Sampling-based methods are widely adopted solutions for robot motion planning. The methods are straightforward to implement, effective in practice for many robotic systems. It is often possible to prove that they have desirable properties, such as probabilistic completeness and asymptotic optimality. Nevertheless, they still face challenges as the complexity of the underlying planning problem increases, especially under tight computation time constraints, which impact the quality of returned solutions or given inaccurate models. This has motivated machine learning to improve the computational efficiency and applicability of Sampling-Based Motion Planners (SBMPs). This survey reviews such integrative efforts and aims to provide a classification of the alternative directions that have been explored in the literature. It first discusses how learning has been used to enhance key components of SBMPs, such as node sampling, collision detection, distance or nearest neighbor computation, local planning, and termination conditions. Then, it highlights planners that use learning to adaptively select between different implementations of such primitives in response to the underlying problem's features. It also covers emerging methods, which build complete machine learning pipelines that reflect the traditional structure of SBMPs. It also discusses how machine learning has been used to provide data-driven models of robots, which can then be used by a SBMP. Finally, it provides a comparative discussion of the advantages and disadvantages of the approaches covered, and insights on possible future directions of research. An online version of this survey can be found at: https://prx-kinodynamic.github.io/

As robots become increasingly capable of manipulation and long-term autonomy, long-horizon task and motion planning problems are becoming increasingly important. A key challenge in such problems is that early actions in the plan may make future actions infeasible. When reaching a dead-end in the search, most existing planners use backtracking, which exhaustively reevaluates motion-level actions, often resulting in inefficient planning, especially when the search depth is large. In this paper, we propose to learn backjumping heuristics which identify the culprit action directly using supervised learning models to guide the task-level search. Based on evaluations on two different tasks, we find that our method significantly improves planning efficiency compared to backtracking and also generalizes to problems with novel numbers of objects.

Abstract: The key to Black-Box Optimization is to efficiently search through input regions with potentially widely-varying numerical properties, to achieve low-regret descent and fast progress toward the optima. Monte Carlo Tree Search (MCTS) methods have recently been introduced to improve Bayesian optimization by computing better partitioning of the search space that balances exploration and exploitation. Extending this promising framework, we study how to further integrate sample-based descent for faster optimization. We design novel ways of expanding Monte Carlo search trees, with new descent methods at vertices that incorporate stochastic search and Gaussian Processes. We propose the corresponding rules for balancing progress and uncertainty, branch selection, tree expansion, and backpropagation.

Mapping and exploration of a Martian terrain with an aerial vehicle has become an emerging research direction, since the successful flight demonstration of the Mars helicopter Ingenuity. Although the autonomy and navigation capability of the state of the art Mars helicopter has proven to be efficient in an open environment, the next area of interest for exploration on Mars are caves or ancient lava tube like environments, especially towards the never-ending search of life on other planets. This article presents an autonomous exploration mission based on a modified frontier approach along with a risk aware planning and integrated collision avoidance scheme with a special focus on energy aspects of a custom designed Mars Coaxial Quadrotor (MCQ) in a Martian simulated lava tube. One of the biggest novelties of the article stems from addressing the exploration capability, while rapidly exploring in local areas and intelligently global re-positioning of the MCQ when reaching dead ends in order to to efficiently use the battery based consumed energy, while increasing the volume of the exploration. The proposed novel algorithm for the Martian exploration is able to select the next way point of interest, such that the MCQ keeps its heading towards the local exploration direction where it will find maximum information about the surroundings. The proposed three layer cost based global re-position point selection assists in rapidly redirecting the MCQ to previously partially seen areas that could lead to more unexplored part of the lava tube. The Martian fully simulated mission presented in this article takes into consideration the fidelity of physics of Mars condition in terms of thin atmosphere, low surface pressure and low gravity of the planet, while proves the efficiency of the proposed scheme in exploring an area that is particularly challenging due to the subterranean-like environment. The proposed exploration-planning framework is also validated in simulation by comparing it against the graph based exploration planner. Intensive simulations with true Mars conditions are carried out in order to validate and benchmark our approach in a utmost realistic Mars lava tube exploration scenario using a Mars Coaxial Quadrotor.

This paper explores a rapid, optimal smooth path-planning algorithm for robots (e.g., autonomous vehicles) in point cloud environments. Derivative maps such as dense point clouds, mesh maps, Octomaps, etc. are frequently used for path planning purposes. A bi-directional target-oriented point planning algorithm, directly using point clouds to compute the optimized and dynamically feasible trajectories, is presented in this paper. This approach searches for obstacle-free, low computational cost, smooth, and dynamically feasible paths by analyzing a point cloud of the target environment, using a modified bi-directional and RRT-connect-based path planning algorithm, with a k-d tree-based obstacle avoidance strategy and three-step optimization. This presented approach bypasses the common 3D map discretization, directly leveraging point cloud data and it can be separated into two parts: modified RRT-based algorithm core and the three-step optimization. Simulations on 8 2D maps with different configurations and characteristics are presented to show the efficiency and 2D performance of the proposed algorithm. Benchmark comparison and evaluation with other RRT-based algorithms like RRT, B-RRT, and RRT star are also shown in the paper. Finally, the proposed algorithm successfully achieved different levels of mission goals on three 3D point cloud maps with different densities. The whole simulation proves that not only can our algorithm achieves a better performance on 2D maps compared with other algorithms, but also it can handle different tasks(ground vehicles and UAV applications) on different 3D point cloud maps, which shows the high performance and robustness of the proposed algorithm. The algorithm is open-sourced at \url{https://github.com/zhz03/BTO-RRT}

