Safe navigation is essential for autonomous systems operating in hazardous environments. Traditional planning methods excel at long-horizon tasks but rely on a predefined graph with fixed distance metrics. In contrast, safe Reinforcement Learning (RL) can learn complex behaviors without relying on manual heuristics but fails to solve long-horizon tasks, particularly in goal-conditioned and multi-agent scenarios. In this paper, we introduce a novel method that integrates the strengths of both planning and safe RL. Our method leverages goal-conditioned RL and safe RL to learn a goal-conditioned policy for navigation while concurrently estimating cumulative distance and safety levels using learned value functions via an automated self-training algorithm. By constructing a graph with states from the replay buffer, our method prunes unsafe edges and generates a waypoint-based plan that the agent follows until reaching its goal, effectively balancing faster and safer routes over extended distances. Utilizing this unified high-level graph and a shared low-level goal-conditioned safe RL policy, we extend this approach to address the multi-agent safe navigation problem. In particular, we leverage Conflict-Based Search (CBS) to create waypoint-based plans for multiple agents allowing for their safe navigation over extended horizons. This integration enhances the scalability of goal-conditioned safe RL in multi-agent scenarios, enabling efficient coordination among agents. Extensive benchmarking against state-of-the-art baselines demonstrates the effectiveness of our method in achieving distance goals safely for multiple agents in complex and hazardous environments. Our code and further details about or work is available at https://safe-visual-mapf-mers.csail.mit.edu/.
The primary goal of reinforcement learning is to develop decision-making policies that prioritize optimal performance, frequently without considering risk or safety. In contrast, safe reinforcement learning seeks to reduce or avoid unsafe states. This letter introduces a risk-averse temporal difference algorithm that uses optimal transport theory to direct the agent toward predictable behavior. By incorporating a risk indicator, the agent learns to favor actions with predictable consequences. We evaluate the proposed algorithm in several case studies and show its effectiveness in the presence of uncertainty. The results demonstrate that our method reduces the frequency of visits to risky states while preserving performance. A Python implementation of the algorithm is available at https:// github.com/SAILRIT/Risk-averse-TD-Learning.
Goal-Conditioned Reinforcement Learning (GCRL) provides a versatile framework for developing unified controllers capable of handling wide ranges of tasks, exploring environments, and adapting behaviors. However, its reliance on trial-and-error poses challenges for real-world applications, as errors can result in costly and potentially damaging consequences. To address the need for safer learning, we propose a method that enables agents to learn goal-conditioned behaviors that explore without the risk of making harmful mistakes. Exploration without risks can seem paradoxical, but environment dynamics are often uniform in space, therefore a policy trained for safety without exploration purposes can still be exploited globally. Our proposed approach involves two distinct phases. First, during a pretraining phase, we employ safe reinforcement learning and distributional techniques to train a safety policy that actively tries to avoid failures in various situations. In the subsequent safe exploration phase, a goal-conditioned (GC) policy is learned while ensuring safety. To achieve this, we implement an action-selection mechanism leveraging the previously learned distributional safety critics to arbitrate between the safety policy and the GC policy, ensuring safe exploration by switching to the safety policy when needed. We evaluate our method in simulated environments and demonstrate that it not only provides substantial coverage of the goal space but also reduces the occurrence of mistakes to a minimum, in stark contrast to traditional GCRL approaches. Additionally, we conduct an ablation study and analyze failure modes, offering insights for future research directions.
The urgent energy transition requirements towards a sustainable future stretch across various industries and are a significant challenge facing humanity. Hydrogen promises a clean, carbon-free future, with the opportunity to integrate with existing solutions in the transportation sector. However, adding hydrogen to existing technologies such as diesel engines requires additional modeling effort. Reinforcement Learning (RL) enables interactive data-driven learning that eliminates the need for mathematical modeling. The algorithms, however, may not be real-time capable and need large amounts of data to work in practice. This paper presents a novel approach which uses offline model learning with RL to demonstrate safe control of a 4.5 L Hydrogen Diesel Dual-Fuel (H2DF) engine. The controllers are demonstrated to be constraint compliant and can leverage a novel state-augmentation approach for sample-efficient learning. The offline policy is subsequently experimentally validated on the real engine where the control algorithm is executed on a Raspberry Pi controller and requires 6 times less computation time compared to online Model Predictive Control (MPC) optimization.
