ICRA 2015 Papers

This paper proposes a method for path planning of high-degree of freedom (DOF) articulated robots with adaptive dimensionality. For an efficient path planning in high-dimensional C-space (configuration space), first we describe an adaptive body selection that selects the robot bodies and joints depending on the complexity of path planning. It means that the robot may use necessary DOF to achieve a path planning task. The adaptive body selection method builds the C-space with adaptive dimensionality for a sampling-based path planner. Next, by using the adaptive body selection, the adaptive Rapidly-Exploring Random Tree (RRT) algorithm is introduced, which incrementally grows RRTs in the adaptive dimensional C-space. And we show through several simulation results that the proposed method is more efficient than the original RRT-based path planner, which requires full-dimensional planning.

We propose a robust and efficient method to estimate the pose of a camera with respect to complex 3D textured models of the environment that can potentially contain more than 100; 000 points. To tackle this problem we follow a top down approach where we combine high-level deep network classifiers with low level geometric approaches to come up with a solution that is fast, robust and accurate. Given an input image, we initially use a pre-trained deep network to compute a rough estimation of the camera pose. This initial estimate constrains the number of 3D model points that can be seen from the camera viewpoint. We then establish 3D-to-2D correspondences between these potentially visible points of the model and the 2D detected image features. Accurate pose estimation is finally obtained from the 2D-to-3D correspondences using a novel PnP algorithm that rejects outliers without the need to use a RANSAC strategy, and which is between 10 and 100 times faster than other methods that use it. Two real experiments dealing with very large and complex 3D models demonstrate the effectiveness of the approach.

We present a new approach to the design of smooth trajectories for quadrotor unmanned aerial vehicles (UAVs), which are free of collisions with obstacles along their entire length. To avoid the non-convex constraints normally required for obstacle-avoidance, we perform a mixed-integer optimization in which polynomial trajectories are assigned to convex regions which are known to be obstacle-free. Prior approaches have used the faces of the obstacles themselves to define these convex regions. We instead use IRIS, a recently developed technique for greedy convex segmentation [1], to pre-compute convex regions of safe space. This results in a substantially reduced number of integer variables, which improves the speed with which the optimization can be solved to its global optimum, even for tens or hundreds of obstacle faces. In addition, prior approaches have typically enforced obstacle avoidance at a finite set of sample or knot points. We introduce a technique based on sums-of-squares (SOS) programming that allows us to ensure that the entire piecewise polynomial trajectory is free of collisions using convex constraints. We demonstrate this technique in 2D and in 3D using a dynamical model in the Drake toolbox for Matlab [2].

The problem of gas detection is relevant to many real-world applications, such as leak detection in industrial settings and surveillance. In this paper we address the problem of gas detection in large areas with a mobile robotic platform equipped with a remote gas sensor. We propose a novel method based on convex relaxation for quickly finding an exploration plan that guarantees a complete coverage of the environment. Our method proves to be highly efficient in terms of computational requirements and to provide nearly-optimal solutions. We validate our approach both in simulation and in real environments, thus demonstrating its applicability to real-world problems.

The paper considers the efficient selection of leader agents in a swarm running a distributed bearing-compass formation controller. The leaders apply external control which induces translation and scaling of the formation, providing manipulation methods useful to a human operator. The selection algorithm for maximizing translation and scale draws from modularity and submodularity theory. Consequently, the algorithms exhibit guaranteed optimal and suboptimal performance, respectively. For more restricted human-swarm interaction requiring pure translation and scale, a relaxed integer programming algorithm is described to reduce the combinatorial optimization problem to a computationally tractable semidefinite program. The leader selection strategies are supported through demonstration on a swarm testbed.

Sampling-based motion-planning algorithms typically rely on nearest-neighbor (NN) queries when constructing a roadmap. Recent results suggest that in various settings NN queries may be the computational bottleneck of such algorithms. Moreover, in several asymptotically-optimal algorithms these NN queries are of a specific form: Given a set of points and a radius r report all pairs of points whose distance is at most r. This calls for an application-specific NN data structure tailored to efficiently answering this type of queries. Randomly transformed grids (RTG) were recently proposed by Aiger et al. [1] as a tool to answer such queries in Euclidean spaces and have been shown to outperform common implementations of NN data structures for this type of queries. In this work we employ RTG for sampling-based motion-planning algorithms and describe an efficient implementation of the approach. We show that for motion planning, RTG allow for faster convergence to high-quality solutions when compared to existing NN data structures. Additionally, RTG enable significantly shorter construction times for batched-PRM variants; specifically, we demonstrate a speedup by a factor of two to three for some scenarios.

