The success of intelligent robotic missions relies on integrating various research tasks, each demanding distinct representations. Designing task-specific representations for each task is costly and impractical. Unified representations suitable for multiple tasks remain unexplored. My outline introduces a series of research outcomes of GP-based probabilistic distance field (GPDF) representation that mathematically models the fundamental property of Euclidean distance field (EDF) along with gradients, surface normals and dense reconstruction. The progress to date and ongoing future works show that GPDF has the potential to offer a unified solution of representation for multiple tasks such as localisation, mapping, motion planning, obstacle avoidance, grasping, human-robot collaboration, and dense visualisation. I believe that GPDF serves as the cornerstone for robots to accomplish more complex and challenging tasks. By leveraging GPDF, robots can navigate through intricate environments, understand spatial relationships, and interact with objects and humans seamlessly.

Despite being successful in board games and reinforcement learning (RL), UCT, a Monte-Carlo Tree Search (MCTS) combined with UCB1 Multi-Armed Bandit (MAB), has had limited success in domain-independent planning until recently. Previous work showed that UCB1, designed for $[0,1]$-bounded rewards, is not appropriate for estimating the distance-to-go which are potentially unbounded in $\mathbb{R}$, such as heuristic functions used in classical planning, then proposed combining MCTS with MABs designed for Gaussian reward distributions and successfully improved the performance. In this paper, we further sharpen our understanding of ideal bandits for planning tasks. Existing work has two issues: First, while Gaussian MABs no longer over-specify the distances as $h\in [0,1]$, they under-specify them as $h\in [-\infty,\infty]$ while they are non-negative and can be further bounded in some cases. Second, there is no theoretical justifications for Full-Bellman backup (Schulte & Keller, 2014) that backpropagates minimum/maximum of samples. We identified \emph{extreme value} statistics as a theoretical framework that resolves both issues at once and propose two bandits, UCB1-Uniform/Power, and apply them to MCTS for classical planning. We formally prove their regret bounds and empirically demonstrate their performance in classical planning.

Autonomous navigation across unstructured terrains, including forests and construction areas, faces unique challenges due to intricate obstacles and the element of the unknown. Lacking pre-existing maps, these scenarios necessitate a motion planning approach that combines agility with efficiency. Critically, it must also incorporate the robot's kinematic constraints to navigate more effectively through complex environments. This work introduces a novel planning method for center-articulated vehicles (CAV), leveraging motion primitives within a receding horizon planning framework using onboard sensing. The approach commences with the offline creation of motion primitives, generated through forward simulations that reflect the distinct kinematic model of center-articulated vehicles. These primitives undergo evaluation through a heuristic-based scoring function, facilitating the selection of the most suitable path for real-time navigation. To augment this planning process, we develop a pose-stabilizing controller, tailored to the kinematic specifications of center-articulated vehicles. During experiments, our method demonstrates a $67\%$ improvement in SPL (Success Rate weighted by Path Length) performance over existing strategies. Furthermore, its efficacy was validated through real-world experiments conducted with a tree harvester vehicle - SAHA.

Efficient path planning is key for safe autonomous navigation over complex and unknown terrains. Lunar Zebro (LZ), a project of the Delft University of Technology, aims to deploy a compact rover, no larger than an A4 sheet of paper and weighing not more than 3 kilograms. In this work, we introduce a Robust Artificial Potential Field (RAPF) algorithm, a new path-planning algorithm for reliable local navigation solution for lunar microrovers. RAPF leverages and improves state of the art Artificial Potential Field (APF)-based methods by incorporating the position of the robot in the generation of bacteria points and considering local minima as regions to avoid. We perform both simulations and on field experiments to validate the performance of RAPF, which outperforms state-of-the-art APF-based algorithms by over 15% in reachability within a similar or shorter planning time. The improvements resulted in a 200% higher success rate and 50% lower computing time compared to the conventional APF algorithm. Near-optimal paths are computed in real-time with limited available processing power. The bacterial approach of the RAPF algorithm proves faster to execute and smaller to store than path planning algorithms used in existing planetary rovers, showcasing its potential for reliable lunar exploration with computationally constrained and energy constrained robotic systems.

