Mobile manipulation problems involving many objects are challenging to solve due to the high dimensionality and multi-modality of their hybrid configuration spaces. Planners that perform a purely geometric search are prohibitively slow for solving these problems because they are unable to factor the configuration space. Symbolic task planners can efficiently construct plans involving many variables but cannot represent the geometric and kinematic constraints required in manipulation. We present the FFROB algorithm for solving task and motion planning problems. First, we introduce Extended Action Specification (EAS) as a general purpose planning representation that supports arbitrary predicates as conditions. We adapt existing heuristic search ideas for solving STRIPS planning problems, particularly delete-relaxations, to solve EAS problem instances. We then apply the EAS representation and planners to manipulation problems resulting in FFROB. FFROB iteratively discretizes task and motion planning problems using batch sampling of manipulation primitives and a multi-query roadmap structure that can be conditionalized to evaluate reachability under different placements of movable objects. This structure enables the EAS planner to efficiently compute heuristics that incorporate geometric and kinematic planning constraints to give a tight estimate of the distance to the goal. Additionally, we show FFROB is probabilistically complete and has finite expected runtime. Finally, we empirically demonstrate FFROB's effectiveness on complex and diverse task and motion planning tasks including rearrangement planning and navigation among movable objects.
translated by 谷歌翻译
We describe an integrated strategy for planning, perception, state-estimation and action in complex mobile manipulation domains based on planning in the belief space of probability distributions over states using hierarchical goal regression (pre-image back-chaining). We develop a vocabulary of logical expressions that describe sets of belief states, which are goals and subgoals in the planning process. We show that a relatively small set of symbolic operators can give rise to task-oriented perception in support of the manipulation goals. An implementation of this method is demonstrated in simulation and on a real PR2 robot, showing robust, flexible solution of mobile manipulation problems with multiple objects and substantial uncertainty.
translated by 谷歌翻译
贝叶斯优化通常假设给出贝叶斯先验。然而,贝叶斯优化中强有力的理论保证在实践中经常因为先验中的未知参数而受到损害。在本文中,我们采用经验贝叶斯的变量并表明,通过估计从同一个先前采样的离线数据之前的高斯过程和构建后验的无偏估计,GP-UCB的变体和改进概率实现近乎零的后悔界限,其随着离线数据和离线数据的数量减少到与观测噪声成比例的常数。在线评估的数量增加。根据经验,我们已经验证了我们的方法,以挑战模拟机器人问题为特色的任务和运动规划。
translated by 谷歌翻译
这项工作的目的是通过学习来增加机器人的基本能力,以使用新的感觉运动原语来解决复杂的长期问题。解决复杂领域中的长期问题需要灵活的生成规划,这种规划可以结合原始能力的新组合来解决世界上出现的问题。为了将原始行为结合起来,我们必须有先决条件的模型来影响这些行为:在什么情况下执行这个原则会在世界上产生某种特殊的影响?我们使用并开发了最先进的方法进行有趣学习和采样的新颖改进。我们使用高斯过程方法从机器人实验中收集的少量昂贵的训练样本中学习操作员有效性的条件。我们开发了自适应采样方法,用于在规划解决新任务期间生成连续集的各种元素(例如机器人配置和对象姿势),从而使计划尽可能高效。我们在一个集成系统中演示了这些方法,将新学习的模型与高效的连续空间机器人任务和运动规划器相结合,学习如何比以前更有效地解决长期问题。
translated by 谷歌翻译
本文讨论了非自适应重排计划问题,其中机器人的任务是重新排列平面上障碍物之间的物体。我们提出了一种有效的规划算法,该算法旨在对机器人的非适应性操纵能力施加一些假设,并且易于适应不同的机器人实施例。为此,我们将基于抽样的运动计划与强化学习和生成建模相结合。 Ouralgorithm将物体和机器人的复合配置空间作为机器人动作的研究,在物理模型中进行模拟。该搜索由生成模型引导,该生成模型提供可以将对象从其运送到期望状态的机器人状态,以及提供相应的机器人动作的学习策略。作为一种有效的生成模型,我们应用了生成对抗网络。我们实施并评估了我们在SE(2)中赋予配置空间的机器人方法。我们通过经验证明了我们的算法设计选择的有效性,并且与最先进的方法相比,在各种测试场景中观察了超过2倍的加速计划时间。
translated by 谷歌翻译
In this paper, we present CHOMP (Covariant Hamiltonian Optimization for Motion Planning), a method for trajectory optimization invariant to reparametrization. CHOMP uses functional gradient techniques to iteratively improve the quality of an initial trajectory, optimizing a functional that trades off between a smoothness and an obstacle avoidance component. CHOMP can be used to locally optimize feasible trajectories, as well as to solve motion planning queries, converging to low-cost trajectories even when initialized with infeasible ones. It uses Hamiltonian Monte Carlo to alleviate the problem of convergence to high-cost local minima (and for probabilistic completeness), and is capable of respecting hard constraints along the trajectory. We present extensive experiments with CHOMP on manipulation and locomotion tasks, using 7-DOF manipulators and a rough-terrain quadruped robot.
