Single and Multiple View Support Order Prediction in Clutter for Manipulation

Robotic manipulation of objects in clutter remains a challenging problem to date. The challenge is posed by various levels of complexity involved in interaction among objects. Understanding these semantic interactions among different objects is important to manipulate in complex settings. It can play a significant role in extending the scope of manipulation to cluttered environment involving generic objects, and both direct and indirect physical contact. In our work, we aim at learning semantic interaction among objects of generic shapes and sizes lying in clutter involving physical contact. We infer three types of support relationships: “support from below”, “support from side”, and “containment”. Subsequently, the learned semantic interaction or support relationship is used to derive a sequence or order in which the objects surrounding the object of interest should be removed without causing damage to the environment. The generated sequence is called support order. We also extend understanding of semantic interaction from single view to multiple views and predict support order in multiple views. Using multiple views addresses those cases that are not handled when using single view such as scenarios of occlusion or missing support relationships. We have created two RGBD datasets for our experiments on support order prediction in single view and multiple views respectively. The datasets contains RGB images, point clouds and depth maps of various objects used in day-to-day life present in clutter with physical contact and overlap. We captured many different cluttered settings involving different kinds of object-object interaction and successfully learned support relationship and performed Support Order Prediction in these settings.     

面向不确定性环境的自动驾驶运动规划：机遇与挑战

运动规划算法作为自动驾驶系统中的重要研究内容,愈发受到研究者们关注.然而目前多数算法仅考虑在确定性结构化环境中的应用,忽视动态交通环境中潜在的不确定性因素.文中面向不确定性环境,将运动规划算法总结为两类：部分可观测马尔可夫决策过程(POMDP)和概率占用栅格图(POGM),从理论基础、求解算法、实际应用三方面进行介绍.基于当前置信状态,POMDP计算使未来折扣奖励最大的策略.POGM使用概率表征对应栅格上的占用情况,衡量车流动态变化的可能性,良好表征不确定性情况.最后,总结不确定性环境中当前运动规划问题面临的主要挑战和未来可能的研究方向.     

A Planning Framework for Systems Development Projects

As part of any systems development project, JS professionals must perform a variety of tasks. One such important task, estimating, is often done incorrectly and, as a result, can lead to errors in the final system or a cancellation of the project. To estimate projects accurately, data on previous project estimates should be readily available. This article describes a systems development planning framework that essentially creates an estimating knowledge data base, which enables IS managers to prepare better project plans.     

Manipulation Planning for Deformable Linear Objects

Research on robotic manipulation has mainly focused on manipulating rigid objects so far. However, many important application domains require manipulating deformable objects, especially deformable linear objects (DLOs), such as ropes, cables, and sutures. Such objects are far more challenging to handle, as they can exhibit a much greater diversity of behaviors, and their manipulation almost inevitably requires two robotic arms, or more, performing well-coordinated motions. This paper describes a new motion planner for manipulating DLOs and tying knots (both self-knots and knots around simple static objects) using two cooperating robotic arms. This planner blends new ideas with preexisting concepts and techniques from knot theory, robot motion planning, and computational modeling. Unlike in traditional motion planning problems, the goal to be achieved by the planner is a topological state of the world, rather than a geometric one. To search for a manipulation path, the planner constructs a topologically biased probabilistic roadmap in the configuration space of the DLO. During roadmap construction, it uses inverse kinematics to determine the successive robot configurations implied by the DLO configurations and tests their feasibility. Also, inspired by the real life, the planner uses static ldquoneedlesrdquo (by analogy to the needles used in knitting) for maintaining the stability of the DLO during manipulation and to make the resulting manipulation plan robust to imperfections in the physical model of the DLO. The implemented planner has been tested both in graphic simulation and on a dual-PUMA-560 hardware platform to achieve various knots, like bowline, neck-tie, bow (shoe-lace), and stun-sail.     

不确定层次任务网络规划研究综述

层次任务网络(Hierarchical task network, HTN)规划作为一项重要的智能规划技术被广泛应用于实际规划问题中, 传统的HTN规划无法处理不确定规划问题.然而, 现实世界不可避免地存在无法确定或无法预测的信息, 这使许多学者开始关注不确定规划问题, 不确定HTN规划研究也成为HTN规划研究的前沿.本文从HTN规划过程出发分析了不确定HTN规划问题中涉及的三类不确定, 即状态不确定、动作效果不确定和任务分解不确定; 总结了系统状态、动作效果和任务分解等不确定需要扩展确定性HTN规划模型的工作, 以此对现有不确定HTN规划的研究工作加以梳理和归类; 最后,对不确定HTN规划研究中仍需要解决的问题和未来的研究方向作了进一步展望.     

