ICRA 2015 论文列表

This paper presents a Therapist-in-the-Loop (TIL) framework for robotics-assisted mirror rehabilitation therapy integrated with adaptive Assist-as-Needed (ANN) training, to be adjusted based on the impairment and disability level of the patient's affected limb. Closed-loop system stability has been investigated using a combination of the Circle Criterion and the Small-Gain Theorem to account both for time-delay and the time-varying adaptive ANN training. Experiments to investigate the performance of the proposed framework are reported.

This paper addresses a synchronization problem that arises when a team of aerial robots (ARs) need to communicate while performing assigned tasks in a cooperative scenario. Each robot has a limited communication range and flies within a previously assigned closed path. When two robots are close enough, a communication link may be established allowing the robots to share information. The goal is to schedule the flights such that the entire system can be synchronized for maximum information exchange, that is, every pair of neighbors are on the feasible communication link at the same time. We propose an algorithm for scheduling a team of robots in this scenario and propose a robust framework where the synchronization of a large team of robots is assured. The approach allows us to design a fault-tolerant system that can be used for multiple tasks such as surveillance, area exploration, searching for targets in a hazardous environment, and assembly and structure construction, to name a few.

Autonomous systems that navigate in unknown environments encounter a variety of planning problems. The success of any one particular planning strategy depends on the validity of assumptions it leverages about the structure of the problem, e.g., Is the cost map locally convex? Does the feasible state space have good connectivity? We address the problem of determining suitable motion planning strategies that can work on a diverse set of applications. We have developed a planning system that does this by running competing planners in parallel. In this paper, we present an approach that constructs a planner ensemble - a set of complementary planners that lever-age a diverse set of assumptions. Our approach optimizes the submodular selection criteria with a greedy approach and lazy evaluation. We seed our selection with learnt priors on planner performance, thus allowing us to solve new applications without evaluating every planner on that application. We present results in simulation where the selected ensemble outperforms the best single planner and does almost as well as an off-line planner. We also present results from an autonomous helicopter that has flown missions several kilometers long at speeds of up to 56m/s which involved avoiding unmapped mountains, no-fly zones and landing in cluttered areas with trees and buildings. This work opens the door on the more general problem of adaptive motion planning.

We consider in this paper cable driven parallel robot (CDPR) with m catenary cables that present a significant deformation due to their elasticity and own mass and we address the kinematics of such a robot under the assumption of static state. We show that the inverse kinematics (IK) requires to solve a square system of equations if the robot has up to 6 cables and an underconstrained system if m > 6. We first show that usual methods for solving the IK of CDPR with other cable model cannot be used and we propose an algorithm that allows one to determine all possible solutions. We then address the forward kinematics (FK) that is known to be usually a difficult task for parallel robots. We show that solving the FK always amounts to solve a square system of equations and that the problem is even more complex than solving the FK of parallel robots with rigid legs as the system is non algebraic and has a larger number of equations. We then propose a generic algorithm for solving the FK for any number of cables. We illustrate the use of the IK and FK solver on a CDPR with 8 cables.

Robust, scalable place recognition is a core competency for many robotic applications. However, when revisiting places over and over, many state-of-the-art approaches exhibit reduced performance in terms of computation and memory complexity and in terms of accuracy. For successful deployment of robots over long time scales, we must develop algorithms that get better with repeated visits to the same environment, while still working within a fixed computational budget. This paper presents and evaluates an algorithm that alternates between online place recognition and offline map maintenance with the goal of producing the best performance with a fixed map size. At the core of the algorithm is the concept of a Summary Map, a reduced map representation that includes only the landmarks that are deemed most useful for place recognition. To assign landmarks to the map, we use a scoring function that ranks the utility of each landmark and a sampling policy that selects the landmarks for each place. The Summary Map can then be used by any descriptor-based inference method for constant-complexity online place recognition. We evaluate a number of scoring functions and sampling policies and show that it is possible to build and maintain maps of a constant size and that place-recognition performance improves over multiple visits.

In this paper, we show that the free configuration space of a Kirchhoff elastic rod is path-connected. By free configuration space, we mean the set of all equilibrium configurations of the rod that are stable (i.e. locally minimize elastic potential energy) and do not experience self-intersections. We also provide semi-analytical expressions for paths in the free configuration space that connect any two stable equilibrium configurations that do not contain self-intersections. These results are applied to the problem of manipulation planning for deformable objects.

In this paper, we explore the exponential map and its inverse, the logarithm map, for the group SIM(n) of similarity transformations in ℝn which are the composition of a rotation, a translation and a uniform scaling. We give a formula for the exponential map and we prove that it is surjective. We give an explicit formula for the case of n = 3 and show how to efficiently compute the logarithmic map. As an application, we use these algorithms to perform motion interpolation. Given a sequence of similarity transformations, we compute a sequence of logarithms, then fit a cubic spline that interpolates the logarithms and finally, we compute the interpolating curve in SIM(3).