Safe reinforcement learning has traditionally relied on predefined constraint functions to ensure safety in complex real-world tasks, such as autonomous driving. However, defining these functions accurately for varied tasks is a persistent challenge. Recent research highlights the potential of leveraging pre-acquired task-agnostic knowledge to enhance both safety and sample efficiency in related tasks. Building on this insight, we propose a novel method to learn shared constraint distributions across multiple tasks. Our approach identifies the shared constraints through imitation learning and then adapts to new tasks by adjusting risk levels within these learned distributions. This adaptability addresses variations in risk sensitivity stemming from expert-specific biases, ensuring consistent adherence to general safety principles even with imperfect demonstrations. Our method can be applied to control and navigation domains, including multi-task and meta-task scenarios, accommodating constraints such as maintaining safe distances or adhering to speed limits. Experimental results validate the efficacy of our approach, demonstrating superior safety performance and success rates compared to baselines, all without requiring task-specific constraint definitions. These findings underscore the versatility and practicality of our method across a wide range of real-world tasks.
This work introduces a toolchain for applying Reinforcement Learning (RL), specifically the Deep Deterministic Policy Gradient (DDPG) algorithm, in safety-critical real-world environments. As an exemplary application, transient load control is demonstrated on a single-cylinder internal combustion engine testbench in Homogeneous Charge Compression Ignition (HCCI) mode, that offers high thermal efficiency and low emissions. However, HCCI poses challenges for traditional control methods due to its nonlinear, autoregressive, and stochastic nature. RL provides a viable solution, however, safety concerns, such as excessive pressure rise rates, must be addressed when applying to HCCI. A single unsuitable control input can severely damage the engine or cause misfiring and shut down. Additionally, operating limits are not known a priori and must be determined experimentally. To mitigate these risks, real-time safety monitoring based on the k-nearest neighbor algorithm is implemented, enabling safe interaction with the testbench. The feasibility of this approach is demonstrated as the RL agent learns a control policy through interaction with the testbench. A root mean square error of 0.1374 bar is achieved for the indicated mean effective pressure, comparable to neural network-based controllers from the literature. The toolchain's flexibility is further demonstrated by adapting the agent's policy to increase ethanol energy shares, promoting renewable fuel use while maintaining safety. This RL approach addresses the longstanding challenge of applying RL to safety-critical real-world environments. The developed toolchain, with its adaptability and safety mechanisms, paves the way for future applicability of RL in engine testbenches and other safety-critical settings.
Reinforcement learning (RL) in the real world necessitates the development of procedures that enable agents to explore without causing harm to themselves or others. The most successful solutions to the problem of safe RL leverage offline data to learn a safe-set, enabling safe online exploration. However, this approach to safe-learning is often constrained by the demonstrations that are available for learning. In this paper we investigate the influence of the quantity and quality of data used to train the initial safe learning problem offline on the ability to learn safe-RL policies online. Specifically, we focus on tasks with spatially extended goal states where we have few or no demonstrations available. Classically this problem is addressed either by using hand-designed controllers to generate data or by collecting user-generated demonstrations. However, these methods are often expensive and do not scale to more complex tasks and environments. To address this limitation we propose an unsupervised RL-based offline data collection procedure, to learn complex and scalable policies without the need for hand-designed controllers or user demonstrations. Our research demonstrates the significance of providing sufficient demonstrations for agents to learn optimal safe-RL policies online, and as a result, we propose optimistic forgetting, a novel online safe-RL approach that is practical for scenarios with limited data. Further, our unsupervised data collection approach highlights the need to balance diversity and optimality for safe online exploration.
Many real-world multiagent learning problems involve safety concerns. In these setups, typical safe reinforcement learning algorithms constrain agents' behavior, limiting exploration -- a crucial component for discovering effective cooperative multiagent behaviors. Moreover, the multiagent literature typically models individual constraints for each agent and has yet to investigate the benefits of using joint team constraints. In this work, we analyze these team constraints from a theoretical and practical perspective and propose entropic exploration for constrained multiagent reinforcement learning (E2C) to address the exploration issue. E2C leverages observation entropy maximization to incentivize exploration and facilitate learning safe and effective cooperative behaviors. Experiments across increasingly complex domains show that E2C agents match or surpass common unconstrained and constrained baselines in task performance while reducing unsafe behaviors by up to $50\%$.