translated by 谷歌翻译
The Partially Observable Markov Decision Process has long been recognized as a rich framework for real-world planning and control problems, especially in robotics. However exact solutions in this framework are typically computationally intractable for all but the smallest problems. A well-known technique for speeding up POMDP solving involves performing value backups at specific belief points, rather than over the entire belief simplex. The efficiency of this approach, however, depends greatly on the selection of points. This paper presents a set of novel techniques for selecting informative belief points which work well in practice. The point selection procedure is combined with point-based value backups to form an effective anytime POMDP algorithm called Point-Based Value Iteration (PBVI). The first aim of this paper is to introduce this algorithm and present a theoretical analysis justifying the choice of belief selection technique. The second aim of this paper is to provide a thorough empirical comparison between PBVI and other state-of-the-art POMDP methods, in particular the Perseus algorithm, in an effort to highlight their similarities and differences. Evaluation is performed using both standard POMDP domains and realistic robotic tasks.
translated by 谷歌翻译
This paper presents a novel randomized motion planner for robots that must achieve a specified goal under kinematic and/or dynamic motion constraints while avoiding collision with moving obstacles with known trajectories. The planner encodes the motion constraints on the robot with a control system and samples the robot's state×time space by picking control inputs at random and integrating its equations of motion. The result is a probabilistic roadmap of sampled state×time points, called milestones, connected by short admissible trajectories. The planner does not precompute the roadmap; instead, for each planning query, it generates a new roadmap to connect an initial and a goal state×time point. The paper presents a detailed analysis of the planner's convergence rate. It shows that, if the state×time space satisfies a geometric property called expansive-ness, then a slightly idealized version of our implemented planner is guaranteed to find a trajectory when one exists, with probability quickly converging to 1, as the number of milestones increases. Our planner was tested extensively not only in simulated environments, but also on a real robot. In the latter case, a vision module estimates obstacle motions just before planning starts. The planner is then allocated a small, fixed amount of time to compute a trajectory. If a change in the expected motion of the obstacles is detected while the robot executes the planned trajectory, the planner recomputes a trajectory on the fly. Experiments on the real robot led to several extensions of the planner in order to deal with time delays and uncertainties that are inherent to an integrated robotic system interacting with the physical world.
translated by 谷歌翻译
We present a new optimization-based approach for robotic motion planning among obstacles. Like CHOMP (Covariant Hamiltonian Optimization for Motion Planning), our algorithm can be used to find collision-free trajectories from naïve, straight-line initializations that might be in collision. At the core of our approach are (a) a sequential convex optimization procedure, which penalizes collisions with a hinge loss and increases the penalty coefficients in an outer loop as necessary, and (b) an efficient formulation of the no-collisions constraint that directly considers continuous-time safety Our algorithm is implemented in a software package called TrajOpt. We report results from a series of experiments comparing TrajOpt with CHOMP and randomized planners from OMPL, with regard to planning time and path quality. We consider motion planning for 7 DOF robot arms, 18 DOF full-body robots, statically stable walking motion for the 34 DOF Atlas humanoid robot, and physical experiments with the 18 DOF PR2. We also apply TrajOpt to plan curvature-constrained steerable needle trajectories in the SE(3) configuration space and multiple non-intersecting curved channels within 3D-printed implants for intracavitary brachytherapy. Details, videos, and source code are freely available at: http://rll.berkeley.edu/trajopt/ijrr.