A Stochastic Multi-agent Optimization Model for Energy Infrastructure Planning under Uncertainty in An Oligopolistic Market

This paper presents a mathematical model for analyzing long-term infrastructure investment decisions in a deregulated electricity market, such as the case in the United States. The interdependence between different decision entities in the system is captured in a network-based stochastic multi-agent optimization model, where new entrants of investors compete among themselves and with existing generators for natural resources, transmission capacities, and demand markets. To overcome computational challenges involved in stochastic multi-agent optimization problems, we have developed a solution method by combining stochastic decomposition and variational inequalities, which converts the original problem to many smaller problems that can be solved more easily.     

Parallel Planning: A New Motion Planning Framework for Autonomous Driving

Motion planning is one of the most significant technologies for autonomous driving. To make motion planning models able to learn from the environment and to deal with emergency situations, a new motion planning framework called as "parallel planning" is proposed in this paper. In order to generate sufficient and various training samples, artificial traffic scenes are firstly constructed based on the knowledge from the reality. A deep planning model which combines a convolutional neural network (CNN) with the Long Short-Term Memory module (LSTM) is developed to make planning decisions in an end-toend mode. This model can learn from both real and artificial traffic scenes and imitate the driving style of human drivers. Moreover, a parallel deep reinforcement learning approach is also presented to improve the robustness of planning model and reduce the error rate. To handle emergency situations, a hybrid generative model including a variational auto-encoder (VAE) and a generative adversarial network (GAN) is utilized to learn from virtual emergencies generated in artificial traffic scenes. While an autonomous vehicle is moving, the hybrid generative model generates multiple video clips in parallel, which correspond to different potential emergency scenarios. Simultaneously, the deep planning model makes planning decisions for both virtual and current real scenes. The final planning decision is determined by analysis of real observations. Leveraging the parallel planning approach, the planner is able to make rational decisions without heavy calculation burden when an emergency occurs.      

Non-decoupled Locomotion and Manipulation Planning for Low-Dimensional Systems

We demonstrate the possibility of solving planning problems by interleaving locomotion and manipulation in a non-decoupled way. We choose three low-dimensional minimalistic robotic systems and use them to illustrate our paradigm: a basic one-legged locomotor, a two-link manipulator with a manipulated object, and a simultaneous locomotion-and-manipulation system. Using existing motion planning and control methods initially designed for either locomotion or manipulation tasks, we see how they apply to both our locomotion-only and manipulation-only systems through parallel derivations, and extend them to the simultaneous locomotion-and-manipulation system. Motion planning is solved for these three systems using two different methods: (i) a geometric path-planning-based one, and (ii) a kinematic control-theoretic-based one. Motion control is then derived by dynamically realizing the geometric paths or kinematic trajectories under the Couloumb friction model using torques as control inputs. All three methods apply successfully to all three systems, showing that the non-decoupled planning is possible.     

基于动态约束满足框架的强表达时态规划算法

智能规划已经成为人工智能领域最热门的研究主题之一。近年来,智能规划在现实领域的应用越来越广泛,这对规划器的处理能力和效率提出了很大的挑战。以一类强表达时态规划——基于约束区间规划为研究对象,基于动态约束满足框架设计和实现了一个基于约束区间的规划算法LP-TPOP;对算法的可靠性和完备性进行了证明;最后以一个规划实例演示了算法的运行过程。     

A General Framework for Planning Landmark-Based Motions for Mobile Robots

In this paper, we present a novel generic approach for planning landmark-based motion. The method consists in selecting automatically the most relevant landmarks along a preplanned geometric path. It proposes a strategy to correct the trajectory and to smoothly switch among the landmarks of the environment. Experimental results highlight the relevance of the proposed formalism.     

A General Approach for Dexterous Manipulation Planning Based on Configuration Space Structuring

In this paper, we propose a new method for the dexterous manipulation planning problem, under quasi-static movement assumption. This method computes both object and finger trajectories as well as finger relocation sequence and applies to every object shape and hand geometry. It relies on the exploration of the particular subspaces GS k that are the subspaces of all the grasps that can be achieved for a given set of k grasping fingers. The originality is to use continuous paths in these subspaces to directly link two configurations. The proposed approach captures the GS k connectivity in a graph structure. The answer of the manipulation planning query is then given by searching a path in the computed graph. Another specificity of our technique is that it considers manipulated object and hand as an only system, unlike most existing methods that first compute object trajectory then fingers trajectories and thus can not find a solution in all situations. Simulation experiments were conducted for different dexterous manipulation task examples to validate the proposed method.     