Offline safe reinforcement learning (OSRL) involves learning a decision-making policy to maximize rewards from a fixed batch of training data to satisfy pre-defined safety constraints. However, adapting to varying safety constraints during deployment without retraining remains an under-explored challenge. To address this challenge, we introduce constraint-adaptive policy switching (CAPS), a wrapper framework around existing offline RL algorithms. During training, CAPS uses offline data to learn multiple policies with a shared representation that optimize different reward and cost trade-offs. During testing, CAPS switches between those policies by selecting at each state the policy that maximizes future rewards among those that satisfy the current cost constraint. Our experiments on 38 tasks from the DSRL benchmark demonstrate that CAPS consistently outperforms existing methods, establishing a strong wrapper-based baseline for OSRL. The code is publicly available at https://github.com/yassineCh/CAPS.
Offline safe reinforcement learning (RL) has emerged as a promising approach for learning safe behaviors without engaging in risky online interactions with the environment. Most existing methods in offline safe RL rely on cost constraints at each time step (derived from global cost constraints) and this can result in either overly conservative policies or violation of safety constraints. In this paper, we propose to learn a policy that generates desirable trajectories and avoids undesirable trajectories. To be specific, we first partition the pre-collected dataset of state-action trajectories into desirable and undesirable subsets. Intuitively, the desirable set contains high reward and safe trajectories, and undesirable set contains unsafe trajectories and low-reward safe trajectories. Second, we learn a policy that generates desirable trajectories and avoids undesirable trajectories, where (un)desirability scores are provided by a classifier learnt from the dataset of desirable and undesirable trajectories. This approach bypasses the computational complexity and stability issues of a min-max objective that is employed in existing methods. Theoretically, we also show our approach's strong connections to existing learning paradigms involving human feedback. Finally, we extensively evaluate our method using the DSRL benchmark for offline safe RL. Empirically, our method outperforms competitive baselines, achieving higher rewards and better constraint satisfaction across a wide variety of benchmark tasks.
Real-world accidents in learning-enabled CPS frequently occur in challenging corner cases. During the training of deep reinforcement learning (DRL) policy, the standard setup for training conditions is either fixed at a single initial condition or uniformly sampled from the admissible state space. This setup often overlooks the challenging but safety-critical corner cases. To bridge this gap, this paper proposes a physics-model-guided worst-case sampling strategy for training safe policies that can handle safety-critical cases toward guaranteed safety. Furthermore, we integrate the proposed worst-case sampling strategy into the physics-regulated deep reinforcement learning (Phy-DRL) framework to build a more data-efficient and safe learning algorithm for safety-critical CPS. We validate the proposed training strategy with Phy-DRL through extensive experiments on a simulated cart-pole system, a 2D quadrotor, a simulated and a real quadruped robot, showing remarkably improved sampling efficiency to learn more robust safe policies.
Safe reinforcement learning (RL) is a popular and versatile paradigm to learn reward-maximizing policies with safety guarantees. Previous works tend to express the safety constraints in an expectation form due to the ease of implementation, but this turns out to be ineffective in maintaining safety constraints with high probability. To this end, we move to the quantile-constrained RL that enables a higher level of safety without any expectation-form approximations. We directly estimate the quantile gradients through sampling and provide the theoretical proofs of convergence. Then a tilted update strategy for quantile gradients is implemented to compensate the asymmetric distributional density, with a direct benefit of return performance. Experiments demonstrate that the proposed model fully meets safety requirements (quantile constraints) while outperforming the state-of-the-art benchmarks with higher return.
