the actions are chosen on the basis of the estimation of their long-term value. A prominent example of an action-value method is Q-learning in which the agent iteratively takes actions with the highest estimated values and updates the action-state value function on the basis of new observations. Policy gradient methods are another class of reinforcement learning methods that seek to optimize the policy directly instead of choosing actions on the basis of their respective estimated value. Such methods could be advantageous in health care applications that entail a large number of possible actions, e.g., when recommending a wide range of drug dosages or treatment options. Clinical Applications Reinforcement learning frameworks and methods are broadly applicable to clinical settings in which decisions are made sequentially. A prominent clinical application of reinforcement learning is for treatment recommendation, which has been studied across a variety of diseases and treatments including radiation and chemotherapy for cancer, brain stimulation for epilepsy, and treatment strategies for sepsis.2–5 In such treatment recommendation settings, a policy is commonly known as a dynamic treatment regime. There are various other clinical applications of reinforcement learning including diagnosis, medical imaging, and decision support tools (see refs. 2–5 and the references therein). Reinforcement Learning in Nephrology Although there have been recent applications of machine learning in nephrology,6,7 to the best of the authors' knowledge, the application of reinforcement learning to nephrology has been primarily limited to optimizing the erythropoietin dosage in hemodialysis patients.8,9 However, there are other settings where reinforcement learning has the potential to improve patient care in nephrology. For example, reinforcement learning methods can be adopted in the treatment of the complications of AKI or CKD (Figure 1). In this problem, the state models the conditions of the patient (e.g., vital signs, laboratory test results including urine and blood tests, and urine output measurements). The action refers to the treatment options (e.g., the dosage of medications such as sodium polystyrene sulfonate, and hemodialysis). The reward models the improvement in patient conditions. Similarly, reinforcement learning can help automate and optimize the dosage of immunosuppressive drugs in kidney transplants. Challenges and Opportunities Despite the success of reinforcement learning in several simplified clinical settings, their large-scale application to patient care faces several open challenges. The complexity of human biology complicates modeling clinical decision making as a reinforcement learning problem. The state space in such settings is often enormous, which could make a purely computational approach infeasible. Moreover, modeling all potential objectives a priori as a reward function may not be feasible. To overcome these challenges and realize the potential of reinforcement learning, clinical insight can play a pivotal role. More specifically, restricting the state space to only include highly relevant clinical variables could greatly reduce the computational complexity. Furthermore, using inverse reinforcement learning,2 relevant reward functions can be learned from retrospective studies assuming the optimality of clinical decisions. Another critical challenge is addressing moral and ethical concerns. It is imperative to ensure that reinforcement learning methods do not cause harm to the patient. To this end, there exists a need for a thorough validation of such methods before their use in patient care. Hence, there is a need to go beyond retrospective studies that have been used for the proof of concept of most existing reinforcement learning methods in health care applications.2,3 The lessons learned from the success of reinforcement learning in other application areas (e.g., self-driving cars) can help navigate the path to realizing its potential in health care. Accessible open-source simulation environments that enable researchers to compare various approaches are essential to the field of reinforcement learning. OpenAI Gym is currently the leading toolkit containing a wide range of simulated environments, e.g., surgical robotics.10 The development of high-quality and reliable simulation environments for nephrology and other health care applications can facilitate the development and validation of reinforcement learning methods beyond limited retrospective studies. The adoption of methods validated in such simulation environments in actual clinical settings will require clinicians' oversight. Similar to how self-driving cars require a human driver to ensure collision avoidance, clinicians' oversight is critical to ensure the safety of the patients, especially in the early stages of the adoption of reinforcement learning methods. The data from clinicians' decisions (e.g., overruling the automated treatment recommendation) can be used to improve the reliability of autonomous systems over time and reduce the burden of clinicians' oversight.

The steep rise in reinforcement learning (RL) in various applications in energy as well as the penetration of home automation in recent years are the motivation for this article. It surveys the use of RL in various home energy management system (HEMS) applications. There is a focus on deep neural network (DNN) models in RL. The article provides an overview of reinforcement learning. This is followed with discussions on state-of-the-art methods for value, policy, and actor–critic methods in deep reinforcement learning (DRL). In order to make the published literature in reinforcement learning more accessible to the HEMS community, verbal descriptions are accompanied with explanatory figures as well as mathematical expressions using standard machine learning terminology. Next, a detailed survey of how reinforcement learning is used in different HEMS domains is described. The survey also considers what kind of reinforcement learning algorithms are used in each HEMS application. It suggests that research in this direction is still in its infancy. Lastly, the article proposes four performance metrics to evaluate RL methods.