灵巧手操作的重构规划

根据人类利用滑动或滚动方式进行灵巧操作的几种模式 ,在多指灵巧手—物体的抓持系统中 ,把实现这些灵巧操作模式的规划问题作为一个位形空间的重构问题进行全局和局部级的操作规划 ,开发了实现灵巧操作的规划算法 .仿真结果表明了方法的有效性和正确性 .     

A General Framework for Expressing Preferences in Causal Reasoning and Planning

We consider the problem of representing arbitrary preferencesin causal reasoning and planning systems. In planning, a preferencemay be seen as a goal or constraint that is desirable, but notnecessary, to satisfy. To begin, we define a very general querylanguage for histories, or interleaved sequences of world statesand actions. Based on this, we specify a second language inwhich preferences are defined. A single preference defines abinary relation on histories, indicating that one history ispreferred to the other. From this, one can define global preferenceorderings on the set of histories, the maximal elements of whichare the preferred histories. The approach is very general andflexible; thus it constitutes a ‘base’ languagein terms of which higher-level preferences may be defined. Tothis end, we investigate two fundamental types of preferencesthat we call choice and temporal preferences. We consider concretestrategies for these types of preferences and encode them interms of our framework. We suggest how to express aggregatesin the approach, allowing, e.g. the expression of a preferencefor histories with lowest total action costs. Last, our approachcan be used to express other approaches and so serves as a commonframework in which such approaches can be expressed and compared.We illustrate this by indicating how an approach due to Sonand Pontelli can be encoded in our approach, as well as thelanguage PDDL3.     

顺序价值迭代算法求解不确定规划

基于Markov决策过程(MDP)的规划方法可以处理多种不确定规划问题,价值迭代算法(VI)是求解MDP的经典算法,但VI需要计算更新每个状态的值,求解过程相当缓慢。在分析了MDP状态图本身的因果依赖关系的基础上,提出一种改进的价值迭代算法,称为顺序价值迭代算法(SVI)。它先将一个MDP分解成多个拓扑有序的强连通分量,然后应用价值迭代算法顺序求解各个分量,这样处理可以避免对大量无用状态的计算并使得可用状态排成拓扑序列。对比实验结果证明了该算法的有效性及优异性能。     

非确定规划及带有时间和资源的规划的研究

智能规划是人工智能近年来的研究热点, 早期的工作主要是围绕着具有较强约束的经典规划展开, 最近的工作放宽了这些假设, 使智能规划逐渐走向应用。在分析经典规划特点的基础上, 介绍了非确定规划的研究进展和带有时间和资源的规划的研究, 并对智能规划的进一步工作和存在的问题提出了一些看法。     

A Motion Planning Framework with Connectivity Management for Multiple Cooperative Robots

This paper proposes a decentralized motion planning algorithm for multiple cooperative robots subject to constraints imposed by sensors and the communication network. It consists of decentralized receding horizon planners that reside on each vehicle to navigate to individual target positions. A routing algorithm which modify the network topology based on the position of the robots and the limited range of transmitters and receivers, enables to reduce the communication link failures. A comparative study between the proposed algorithm and other existing algorithms is provided in order to show the advantages especially in terms of traveling time and communication link failure.     

A General Framework for Assembly Planning: The Motion Space Approach

Assembly planning is the problem of finding a sequence of motions to assemble a product from its parts. We present a general framework for finding assembly motions based on the concept of motion space . Assembly motions are parameterized such that each point in motion space represents a mating motion that is independent of the moving part set. For each motion we derive blocking relations that explicitly state which parts collide with other parts; each subassembly (rigid subset of parts) that does not collide with the rest of the assembly can easily be derived from the blocking relations. Motion space is partitioned into an arrangement of cells such that the blocking relations are fixed within each cell. We apply the approach to assembly motions of several useful types, including one-step translations, multistep translations, and infinitesimal rigid motions. Several efficiency improvements are described, as well as methods to include additional assembly constraints into the framework. The resulting algorithms have been implemented and tested extensively on complex assemblies. We conclude by describing some remaining open problems. Received November 15, 1996; revised January 15, 1998.     

SCM环境中生产计划信息不确定性问题研究

本文探讨在供应链环境下ERP生产计划信息不确定性产生的原因及对生产计划的影响;从数据模式转化机制与数据库集成机制两方面探讨了供应链环境下生产计划信息集成机制问题,并建立了基于数据传递、数据共享、应用集成三个层次的信息共享机制。