While robust optimal control theory provides a rigorous framework to compute robot control policies that are provably safe, it struggles to scale to high-dimensional problems, leading to increased use of deep learning for tractable synthesis of robot safety. Unfortunately, existing neural safety synthesis methods often lack convergence guarantees and solution interpretability. In this paper, we present Minimax Actors Guided by Implicit Critic Stackelberg (MAGICS), a novel adversarial reinforcement learning (RL) algorithm that guarantees local convergence to a minimax equilibrium solution. We then build on this approach to provide local convergence guarantees for a general deep RL-based robot safety synthesis algorithm. Through both simulation studies on OpenAI Gym environments and hardware experiments with a 36-dimensional quadruped robot, we show that MAGICS can yield robust control policies outperforming the state-of-the-art neural safety synthesis methods.

Preference-based reinforcement learning (PbRL) has shown significant promise for personalization in human-robot interaction (HRI) by explicitly integrating human preferences into the robot learning process. However, existing practices often require training a personalized robot policy from scratch, resulting in inefficient use of human feedback. In this paper, we propose preference-based action representation learning (PbARL), an efficient fine-tuning method that decouples common task structure from preference by leveraging pre-trained robot policies. Instead of directly fine-tuning the pre-trained policy with human preference, PbARL uses it as a reference for an action representation learning task that maximizes the mutual information between the pre-trained source domain and the target user preference-aligned domain. This approach allows the robot to personalize its behaviors while preserving original task performance and eliminates the need for extensive prior information from the source domain, thereby enhancing efficiency and practicality in real-world HRI scenarios. Empirical results on the Assistive Gym benchmark and a real-world user study (N=8) demonstrate the benefits of our method compared to state-of-the-art approaches.

Preference-based reinforcement learning (PbRL) shows promise in aligning robot behaviors with human preferences, but its success depends heavily on the accurate modeling of human preferences through reward models. Most methods adopt Markovian assumptions for preference modeling (PM), which overlook the temporal dependencies within robot behavior trajectories that impact human evaluations. While recent works have utilized sequence modeling to mitigate this by learning sequential non-Markovian rewards, they ignore the multimodal nature of robot trajectories, which consist of elements from two distinctive modalities: state and action. As a result, they often struggle to capture the complex interplay between these modalities that significantly shapes human preferences. In this paper, we propose a multimodal sequence modeling approach for PM by disentangling state and action modalities. We introduce a multimodal transformer network, named PrefMMT, which hierarchically leverages intra-modal temporal dependencies and inter-modal state-action interactions to capture complex preference patterns. We demonstrate that PrefMMT consistently outperforms state-of-the-art PM baselines on locomotion tasks from the D4RL benchmark and manipulation tasks from the Meta-World benchmark.

Parkour poses a significant challenge for legged robots, requiring navigation through complex environments with agility and precision based on limited sensory inputs. In this work, we introduce a novel method for training end-to-end visual policies, from depth pixels to robot control commands, to achieve agile and safe quadruped locomotion. We formulate robot parkour as a constrained reinforcement learning (RL) problem designed to maximize the emergence of agile skills within the robot's physical limits while ensuring safety. We first train a policy without vision using privileged information about the robot's surroundings. We then generate experience from this privileged policy to warm-start a sample efficient off-policy RL algorithm from depth images. This allows the robot to adapt behaviors from this privileged experience to visual locomotion while circumventing the high computational costs of RL directly from pixels. We demonstrate the effectiveness of our method on a real Solo-12 robot, showcasing its capability to perform a variety of parkour skills such as walking, climbing, leaping, and crawling.

This paper proposes an assembly sequence planning framework, named Subassembly to Assembly (S2A). The framework is designed to enable a robotic manipulator to assemble multiple parts in a prespecified structure by leveraging object manipulation actions. The primary technical challenge lies in the exponentially increasing complexity of identifying a feasible assembly sequence as the number of parts grows. To address this, we introduce a graph-based reinforcement learning approach, where a graph attention network is trained using a delayed reward assignment strategy. In this strategy, rewards are assigned only when an assembly action contributes to the successful completion of the assembly task. We validate the framework's performance through physics-based simulations, comparing it against various baselines to emphasize the significance of the proposed reward assignment approach. Additionally, we demonstrate the feasibility of deploying our framework in a real-world robotic assembly scenario.