The relation between reinforcement learning parameters and the influence of reinforcement history on choice behavior

Reinforcement learning (RL) models have been broadly used in modeling the choice behavior of humans and other animals. In standard RL models, the action values are assumed to be updated according to the reward prediction error (RPE), i.e., the difference between the obtained reward and the expected reward. Numerous studies have noted that the magnitude of the update is biased depending on the sign of the RPE. The bias is represented in RL models by differential learning rates for positive and negative RPEs. However, which aspect of behavioral data that the estimated differential learning rates reflect is not well understood. In this study, we investigate how the differential learning rates influence the statistical properties of choice behavior (i.e., the relation between past experiences and the current choice) based on theoretical considerations and numerical simulations. We clarify that when the learning rates differ, the impact of a past outcome depends on the subsequent outcomes, in contrast to standard RL models with symmetric value updates. Based on the results, we propose a model-neutral statistical test to validate the hypothesis that value updates are asymmetric. The asymmetry in the value updates induces the autocorrelation of choice (i.e., the tendency to repeat the same choice or to switch the choice irrespective of past rewards). Conversely, if an RL model without an intrinsic autocorrelation factor is fitted to data that possess an intrinsic autocorrelation, a statistical bias to overestimate the difference in learning rates arises. We demonstrate that this bias can cause a statistical artifact in RL-model fitting leading to a “pseudo-positivity bias” and a “pseudo-confirmation bias.”

Experimental research such as cell cultures and carbon nanotube (CNT) growth are largely governed by following predefined execution protocols with fine-tuned control of parameters. There are promising opportunities to apply reinforcement learning (RL), an established learning technique in the area of artificial intelligence, in order to automate CNT growth process and accelerate related scientific breakthroughs in material discovery. Although there are benefits in RL-based exploration and exploitation methodologies, there are also challenges in developing relevant learning policies in experimental settings relating to CNT growth. In this paper, we present a novel data-driven RL approach for assisting experimental CNT growth. Our approach focuses on developing an RL model to learn from simulation-based images and characteristics of temporal CNT growth, considering various growth parameters. Our RL model learns from CNT growth variation in a simulation-based environment where critical control parameters, such as density, growth rate, tube radius, tube stiffness, and Van der Waals forces, are used. Our RL model enables CNT growth automation in order to explore of a wider range of growth conditions, and improve reproducibility. The ultimate goal of our RL model is to achieve desired CNT growth by dynamically controlling growth parameters throughout a sequence of experiments. We evaluate the effectiveness of our RL approach by measuring the improvement in the maximum compressive strength of a carbon nanotube ‘with’ and ‘without’ the RL model. Our results show the effectiveness of course correction recommended by our RL approach when controlling the parameters of angular deviation and rate of growth of carbon nanotubes, when compared against non-regulated CNT growth.

Healthcare applications pose significant challenges to existing reinforcement learning (RL) methods due to implementation risks, limited data availability, short treatment episodes, sparse rewards, partial observations, and heterogeneous treatment effects. Despite significant interest in using RL to generate dynamic treatment regimes for longitudinal patient care scenarios, no standardized benchmark has yet been developed. To fill this need we introduce Episodes of Care (EpiCare), a benchmark designed to mimic the challenges associated with applying RL to longitudinal healthcare settings. We leverage this benchmark to test five state-of-the-art offline RL models as well as five common off-policy evaluation (OPE) techniques. Our results suggest that while offline RL may be capable of improving upon existing standards of care given sufficient data, its applicability does not appear to extend to the moderate to low data regimes typical of current healthcare settings. Additionally, we demonstrate that several OPE techniques standard in the the medical RL literature fail to perform adequately on our benchmark. These results suggest that the performance of RL models in dynamic treatment regimes may be difficult to meaningfully evaluate using current OPE methods, indicating that RL for this application domain may still be in its early stages. We hope that these results along with the benchmark will facilitate better comparison of existing methods and inspire further research into techniques that increase the practical applicability of medical RL.