A key aspect of Safe Reinforcement Learning (Safe RL) involves estimating the constraint condition for the next policy, which is crucial for guiding the optimization of safe policy updates. However, the existing Advantage-based Estimation (ABE) method relies on the infinite-horizon discounted advantage function. This dependence leads to catastrophic errors in finite-horizon scenarios with non-discounted constraints, resulting in safety-violation updates. In response, we propose the first estimation method for finite-horizon non-discounted constraints in deep Safe RL, termed Gradient-based Estimation (GBE), which relies on the analytic gradient derived along trajectories. Our theoretical and empirical analyses demonstrate that GBE can effectively estimate constraint changes over a finite horizon. Constructing a surrogate optimization problem with GBE, we developed a novel Safe RL algorithm called Constrained Gradient-based Policy Optimization (CGPO). CGPO identifies feasible optimal policies by iteratively resolving sub-problems within trust regions. Our empirical results reveal that CGPO, unlike baseline algorithms, successfully estimates the constraint functions of subsequent policies, thereby ensuring the efficiency and feasibility of each update.
In safe offline reinforcement learning (RL), the objective is to develop a policy that maximizes cumulative rewards while strictly adhering to safety constraints, utilizing only offline data. Traditional methods often face difficulties in balancing these constraints, leading to either diminished performance or increased safety risks. We address these issues with a novel approach that begins by learning a conservatively safe policy through the use of Conditional Variational Autoencoders, which model the latent safety constraints. Subsequently, we frame this as a Constrained Reward-Return Maximization problem, wherein the policy aims to optimize rewards while complying with the inferred latent safety constraints. This is achieved by training an encoder with a reward-Advantage Weighted Regression objective within the latent constraint space. Our methodology is supported by theoretical analysis, including bounds on policy performance and sample complexity. Extensive empirical evaluation on benchmark datasets, including challenging autonomous driving scenarios, demonstrates that our approach not only maintains safety compliance but also excels in cumulative reward optimization, surpassing existing methods. Additional visualizations provide further insights into the effectiveness and underlying mechanisms of our approach.
Safe reinforcement learning (RL) requires the agent to finish a given task while obeying specific constraints. Giving constraints in natural language form has great potential for practical scenarios due to its flexible transfer capability and accessibility. Previous safe RL methods with natural language constraints typically need to design cost functions manually for each constraint, which requires domain expertise and lacks flexibility. In this paper, we harness the dual role of text in this task, using it not only to provide constraint but also as a training signal. We introduce the Trajectory-level Textual Constraints Translator (TTCT) to replace the manually designed cost function. Our empirical results demonstrate that TTCT effectively comprehends textual constraint and trajectory, and the policies trained by TTCT can achieve a lower violation rate than the standard cost function. Extra studies are conducted to demonstrate that the TTCT has zero-shot transfer capability to adapt to constraint-shift environments.
The high costs and risks involved in extensive environment interactions hinder the practical application of current online safe reinforcement learning (RL) methods. While offline safe RL addresses this by learning policies from static datasets, the performance therein is usually limited due to reliance on data quality and challenges with out-of-distribution (OOD) actions. Inspired by recent successes in offline-to-online (O2O) RL, it is crucial to explore whether offline safe RL can be leveraged to facilitate faster and safer online policy learning, a direction that has yet to be fully investigated. To fill this gap, we first demonstrate that naively applying existing O2O algorithms from standard RL would not work well in the safe RL setting due to two unique challenges: \emph{erroneous Q-estimations}, resulted from offline-online objective mismatch and offline cost sparsity, and \emph{Lagrangian mismatch}, resulted from difficulties in aligning Lagrange multipliers between offline and online policies. To address these challenges, we introduce \textbf{Marvel}, a novel framework for O2O safe RL, comprising two key components that work in concert: \emph{Value Pre-Alignment} to align the Q-functions with the underlying truth before online learning, and \emph{Adaptive PID Control} to effectively adjust the Lagrange multipliers during online finetuning. Extensive experiments demonstrate that Marvel significantly outperforms existing baselines in both reward maximization and safety constraint satisfaction. By introducing the first policy-finetuning based framework for O2O safe RL, which is compatible with many offline and online safe RL methods, our work has the great potential to advance the field towards more efficient and practical safe RL solutions.