translated by 谷歌翻译
In a realistic mobile push-manipulation scenario , it becomes non-trivial and infeasible to build analytical models that will capture the complexity of the interactions between the environment, each of the objects, and the robot as the variety of objects to be manipulated increases. We present an experience-based push-manipulation approach that enables the robot to acquire experimental models regarding how pushable real world objects with complex 3D structures move in response to various pushing actions. These experimentally acquired models can then be used either (1) for trying to track a collision-free guideline path generated for the object by reiterating pushing actions that result in the best locally-matching object trajectories until the goal is reached, or (2) as building blocks for constructing achievable push plans via a Rapidly-exploring Random Trees variant planning algorithm we contribute and executing them by reiterating the corresponding trajectories. We extensively experiment with these two methods in a 3D simulation environment and demonstrate the superiority of the achievable planning and execution concept through safe and successful push-manipulation of a variety of passively mobile pushable objects. Additionally, our preliminary tests in a real world scenario, where the robot is asked to arrange a set of chairs around a table through achievable push-manipulation, also show promising results despite the increased perception and action uncertainty, and verify the validity of our contributed method.
translated by 谷歌翻译
本文讨论了规划具有定向障碍物检测传感器的机器人的问题,该传感器必须在杂乱的环境中移动。规划目标是通过找到包括传感器在内的完整机器人的路径来保持安全,该路径确保机器人在传感器看到之前不会移动到工作空间的任何部分。虽然大量的工作已经解决了这个问题的一个版本,其中传感器的“视野”是机器人周围的球体,但是很少有工作机器人具有狭窄或遮挡的视野。我们给出了问题的正式定义,几种具有不同计算权衡的解决方法,以及示例域中的实验结果。
translated by 谷歌翻译
We present a tutorial on Bayesian optimization, a method of finding the maximum of expensive cost functions. Bayesian optimization employs the Bayesian technique of setting a prior over the objective function and combining it with evidence to get a posterior function. This permits a utility-based selection of the next observation to make on the objective function, which must take into account both exploration (sampling from areas of high uncertainty) and exploitation (sampling areas likely to offer improvement over the current best observation). We also present two detailed extensions of Bayesian optimization, with experiments-active user modelling with preferences, and hierarchical reinforcement learning-and a discussion of the pros and cons of Bayesian optimization based on our experiences.
translated by 谷歌翻译
We propose a learning-from-demonstration approach for grounding actions from expert data and an algorithm for using these actions to perform a task in new environments. Our approach is based on an application of sampling-based motion planning to search through the tree of discrete, high-level actions constructed from a symbolic representation of a task. Recursive sampling-based planning is used to explore the space of possible continuous-space instantiations of these actions. We demonstrate the utility of our approach with a magnetic structure assembly task, showing that the robot can intelligently select a sequence of actions in different parts of the workspace and in the presence of obstacles. This approach can better adapt to new environments by selecting the correct high-level actions for the particular environment while taking human preferences into account.
translated by 谷歌翻译
Intelligent mobile robots have recently become able to operate autonomously in large-scale indoor environments for extended periods of time. Task planning in such environments involves sequencing the robot's high-level goals and subgoals, and typically requires reasoning about the locations of people, rooms, and objects in the environment, and their interactions to achieve a goal. One of the prerequisites for optimal task planning that is often overlooked is having an accurate estimate of the actual distance (or time) a robot needs to navigate from one location to another. State-of-the-art motion planners, though often computationally complex, are designed exactly for this purpose of finding routes through constrained spaces. In this work, we focus on integrating task and motion planning (TMP) to achieve task-level optimal planning for robot navigation while maintaining manageable computational efficiency. To this end, we introduce TMP algorithm PETLON (Planning Efficiently for Task-Level-Optimal Navigation) for everyday service tasks using a mobile robot. PETLON is more efficient than planning approaches that pre-compute motion costs of all possible navigation actions, while still producing plans that are optimal at the task level.