Fog computing extends cloud services to the network edge, enabling low-latency processing for Internet of Things (IoT) applications. However, this distributed approach is vulnerable to a wide range of attacks, necessitating advanced intrusion detection systems (IDSs) that operate under resource constraints. This study proposes integrating self-awareness (online learning and concept drift adaptation) into a lightweight RL (reinforcement learning)-based IDS for fog networks and quantitatively comparing it with non-RL static thresholds and bandit-based approaches in real time. Novel self-aware reinforcement learning (RL) models, the Hierarchical Adaptive Thompson Sampling–Reinforcement Learning (HATS-RL) model, and the Federated Hierarchical Adaptive Thompson Sampling–Reinforcement Learning (F-HATS-RL), were proposed for real-time intrusion detection in a fog network. These self-aware RL policies integrated online uncertainty estimation and concept-drift detection to adapt to evolving attacks. The RL models were benchmarked against the static threshold (ST) model and a widely adopted linear bandit (Linear Upper Confidence Bound/LinUCB). A realistic fog network simulator with heterogeneous nodes and streaming traffic, including multi-type attack bursts and gradual concept drift, was established. The models’ detection performance was compared using metrics including latency, energy consumption, detection accuracy, and the area under the precision–recall curve (AUPR) and the area under the receiver operating characteristic curve (AUROC). Notably, the federated self-aware agent (F-HATS-RL) achieved the best AUROC (0.933) and AUPR (0.857), with a latency of 0.27 ms and the lowest energy consumption of 0.0137 mJ, indicating its ability to detect intrusions in fog networks with minimal energy. The findings suggest that self-aware RL agents can detect traffic–dynamic attack methods and adapt accordingly, resulting in more stable long-term performance. By contrast, a static model’s accuracy degrades under drift.

Data quality is a critical aspect of data analytics since it directly influences the accuracy and effectiveness of insights and predictions generated from data. Artificial Intelligence (AI) schemes have grown in the existing era of technological advancement, which provides innovative exposure to healthcare applications. Reinforcement Learning (RL) is a subfield and an influential Machine Learning (ML) model aimed at optimizing decision-making by association with dynamic environments. In healthcare applications, RL can modify conduct strategies, enhance source application, and improve patient investigation history by using various data modalities. The worth of the data quality regulates how effective RL is in healthcare applications. In healthcare, the model predictions have a direct impact on patient's lives, and poor data quality often leads to wrong evaluations that expose patient safety and treatment quality. Biases in data quality have also presented a challenging influence on the RL model's effectiveness and accuracy. RL models have enormous potential in healthcare; however, various strategic limitations prevent their widespread acceptance and deployment. The implementation of RL in healthcare faces serious issues, mostly around data quality, bias, and tactical difficulties. This study delivers a broad survey of these challenges, emphasizing how imbalanced, imperfect, and biased data can affect the generalizability and performance of RL models. We critically assessed the sources of data bias, comprising demographic imbalances and irregularities in electronic health records (EHRs), and their impact on RL algorithms. This survey aims to present a detailed study of the complex circumstances relating to data quality, data biases, and strategic barriers in RL models deploying in healthcare applications. However, the main contribution of the proposed study is that it provides a systematic review of these challenges and delivers a roadmap for future work intended to refine the consistency, fairness, and scalability of RL in healthcare sectors.

Reinforcement learning (RL) models have been influential in characterizing human learning and decision making, but few studies apply them to characterizing human spatial navigation and even fewer systematically compare RL models under different navigation requirements. Because RL can characterize one’s learning strategies quantitatively and in a continuous manner, and one’s consistency of using such strategies, it can provide a novel and important perspective for understanding the marked individual differences in human navigation and disentangle navigation strategies from navigation performance. One-hundred and fourteen participants completed wayfinding tasks in a virtual environment where different phases manipulated navigation requirements. We compared performance of five RL models (3 model-free, 1 model-based and 1 “hybrid”) at fitting navigation behaviors in different phases. Supporting implications from prior literature, the hybrid model provided the best fit regardless of navigation requirements, suggesting the majority of participants rely on a blend of model-free (route-following) and model-based (cognitive mapping) learning in such navigation scenarios. Furthermore, consistent with a key prediction, there was a correlation in the hybrid model between the weight on model-based learning (i.e., navigation strategy) and the navigator’s exploration vs. exploitation tendency (i.e., consistency of using such navigation strategy), which was modulated by navigation task requirements. Together, we not only show how computational findings from RL align with the spatial navigation literature, but also reveal how the relationship between navigation strategy and a person’s consistency using such strategies changes as navigation requirements change.

Many model-based reinforcement learning (RL) methods follow a similar template: fit a model to previously observed data, and then use data from that model for RL or planning. However, models that achieve better training performance (e.g., lower MSE) are not necessarily better for control: an RL agent may seek out the small fraction of states where an accurate model makes mistakes, or it might act in ways that do not expose the errors of an inaccurate model. As noted in prior work, there is an objective mismatch: models are useful if they yield good policies, but they are trained to maximize their accuracy, rather than the performance of the policies that result from them. In this work, we propose a single objective for jointly training the model and the policy, such that updates to either component increase a lower bound on expected return. To the best of our knowledge, this is the first lower bound for model-based RL that holds globally and can be efficiently estimated in continuous settings; it is the only lower bound that mends the objective mismatch problem. A version of this bound becomes tight under certain assumptions. Optimizing this bound resembles a GAN: a classifier distinguishes between real and fake transitions, the model is updated to produce transitions that look realistic, and the policy is updated to avoid states where the model predictions are unrealistic. Numerical simulations demonstrate that optimizing this bound yields reward maximizing policies and yields dynamics that (perhaps surprisingly) can aid in exploration. We also show that a deep RL algorithm loosely based on our lower bound can achieve performance competitive with prior model-based methods, and better performance on certain hard exploration tasks.