As large-scale distributed energy resources are integrated into the active distribution networks (ADNs), effective energy management in ADNs becomes increasingly prominent compared to traditional distribution networks. Although advanced reinforcement learning (RL) methods, which alleviate the burden of complicated modelling and optimization, have greatly improved the efficiency of energy management in ADNs, safety becomes a critical concern for RL applications in real-world problems. Since the design and adjustment of penalty functions, which correspond to operational safety constraints, requires extensive domain knowledge in RL and power system operation, the emerging ADN operators call for a more flexible and customized approach to address the penalty functions so that the operational safety and efficiency can be further enhanced. Empowered with strong comprehension, reasoning, and in-context learning capabilities, large language models (LLMs) provide a promising way to assist safe RL for energy management in ADNs. In this paper, we introduce the LLM to comprehend operational safety requirements in ADNs and generate corresponding penalty functions. In addition, we propose an RL2 mechanism to refine the generated functions iteratively and adaptively through multi-round dialogues, in which the LLM agent adjusts the functions' pattern and parameters based on training and test performance of the downstream RL agent. The proposed method significantly reduces the intervention of the ADN operators. Comprehensive test results demonstrate the effectiveness of the proposed method.
This paper addresses the target-pursuit problem, aiming to ensure each pursuer's safety regarding collision avoidance, sensing range, and input saturation. An input-constrained CBF is proposed to dynamically regulate the pursuer's control, ensuring effective target pursuit even when the target performs evasive maneuvers. To further ensure safety, two sets of CBF constraints are designed to regulate the pursuer's position, enabling it to keep the target within the sensing range while avoiding collision in complex environments with external disturbances. These three CBFs collectively form our safety filter, which filters unsafe outputs from RL by solving a Quadratic Program (QP). Finally, the safety filter, combined with a switch strategy that enhances the feasibility of solving its QP, constitutes the Control Barrier Function (CBF)-Safe Reinforcement Learning (CSRL) algorithm, whose solutions are proven to satisfy the Karush-Kuhn-Tucker (KKT) conditions for all safety constraints. Simulation results validate the effectiveness of the CSRL algorithm, demonstrating its ability to handle complex pursuit scenarios while maintaining safety and improving control performance.
Safe reinforcement learning (RL) is crucial for real-world applications, and multi-agent interactions introduce additional safety challenges. While Probabilistic Logic Shields (PLS) has been a powerful proposal to enforce safety in single-agent RL, their generalizability to multi-agent settings remains unexplored. In this paper, we address this gap by conducting extensive analyses of PLS within decentralized, multi-agent environments, and in doing so, propose Shielded Multi-Agent Reinforcement Learning (SMARL) as a general framework for steering MARL towards norm-compliant outcomes. Our key contributions are: (1) a novel Probabilistic Logic Temporal Difference (PLTD) update for shielded, independent Q-learning, which incorporates probabilistic constraints directly into the value update process; (2) a probabilistic logic policy gradient method for shielded PPO with formal safety guarantees for MARL; and (3) comprehensive evaluation across symmetric and asymmetrically shielded $n$-player game-theoretic benchmarks, demonstrating fewer constraint violations and significantly better cooperation under normative constraints. These results position SMARL as an effective mechanism for equilibrium selection, paving the way toward safer, socially aligned multi-agent systems.
In this paper we propose a formal, model-agnostic meta-learning framework for safe reinforcement learning. Our framework is inspired by how parents safeguard their children across a progression of increasingly riskier tasks, imparting a sense of safety that is carried over from task to task. We model this as a meta-learning process where each task is synchronized with a safeguard that monitors safety and provides a reward signal to the agent. The safeguard is implemented as a finite-state machine based on a safety specification; the reward signal is formally shaped around this specification. The safety specification and its corresponding safeguard can be arbitrarily complex and non-Markovian, which adds flexibility to the training process and explainability to the learned policy. The design of the safeguard is manual but it is high-level and model-agnostic, which gives rise to an end-to-end safe learning approach with wide applicability, from pixel-level game control to language model fine-tuning. Starting from a given set of safety specifications (tasks), we train a model such that it can adapt to new specifications using only a small number of training samples. This is made possible by our method for efficiently transferring safety bias between tasks, which effectively minimizes the number of safety violations. We evaluate our framework in a Minecraft-inspired Gridworld, a VizDoom game environment, and an LLM fine-tuning application. Agents trained with our approach achieve near-minimal safety violations, while baselines are shown to underperform.