This paper concerns itself with a relatively unexplored type of personal robot that operates in the tool space. Handheld robots aim to cooperate with the user to solve tasks and improve what tools can offer enhanced by actuation, sensing, and importantly, task knowledge. To this end, we devised a new lightweight robotic platform that has 4 DoF and uses a cable driven continuum structure. Feedback from the robot to the user is provided in an intuitive, implicit manner by the robot end effector pointing towards the goal, avoiding pointing, and/or refusing to perform an action when it conflicts with the task specification. We evaluate two generic tasks involving aiming in space and picking/placing objects with a number of volunteers. Repeated measures ANOVA is used to analyse results to show in which conditions an increased level of automation in the handheld robot improves task performance or user perception of task load. The robot is offered as an open robotics platform[1] and the results indicate directions to improve on feedback and interaction mechanisms.

Pendulum models have been studied as benchmark problems for development of nonlinear control schemes, as well as reduced-order models for the dynamics analysis of locomotion of humanoid robots. This work provides a generalization of the previously introduced Reaction Mass Pendulum (RMP), which is a multibody inverted pendulum model, to a bipedal model that can better model bipedal locomotion. The RMP consists of an extensible “leg” and a “body” with moving proof masses that give rise to a variable rotational inertia. The Reaction Mass Biped (RMB) introduced here has two legs, one of which takes the role of a stance leg and the other performs as a swing leg during bipedal locomotion. The bipedal walking dynamics model of the RMB is therefore hybrid, with the roles of stance leg and swing leg interchanged after each cycle. The dynamics model is developed using a variational mechanics approach, without using generalized coordinates for the rotational degrees of freedom. This dynamics model has thirteen degrees of freedom, all of which are considered to be actuated in the control design. A set of desired state trajectories that can enable bipedal walking in straight and curved lines are generated. A control scheme is then designed for asymptotically stable tracking of this set of trajectories with an almost global domain of attraction. Numerical simulation results confirm the stability of this tracking control scheme for different walking trajectories of the RMB.

On-orbit servicing involves a new class of space missions in which a servicer spacecraft is launched into the orbit of a target spacecraft, the client. The servicer navigates to the client with the intention of manipulating it, using a robotic arm. Within this framework, this work presents a new robotic experimental facility which was recently built at the DLR to support the development and experimental validation of such orbital servicing robots. The facility allows reproducing a close-proximity scenario under realistic three-dimensional orbital dynamics conditions. Its salient features are described here, to include a fully actuated macro-micro system with multiple sensing capabilities, and analyses on its performance including the amount of space environment volume that can be simulated.

Current multi-agent robotic testbeds are prohibitively expensive or highly specialized and as such their use is limited to a small number of research laboratories. Given the high price tag, what is needed to scale multi-agent testbeds down both in price and size to make them accessible to a larger community? One answer is the GRITSBot, an inexpensive differential drive microrobot designed specifically to lower the entrance barrier to multi-agent robotics. The robot allows for a straightforward transition from current ground-based systems to the GRITSBot testbed because it closely resembles expensive platforms in capabilities and architecture. Additionally, the GRITSBot's support system allows a single user to easily operate and maintain a large collective of robots. These features include automatic sensor calibration, autonomous recharging, wireless reprogramming of the robot, as well as collective control.

We describe the National Museum of Mathematics's Robot Swarm exhibit and our approach for achieving a reliable system for collision avoidance. The Robot Swarm exhibit allows visitors to program behaviors and interact with a “swarm” of small robots. The exhibit supports extended unattended run times, continuous interaction with the public and the demonstration of evocative group behaviors. The exhibit software includes a robust collision avoidance scheme that prevents collisions between robots and collisions between robots and static obstructions in the exhibit space. This system was achieved by building on the Optimal Reciprocal Collision Avoidance (ORCA) algorithm in a novel implementation: the Extended Velocity Obstacle validation system. This paper presents: 1) A collision avoidance algorithm that robustly and efficiently avoids collisions between many robots and static obstacles. 2) A unique hybrid ORCA collision avoidance approach that utilizes global state knowledge without subverting the behavioral independence of each robot. 3) A unique position filtering system which is tailored to an error model in which positions reads can be treated as “ground-truth” and a noise model that is highly discontinuous and non-linear. We present experiments and experimental data that demonstrate the efficacy of our approach.

Robotic navigation applications often require on-line generation of trajectories that respect underactuated non-linear dynamics, while optimizing a cost function that depends only on a low-dimensional workspace (collision avoidance). Approaches to non-linear optimization, such as differential dynamic programming (DDP), suffer from the drawbacks of slow convergence by being limited to stay within the trust-region of the linearized dynamics and having to integrate the dynamics with fine granularity at each iteration. We address the problem of decoupling the workspace optimization from the enforcement of non-linear constraints. In this paper, we introduce the Dynamics Projection Filter, a nonlinear projection operator based approach that first optimizes a workspace trajectory with reduced constraints and then projects (filters) it to a feasible configuration space trajectory that has a bounded sub-optimality guarantee. We show simulation results for various curvature and curvature-derivatives constrained systems, where the dynamics projection filter is able to, on average, produce similar quality solution 50 times faster than DDP. We also show results from flight tests on an autonomous helicopter that solved these problems on-line while avoiding mountains at high speed as well as trees and buildings as it came in to land.