Automated vehicles require the ability to cooperate with humans for smooth integration into today's traffic. While the concept of cooperation is well known, developing a robust and efficient cooperative trajectory planning method is still a challenge. One aspect of this challenge is the uncertainty surrounding the state of the environment due to limited sensor accuracy. This uncertainty can be represented by a Partially Observable Markov Decision Process. Our work addresses this problem by extending an existing cooperative trajectory planning approach based on Monte Carlo Tree Search for continuous action spaces. It does so by explicitly modeling uncertainties in the form of a root belief state, from which start states for trees are sampled. After the trees have been constructed with Monte Carlo Tree Search, their results are aggregated into return distributions using kernel regression. We apply two risk metrics for the final selection, namely a Lower Confidence Bound and a Conditional Value at Risk. It can be demonstrated that the integration of risk metrics in the final selection policy consistently outperforms a baseline in uncertain environments, generating considerably safer trajectories.

As robotic systems continue to address emerging issues in areas such as logistics, mobility, manufacturing, and disaster response, it is increasingly important to rapidly generate safe and energy-efficient trajectories. In this article, we present a new approach to plan energy-optimal trajectories through cluttered environments containing polygonal obstacles. In particular, we develop a method to quickly generate optimal trajectories for a double-integrator system, and we show that optimal path planning reduces to an integer program. To find an efficient solution, we present a distance-informed prefix search to efficiently generate optimal trajectories for a large class of environments. We demonstrate that our approach, while matching the performance of RRT* and Probabilistic Road Maps in terms of path length, outperforms both in terms of energy cost and computational time by up to an order of magnitude. We also demonstrate that our approach yields implementable trajectories in an experiment with a Crazyflie quadrotor.

The exponential growth in demand for digital services drives massive datacenter energy consumption and negative environmental impacts. Promoting sustainable solutions to pressing energy and digital infrastructure challenges is crucial. Several hyperscale cloud providers have announced plans to power their datacenters using renewable energy. However, integrating renewables to power the datacenters is challenging because the power generation is intermittent, necessitating approaches to tackle power supply variability. Hand engineering domain-specific heuristics-based schedulers to meet specific objective functions in such complex dynamic green datacenter environments is time-consuming, expensive, and requires extensive tuning by domain experts. The green datacenters need smart systems and system software to employ multiple renewable energy sources (wind and solar) by intelligently adapting computing to renewable energy generation. We present RARE (Renewable energy Aware REsource management), a Deep Reinforcement Learning (DRL) job scheduler that automatically learns effective job scheduling policies while continually adapting to datacenters' complex dynamic environment. The resulting DRL scheduler performs better than heuristic scheduling policies with different workloads and adapts to the intermittent power supply from renewables. We demonstrate DRL scheduler system design parameters that, when tuned correctly, produce better performance. Finally, we demonstrate that the DRL scheduler can learn from and improve upon existing heuristic policies using Offline Learning.

Planning is an extraordinary ability in which the brain imagines and then enacts evaluated possible futures. Using traditional planning models, computer scientists have attempted to replicate this capacity with some level of success but ultimately face a reoccurring limitation: as the plan grows in steps, the number of different possible futures makes it intractable to determine the right sequence of actions to reach a goal state. Based on prior theoretical work on how the ecology of an animal governs the value of spatial planning, we developed a more efficient biologically-inspired planning algorithm, TLPPO. This algorithm allows us to achieve mouselevel predator evasion performance with orders of magnitude less computation than a widespread algorithm for planning in the situations of partial observability that typify predator-prey interactions. We compared the performance of a real-time agent using TLPPO against the performance of live mice, all tasked with evading a robot predator. We anticipate these results will be helpful to planning algorithm users and developers, as well as to areas of neuroscience where robot-animal interaction can provide a useful approach to studying the basis of complex behaviors.

Exploration and mapping of unknown environments is a fundamental task in applications for autonomous robots. In this article, we present a complete framework for deploying MAVs in autonomous exploration missions in unknown subterranean areas. The main motive of exploration algorithms is to depict the next best frontier for the robot such that new ground can be covered in a fast, safe yet efficient manner. The proposed framework uses a novel frontier selection method that also contributes to the safe navigation of autonomous robots in obstructed areas such as subterranean caves, mines, and urban areas. The framework presented in this work bifurcates the exploration problem in local and global exploration. The proposed exploration framework is also adaptable according to computational resources available onboard the robot which means the trade-off between the speed of exploration and the quality of the map can be made. Such capability allows the proposed framework to be deployed in a subterranean exploration, mapping as well as in fast search and rescue scenarios. The overall system is considered a low-complexity and baseline solution for navigation and object localization in tunnel-like environments. The performance of the proposed framework is evaluated in detailed simulation studies with comparisons made against a high-level exploration-planning framework developed for the DARPA Sub-T challenge as it will be presented in this article.