Navigating in human-filled public spaces is a critical challenge for deploying autonomous robots in real-world environments. This paper introduces NaviDIFF, a novel Hamiltonian-constrained socially-aware navigation framework designed to address the complexities of human-robot interaction and socially-aware path planning. NaviDIFF integrates a port-Hamiltonian framework to model dynamic physical interactions and a diffusion model to manage uncertainty in human-robot cooperation. The framework leverages a spatial-temporal transformer to capture social and temporal dependencies, enabling more accurate pedestrian strategy predictions and port-Hamiltonian dynamics construction. Additionally, reinforcement learning from human feedback is employed to fine-tune robot policies, ensuring adaptation to human preferences and social norms. Extensive experiments demonstrate that NaviDIFF outperforms state-of-the-art methods in social navigation tasks, offering improved stability, efficiency, and adaptability.

Real-time dynamic scheduling is a crucial but notoriously challenging task in modern manufacturing processes due to its high decision complexity. Recently, reinforcement learning (RL) has been gaining attention as an impactful technique to handle this challenge. However, classical RL methods typically rely on human-made dispatching rules, which are not suitable for large-scale factory-wide scheduling. To bridge this gap, this paper applies a leader-follower multi-agent RL (MARL) concept to obtain desired coordination after decomposing the scheduling problem into a set of sub-problems that are handled by each individual agent for scalability. We further strengthen the procedure by proposing a rule-based conversion algorithm to prevent catastrophic loss of production capacity due to an agent's error. Our experimental results demonstrate that the proposed model outperforms the state-of-the-art deep RL-based scheduling models in various aspects. Additionally, the proposed model provides the most robust scheduling performance to demand changes. Overall, the proposed MARL-based scheduling model presents a promising solution to the real-time scheduling problem, with potential applications in various manufacturing industries.

LASTIC pollution in rivers has become a pressing global issue, with 11 million tons of plastic waste entering the ocean annually, 80% of which is caused by 1,000 major polluting rivers [1]. To address this problem, it is desired to develop a solution capable of removing plastic and other waste objects without interfering with the existing flora and fauna essential to river ecosystems [2] . Our Autonomous River Cleanup (ARC) project, initiated in 2019, leverages robotics and automation to remove plastic waste from rivers. In order to increase the capacity at which this can be done, we enhance the existing single arm sorting station [3] with additional robot arms. For multiple robot agents to efficiently sort waste on a conveyor belt, we develop and evaluate novel strategy algorithms using reinforcement learning that assign pick-and-place (PnP) tasks to the respective robot agents (Figure 1). Given a set of objects on the moving conveyor belt, the robot agents are tasked with removing waste objects, whilst bio-matter is ignored and collected at the end of the belt. The challenge is to allocate each robot optimally with PnP operations for objects within its reachable workspace.

In recent years, robots and autonomous systems have become increasingly integral to our daily lives, offering solutions to complex problems across various domains. Their application in search and rescue (SAR) operations, however, presents unique challenges. Comprehensively exploring the disaster-stricken area is often infeasible due to the vastness of the terrain, transformed environment, and the time constraints involved. Traditional robotic systems typically operate on predefined search patterns and lack the ability to incorporate and exploit ground truths provided by human stakeholders, which can be the key to speeding up the learning process and enhancing triage. Addressing this gap, we introduce a system that integrates social interaction via large language models (LLMs) with a hierarchical reinforcement learning (HRL) framework. The proposed system is designed to translate verbal inputs from human stakeholders into actionable RL insights and adjust its search strategy. By leveraging human-provided information through LLMs and structuring task execution through HRL, our approach not only bridges the gap between autonomous capabilities and human intelligence but also significantly improves the agent's learning efficiency and decision-making process in environments characterised by long horizons and sparse rewards.

Autonomous operations of robots in unknown environments are challenging due to the lack of knowledge of the dynamics of the interactions, such as the objects' movability. This work introduces a novel Causal Reinforcement Learning approach to enhancing robotics operations and applies it to an urban search and rescue (SAR) scenario. Our proposed machine learning architecture enables robots to learn the causal relationships between the visual characteristics of the objects, such as texture and shape, and the objects' dynamics upon interaction, such as their movability, significantly improving their decision-making processes. We conducted causal discovery and RL experiments demonstrating the Causal RL's superior performance, showing a notable reduction in learning times by over 24.5% in complex situations, compared to non-causal models.