This paper presents a method for trajectory generation with Bézier curves, by considering the wind field and dynamic constraints for small Unmanned Aerial Systems (UAS). Atmospheric phenomena affect UAS trajectories, so the wind presence should be considered in order efficiently perform a mission. These missions require precise area decomposition, efficient waypoint sequencing, and a smooth trajectory generation to fulfill the required level of safety and reliability. In this context, a path planning algorithm is presented which incorporates the use of an area decomposition scheme which considers complex shapes and restrictions, in order to provide high levels of coverage. A novel trajectory generation algorithm is proposed, which aims to harvest energy from the atmosphere by taking advantage of the previously estimated wind field and the identified wind features, such as shear wind and discrete gusts, by taking into account flight envelope restrictions. Different test cases and scenarios, including Software-In-The-Loop (SITL) simulations and real telemetry data analysis, are presented to assess the energy gain for the implemented algorithms. The results indicate promising energy gains of almost 15% in voltage saving if the mentioned factors are considered to re-plan the path with the energy-efficient trajectories.

The widespread deployment of smart meters and communication technologies brings opportunities to improve the adaptability and flexibility of future manufacturing systems under growing complexity and uncertainty. As a result, data-driven approaches, especially Reinforcement Learning (RL) methods, are gaining wide attention in recent years. Although in the current literature, RL approaches show superiority over traditional methods in many applications, a comprehensive review and comparison of different RL methods and their use in discrete manufacturing system control under changing environments have not yet been established. This paper first provides a literature review of RL algorithms for decision-making in discrete manufacturing systems, and then systematically discusses the underlying mechanisms of four most commonly used RL algorithms. In addition, the performance of these RL algorithms is compared by solving a production control problem, which aims to maximize the profit under time-varying production costs. The comparison results show that the single-agent RL methods with discrete action space can provide satisfactory results in small-scale systems, while multi-agent RL algorithms are more suitable to solve problems in complex, large-scale systems to remit the curse of dimensionality. This research sheds light on the impacts of RL method selection on the search for high-quality solutions in manufacturing systems, which is envisioned to drive future research on RL-supported manufacturing systems.

Stress and genetic background regulate different aspects of behavioral learning through the action of stress hormones and neuromodulators. In reinforcement learning (RL) models, meta-parameters such as learning rate, future reward discount factor, and exploitation-exploration factor, control learning dynamics and performance. They are hypothesized to be related to neuromodulatory levels in the brain. We found that many aspects of animal learning and performance can be described by simple RL models using dynamic control of the meta-parameters. To study the effects of stress and genotype, we carried out 5-hole-box light conditioning and Morris water maze experiments with C57BL/6 and DBA/2 mouse strains. The animals were exposed to different kinds of stress to evaluate its effects on immediate performance as well as on long-term memory. Then, we used RL models to simulate their behavior. For each experimental session, we estimated a set of model meta-parameters that produced the best fit between the model and the animal performance. The dynamics of several estimated meta-parameters were qualitatively similar for the two simulated experiments, and with statistically significant differences between different genetic strains and stress conditions.

Metamaterials are engineered composites that can exhibit acoustic, electromagnetic, elasto-dynamic and mechanical properties that are not found in natural materials Due to the complexity of the target objective functions, it is difficult to find the globally optimized solutions in the inverse design of metameterials. This thesis proposes and outlines two model, a gradient-based optimization method combined with generative networks (2D-GLOnets) and a reinforcement learning (RL) model, that can find the optimized metamaterial structures across a wide range of parameters. By perturbing the positions of each cylindrical scatterer in a planar configuration, 2D-GLOnets and the RL model with Deep Deterministic Policy Gradients (DDPG) are developed to design 2D (two-dimensional) broadband acoustic cloaking devices at the desired range of wavenumbers. Both models were implemented using PyTorch, Python libraries, and MATLAB engine. The numerical results are presented and compared with the optimal values produced by fmincon algorithms to verify the validation of our approaches. Our results indicated that both methods reduced the total scattering cross section by 17% to 90% compared to the initial conditions.