The main reported solutions for the problem of multi-robot relative localization require synchronous communication between robots, where the network should communicate each time a relative measurement is logged in the team. This paper proposes a localization method, which can accommodate communication at a low predefined rate rather than forcing communication each time a measurement is logged. This is achieved without explicitly accumulating past measurements locally at each robot. This capability is necessary to support increasing number of robots in a team, under finite communication and computation resources. The design includes a novel fusion strategy, a consistent estimation method, and a state based initialization method, embedded in a distributed target tracking framework. The design is efficient in terms of computation demand, since it scales linearly with the number of robots. Additionally, the design is efficient in terms of communication demand, since communication is neither required to be synchronized with sensor readings, nor constrained to a specific network topology. The paper validates the proposed approach for its initialization capability, consistency of estimates, and robustness of performance, through several numerical simulations and using a publicly available multi-robot data set.

Parallel continuum robots have the potential to provide multi-degree-of-freedom articulation using a structure that is simple, compact, compliant, and highly scalable. These characteristics may be useful in micromanipulation, endoscopic robotic-assisted surgery, and human-robot interaction. Our prior work formulated a kinematic model which treats a parallel continuum robot as a set of multiple Cosserat rods with coupled boundary conditions. In this paper, we detail methods for the efficient numerical solution of this model at rates that enable real-time interactive simulation, motion planning, design optimization, and control. Exploitation of the model structure enables a significant reduction in the number of integrations required to evaluate the boundary value Jacobian matrix used in a shooting method. Our approach is used to teleoperate a prototype robot using real-time inverse kinematics solutions, and simulation tests show that inverse kinematics solutions are consistently computed at rates of several kilohertz using standard desktop computing hardware.

The ability to localize a robot is an important capability and matching of observations under substantial changes is a prerequisite for robust long-term operation. This paper investigates the problem of efficiently coping with seasonal changes in image data. We present an extension of a recent approach [15] to visual image matching using sequence information. Our extension allows for exploiting GPS priors in the matching process to overcome the main computational bottleneck of the previous method and to handle loops within the image sequences. We present an experimental evaluation using real world data containing substantial seasonal changes and show that our approach outperforms the previous method in case a noisy GPS pose prior is available.

This paper presents an efficient framework for the categorization of objects in real-world scenes (captured with an RGB-D sensor). The proposed framework uses ensembles of randomized decision trees in a hierarchical cascaded architecture to compute consistent object-class inferences of unseen objects. Specifically, the proposed framework computes object-class probabilities at three levels of an image hierarchy (i.e., pixel-, surfel-, and object-levels) using Random Forest classifiers. Next, these probabilities are fused together to compute a cumulative probabilistic output which is used to infer object categories. This fusion results in an improved object categorization performance compared with the state-of-the-art methods.

Co-verbal gestures are an important part of human communication, improving its efficiency for information conveyance. A key component of such improvement is the observer's ability to integrate information from the two communication channels, speech and gesture. Whether such integration also occurs when the multi-modal communication information is produced by a humanoid robot, and whether it is as efficient as for a human communicator, is an open question. Here, we present an experiment which, using a fully within subjects design, shows that for a range of iconic gestures, speech and gesture integration occurs with similar efficiency for human and for robot communicators. The gestures for this study were produced on an Aldebaran Robotics NAO robot platform with a Kinect based tele-operation system. We also show that our system is able to produce a range of iconic gestures that are understood by participants in unimodal (gesture only) communication, as well as being efficiently integrated with speech. Hence, we demonstrate the utility of iconic gestures for robotic communicators.

In a teleoperated system, misalignment between the master and slave manipulators can result from clutching, errors in the kinematic model, and/or sensor errors. This study examines the effects of type and magnitude of misalignment on the performance of the teleoperator. We first characterized the magnitude and direction of orientation misalignment created when clutching and unclutching during use of two surgical robots: the Raven II and the da Vinci Research Kit. We then purposely generated typical misalignments in order to measure the impact of such misalignment on user performance of a peg transfer task with the Raven II. Users were able to compensate for misalignment angles up to approximately 20 degrees in both tool orientation and camera viewpoint misalignment. These results can be used to guide the design and control of teleoperated systems for a variety of applications.

In this paper we investigate effects of the mass, kinematic constraints imposed by the joint, and assistance provided by the spring of a pair of quasi-passive knee exoskeletons on the motion of the human body center of mass during normal walking. The exoskeletons implement a spring in parallel with the knee joint in the weight acceptance phase of gait, and allow free rotation during all other phases. We begin with a brief explanation of the exoskeleton design, which employs a friction-based latching mechanism to engage/disengage a spring in parallel with the knee. Additionally, a pair of joint-less mass replicas of the exoskeletons were used to separately investigate the effects of the exoskeleton added mass and articulation. It was found that the exoskeleton mass is the main contributor to the changes in the motion of the center of mass, with more pronounced fluctuations of the center of mass in the mediolateral direction, while the exoskeleton joint and spring had negligible effects over and above those of the mass. Additionally, the exoskeleton mass and assistance conditions respectively resulted in a non-significant increase and a non-significant decrease in the total mechanical work of the body.