translated by 谷歌翻译
When monitoring spatial phenomena, which can often be modeled as Gaussian processes (GPs), choosing sensor locations is a fundamental task. There are several common strategies to address this task, for example, geometry or disk models, placing sensors at the points of highest entropy (vari-ance) in the GP model, and AD D-, or E-optimal design. In this paper, we tackle the combinatorial optimization problem of maximizing the mutual information between the chosen locations and the locations which are not selected. We prove that the problem of finding the configuration that maximizes mutual information is NP-complete. To address this issue, we describe a polynomial-time approximation that is within (1 − 1/e) of the optimum by exploiting the submodularity of mutual information. We also show how submodularity can be used to obtain online bounds, and design branch and bound search procedures. We then extend our algorithm to exploit lazy evaluations and local structure in the GP, yielding significant speedups. We also extend our approach to find placements which are robust against node failures and uncertainties in the model. These extensions are again associated with rigorous theoretical approximation guarantees, exploiting the submodu-larity of the objective function. We demonstrate the advantages of our approach towards optimizing mutual information in a very extensive empirical study on two real-world data sets.
translated by 谷歌翻译
Noisy probabilistic relational rules are a promising world model representation for several reasons. They are compact and generalize over world instantiations. They are usually interpretable and they can be learned effectively from the action experiences in complex worlds. We investigate reasoning with such rules in grounded relational domains. Our algorithms exploit the compactness of rules for efficient and flexible decision-theoretic planning. As a first approach, we combine these rules with the Upper Confidence Bounds applied to Trees (UCT) algorithm based on look-ahead trees. Our second approach converts these rules into a structured dynamic Bayesian network representation and predicts the effects of action sequences using approximate inference and beliefs over world states. We evaluate the effectiveness of our approaches for planning in a simulated complex 3D robot manipulation scenario with an articulated manipulator and realistic physics and in domains of the probabilistic planning competition. Empirical results show that our methods can solve problems where existing methods fail.
translated by 谷歌翻译
连续状态和动作空间中的多对象操纵问题可以通过搜索操作符的连续参数的采样值的规划器来解决。这些规划者的效率关键取决于所用采样器的有效性,但有效采样又取决于机器人,环境和任务的细节。我们的策略是学习称为特殊化器的函数,它们为连续的操作符参数生成值,给定离散参数的状态描述和值。除了尝试从单个任务的大量数据中为每个操作符学习单个特化器之外,我们采用模块化元学习方法。对多个任务进行编辑,并学习各种专业人员,这些专业人员可以使用相对较少的数据快速调整新任务 - 因此,我们的系统可以“快速学习使用这些专业人员快速计划”。我们通过连续状态和动作空间在模拟3D拾取和放置任务中实验性地验证了ourapproach。
translated by 谷歌翻译
We consider the problem of generating motion plans for a robot that are guaranteed to succeed despite uncertainty in the environment, parametric model uncertainty, and disturbances. Furthermore, we consider scenarios where these plans must be generated in real-time, because constraints such as obstacles in the environment may not be known until they are perceived (with a noisy sensor) at runtime. Our approach is to pre-compute a library of "funnels" along different maneuvers of the system that the state is guaranteed to remain within (despite bounded disturbances) when the feedback controller corresponding to the maneuver is executed. We leverage powerful computational machinery from convex optimization (sums-of-squares programming in particular) to compute these funnels. The resulting funnel library is then used to sequentially compose motion plans at runtime while ensuring the safety of the robot. A major advantage of the work presented here is that by explicitly taking into account the effect of uncertainty, the robot can evaluate motion plans based on how vulnerable they are to disturbances. We demonstrate and validate our method using extensive hardware experiments on a small fixed-wing airplane avoiding obstacles at high speed (∼12 mph), along with thorough simulation experiments of ground vehicle and quadrotor models navigating through cluttered environments. To our knowledge, these demonstrations constitute one of the first examples of provably safe and robust control for robotic systems with complex nonlinear dynamics that need to plan in real-time in environments with complex geometric constraints.
translated by 谷歌翻译