Parametric filters, such as the Extended Kalman Filter and the Unscented Kalman Filter, typically scale well with the dimensionality of the problem, but they are known to fail if the posterior state distribution cannot be closely approximated by a density of the assumed parametric form.

The high stiffness of conventional robots is beneficial in attaining highly accurate positioning in free space. High stiffness, however, limits a robot's ability to perform constrained manipulation. Because of the high stiffness, geometric conflict between the robot and task constraints during constrained manipulation can lead to excessive forces and task failure. Variable stiffness actuators can be used to adjust the stiffness of robot joints to allow high stiffness in unconstrained directions and low stiffness in constrained directions. Two important design criteria for variable stiffness actuation are a large range of stiffness and a compact size. A new design, the Arched Flexure VSA, uses a cantilevered beam flexure of variable cross-section and a controllable load location. It allows the joint to have continuously variable stiffness within a finite stiffness range, have zero stiffness for a small range of joint motion, and allow rapid adjustment of stiffness. Using finite element analysis, flexure geometry was optimized to achieve high stiffness in a compact size. A proof-of-concept prototype demonstrated continuously variable stiffness with a ratio of high stiffness to low stiffness of 55.

This paper introduces the Affordance Template ROS package for quickly programming, adjusting, and executing robot applications in the ROS RViz environment. This package extends the capabilities of RViz interactive markers [1] by allowing an operator to specify multiple end-effector waypoint locations and grasp poses in object-centric coordinate frames and to adjust these waypoints in order to meet the run-time demands of the task (specifically, object scale and location). The Affordance Template package stores task specifications in a robot-agnostic JSON description format such that it is trivial to apply a template to a new robot. As such, the Affordance Template package provides a robot-generic ROS tool appropriate for building semi-autonomous, manipulation-based applications. Affordance Templates were developed by the NASA-JSC DARPA Robotics Challenge (DRC) team and have since successfully been deployed on multiple platforms including the NASA Valkyrie and Robonaut 2 humanoids, the University of Texas Dreamer robot and the Willow Garage PR2. In this paper, the specification and implementation of the affordance template package is introduced and demonstrated through examples for wheel (valve) turning, pick-and-place, and drill grasping, evincing its utility and flexibility for a wide variety of robot applications.

The inability of current robotic perception techniques to adequately detect non-geometric terrain hazards is a primary cause of failure for robots operating in natural terrain on Mars, the Moon, and Earth. Classical approaches detect surface appearance but do not measure the underlying mechanical properties that determine wheel-terrain interaction. Diurnal temperature variations of a granular material, however, are strongly correlated with both its surface appearance and subsurface geophysical properties. This paper presents a technique for determining relative differences in looseness and traversability of granular terrain through analysis of thermal imagery. Terrain compaction and traversability are predicted by estimating a material's thermal inertia from observations of thermal transients. Results from a set of experiments in sandy terrain demonstrate the ability of this approach to differentiate between safe, compact and hazardous, loose terrain.

Lung cancer is the most deadly form of cancer, and survival depends on early-stage diagnosis and treatment. Transoral access is preferable to traditional between-the-ribs needle insertion because it is less invasive and reduces risk of lung collapse. Yet many sites in the peripheral zones of the lung or distant from the bronchi cannot currently be accessed transorally, due to the relatively large diameter and lack of sufficient steerablity of current instrumentation. To remedy this, we propose a new robotic system that uses a tendon-actuated device (bronchoscope) as a first stage for deploying a concentric tube robot, which itself is a vehicle through which a bevel steered needle can be introduced into the soft tissue of the lung outside the bronchi. In this paper we present the various components of the system and the workflow we envision for deploying the robot to a target using image guidance. We describe initial validation experiments in which we puncture ex vivo bronchial wall tissue and also target a nodule in a phantom with an average final tip error of 0.72 mm.

This paper proposes a soft, inflatable manipulator that is antagonistically actuated by tendons and pneumatics. The combination of the two actuation mechanisms in this antagonistic robot structure is inspired by the octopus which uses its longitudinal and transversal muscles to steer, elongate, shrink and also stiffen its continuum arms. By “activating” its antagonistic muscle groups at the same time, the octopus can achieve multiple motion patterns as well as stiffen their arms. Being organized in a similar fashion, our robot manipulator uses, on the one hand, pneumatic actuation and, on the other hand, tendon-based actuation - one opposing the other, achieving an overall antagonistic actuation framework. Controlling the pressure inside the robot while at the same time controlling the tendons' displacements, the robot can be moved into a wide range of configurations while simultaneously controlling the arm's stiffness. This paper builds on earlier work by the authors: Here, we present a new conic-shaped manipulator structure and the control architecture. Using a constant curvature model, we have derived an approach suitable for controlling the robot manipulator. The manipulator's reachable workspace is analyzed and proof-of-concept experiments were conducted to show the robot's stiffness control and motion abilities.