We present a new framework for reactive synthesis that considers the dynamics of the robot when synthesizing correct-by-construction controllers for nonlinear systems. Many high-level synthesis approaches employ discrete abstractions to reason about the dynamics of the continuous system in a simplified manner. Often, these abstractions are expensive to compute. We circumvent the need to have detailed abstractions for nonlinear systems by proposing a framework for adapting abstractions based on partial solutions to the low-level controller synthesis problem. The contribution of this paper is a reactive synthesis algorithm that makes use of our adaptation procedure to update the high-level strategy each time the non-deterministic discrete abstraction is modified. We combine this with a verified low-level controller synthesis scheme capable of automatically synthesizing controllers for a wide class of nonlinear systems. This novel synthesis framework is demonstrated on a dynamical robot executing an autonomous inspection task.

This paper presents a repetitive hopping control of biped robots, which has the following properties: 1) It can initiate or cease hopping at any time, 2) The peak altitude can be freely altered during hopping, and 3) The impact at touchdown is suppressed. Hopping motions accompany frequent changes of contact conditions with the environment and lead the discontinuity of dynamics they are governed, therefore, it is difficult to achieve the motion in a way without relying any passive mechanisms such as springs and dampers. Three controllers, that is, (I) a regulator to stabilize, (II) a controller to converge a desired velocity when lifting off the ground and (III) a controller to absorb the landing impact are unified by the global scheme to transit each other depending on a commanded value and the state of the robot.

The goal of this work is to develop a soft robotic manipulation system that is capable of autonomous, dynamic, and safe interactions with humans and its environment. First, we develop a dynamic model for a multi-body fluidic elastomer manipulator that is composed entirely from soft rubber and subject to the self-loading effects of gravity. Then, we present a strategy for independently identifying all unknown components of the system: the soft manipulator, its distributed fluidic elastomer actuators, as well as drive cylinders that supply fluid energy. Next, using this model and trajectory optimization techniques we find locally optimal open-loop policies that allow the system to perform dynamic maneuvers we call grabs. In 37 experimental trials with a physical prototype, we successfully perform a grab 92% of the time. By studying such an extreme example of a soft robot, we can begin to solve hard problems inhibiting the mainstream use of soft machines.

This paper presents a method for the generation of dynamic walking gaits with a 3D Dual-SLIP model and its application to a simulated Hubo+ based humanoid. Previous approaches with the Dual-SLIP model have only focused on the planar case, wherein self-stable gaits can be found. When extended to 3D here, this model has not been found to exhibit self-stable gaits, requiring new methods for gait optimization and control. By taking advantage of a newly discovered symmetry condition for the Dual-SLIP model, this work proposes a quarter period (half step) optimization process to find periodic walking gaits in 3D. An LQR controller is developed to regulate the state of the model at leg midstance (MS) based on its return map dynamics. The Dual-SLIP model is extended by introducing a bio-inspired leg length actuation scheme in order to describe high-speed walking gaits (up to 2 m/s for human-compatible parameters). Finally, the CoM trajectory and footstep positions from the 3D Dual-SLIP are used as a reference in a task-space controller with a Hubo+ based humanoid model. By tracking these references, the methods successfully produce human-like dynamic walking gaits in simulation which are robust to disturbances. The whole-body control system for walking can handle uneven terrain with variation up to 10% of its leg length. This represents the first humanoid dynamic walking approach based on a 3D Dual-SLIP model.

Several sampling-based algorithms have been recently proposed that ensure asymptotic optimality. The convergence of these algorithms can be improved if sampling is guided toward the most promising region of the search space where the solution is more likely to be found. In this paper we propose three sample rejection methods that leverage the classification of the samples according to their potential of being part of the optimal solution to guide the exploration of the motion planner to promising regions of the search space. These sampling strategies are a direct by-product of the exploitation phase of the algorithm, which uses a dynamic programming (DP) step while planning on random graphs as, for example, is done in the RRT# algorithm. It is shown that the proposed sampling strategies are able to compute high-quality solutions, much faster than existing algorithms. We provide numerical results and compare the performance of the proposed algorithm with the original RRT# and the RRT* algorithms.

This paper discusses a dynamic nonprehensile manipulation by using a vibrating plate. The manipulation method, where two DOF (degrees of freedom) motion of an object on the plate surface is controlled by a single active joint, is proposed. For the plate vibration mechanism, we newly introduce an active-passive hybrid joint whose two axes are arranged in nonparallel to each other. This mechanism features the plate vibration trajectory is variable based on the input frequency. We then analytically derive the trajectory of object on the vibrating plate. Through this analysis, we reveal that two DOF motion of the object can be controlled by utilizing the variable vibration effect. Finally, we show experimental result by using the prototype robot, for confirming the validity of the proposed method.