Abstract-- Till now, many path planning algorithms have been proposed in the literature. The objective of these algorithms is to find the quickest path between initial position to the end position in a certain environment. The complexity of these algorithms depends on the internal parameters such as motor speed or sensor range and on other external parameters, including the accuracy of the map, size of the environment, and the number of obstacles. In this paper, we are giving information about how path planning algorithm finds the optimal path in an uneven terrain with a multiple obstacle using TurtleBot3 robot into the Gazebo environment using Dijkstra's and A*. A fundamental task for any mobile robot is its capability to organize collision-free trajectories from point A to point B, a start to a end position or visiting a series of positions, i.e. When provided a map and a end point, path The focus of this project is path-planning, which involves planning includes selecting the best (collision free, if finding the optimal or near-optimal path from point A to point applicable) trajectory that the robot can follow to reach the goal B. The aim here is to find the shortest path, optimally, that position.

When performing manipulation-based activities such as picking objects, a mobile robot needs to position its base at a location that supports successful execution. To address this problem, prominent approaches typically rely on costly grasp planners to provide grasp poses for a target object, which are then are then analysed to identify the best robot placements for achieving each grasp pose. In this paper, we propose instead to first find robot placements that would not result in collision with the environment and from where picking up the object is feasible, then evaluate them to find the best placement candidate. Our approach takes into account the robot's reachability, as well as RGB-D images and occupancy grid maps of the environment for identifying suitable robot poses. The proposed algorithm is embedded in a service robotic workflow, in which a person points to select the target object for grasping. We evaluate our approach with a series of grasping experiments, against an existing baseline implementation that sends the robot to a fixed navigation goal. The experimental results show how the approach allows the robot to grasp the target object from locations that are very challenging to the baseline implementation.

We give a model-based agent that builds a Python program representing its knowledge of the world based on its interactions with the environment. The world model tries to explain its interactions, while also being optimistic about what reward it can achieve. We define this optimism as a logical constraint between a program and a planner. We study our agent on gridworlds, and on task planning, finding our approach is more sample-efficient compared to deep RL, more compute-efficient compared to ReAct-style agents, and that it can transfer its knowledge across environments by editing its code.

A central challenge in quantum information science and technology is achieving real-time estimation and feedforward control of quantum systems. This challenge is compounded by the inherent inhomogeneity of quantum resources, such as qubit properties and controls, and their intrinsically probabilistic nature. This leads to stochastic challenges in error detection and probabilistic outcomes in processes such as heralded remote entanglement. Given these complexities, optimizing the construction of quantum resource states is an NP-hard problem. In this paper, we address the quantum resource scheduling issue by formulating the problem and simulating it within a digitized environment, allowing the exploration and development of agent-based optimization strategies. We employ reinforcement learning agents within this probabilistic setting and introduce a new framework utilizing a Transformer model that emphasizes self-attention mechanisms for pairs of qubits. This approach facilitates dynamic scheduling by providing real-time, next-step guidance. Our method significantly improves the performance of quantum systems, achieving more than a 3$\times$ improvement over rule-based agents, and establishes an innovative framework that improves the joint design of physical and control systems for quantum applications in communication, networking, and computing.

This paper addresses the challenge of collision-free motion planning in automated navigation within complex environments. Utilizing advancements in Deep Reinforcement Learning (DRL) and sensor technologies like LiDAR, we propose the TD3-DWA algorithm, an innovative fusion of the traditional Dynamic Window Approach (DWA) with the Twin Delayed Deep Deterministic Policy Gradient (TD3). This hybrid algorithm enhances the efficiency of robotic path planning by optimizing the sampling interval parameters of DWA to effectively navigate around both static and dynamic obstacles. The performance of the TD3-DWA algorithm is validated through various simulation experiments, demonstrating its potential to significantly improve the reliability and safety of autonomous navigation systems.

As the trend of moving away from high-precision maps gradually emerges in the autonomous driving industry,traditional planning algorithms are gradually exposing some problems. To address the high real-time, high precision, and high trajectory quality requirements posed by the automatic parking task under real-time perceived local maps,this paper proposes an improved automatic parking planning algorithm based on the A* algorithm, and uses Model Predictive Control (MPC) as the control module for automatic parking.The algorithm enhances the planning real-time performance by optimizing heuristic functions, binary heap optimization, and bidirectional search; it calculates the passability of narrow areas by dynamically loading obstacles and introduces the vehicle's own volume during planning; it improves trajectory quality by using neighborhood expansion and Bezier curve optimization methods to meet the high trajectory quality requirements of the parking task. After obtaining the output results of the planning algorithm, a loss function is designed according to the characteristics of the automatic parking task under local maps, and the MPC algorithm is used to output control commands to drive the car along the planned trajectory. This paper uses the perception results of real driving environments converted into maps as planning inputs to conduct simulation tests and ablation experiments on the algorithm. Experimental results show that the improved algorithm proposed in this paper can effectively meet the special requirements of automatic parking under local maps and complete the automatic parking planning and control tasks.