ICRA 2015 论文列表

This paper presents a novel volumetric display based on projection on a non-planer and reconfigurable fog screen. Unlike conventional fog projection systems which produce 2D images on flat screens, our display scatters different parts of the projected image at different depth levels, thus allowing volumetric data to be displayed in the real 3D space. We constructed the fog screen with a 2D array of nozzles that are individually switchable, while the switching pattern is tightly synchronized with the video content. Our system is superior to many existing approaches at many levels. First, our display does not require head tracking, glasses or head-mounted devices while allowing high resolution, full color 3D image to be observed from wide viewing angles by many people at the same time. As compare with various existing approaches, our system is relatively easy to setup and low cost. Most importantly, our immaterial, mid-air display allows users to directly touch and manipulate virtual objects in 3D under marker-free and barrier-free settings which opens up immense tangible and creative interaction possibilities. In this paper, we provide the details of display mechanism and design prototype, as well as a constrained optimization problem to find the projection distance that can maximize the display resolution. A number of real display examples will demonstrate the performance of the proposed system.

An underactuated robotic finger is proposed and studied for capturing non-cooperative objects, featuring an underactuated rigid-mechanism that can brake residual motion of non-cooperative targets. The finger consists of three joints driven by two actuators, and a set of one-input-two-output planetary gears is introduced to realize force transmission by using the ring gear and the planetary carrier to drive the two joints. The rotation of the ring gear and the planetary carrier are governed jointly by the inputs and the external force. In the process of capturing a moving target, the proposed robotic finger can effectively limit the impact force peak due to collision with the target. Moreover, the recovery time of the joint can be adjusted by changing the input torque of the motor, which can be set to achieve optimal recovery process for different tasks. A prototype has been built to study the proposed design of the robotic finger, and experimental results have confirmed its desirable characteristics.

For XYΘ-type flexure-based manipulators with nanometer accuracy, the coupled linear/angular motions of the end effector cause unwanted misalignments to the measurement system, leading to measurement error or even measurement failure. This paper proposes an indirect measuring method for such manipulators, where the 3-PRR topology is adopted to kinematically transfer the coupled motions of the end effector to the linear motions of three prismatic joints. Unlike the other XYΘ manipulators, three linear position sensors are used to measure the linear motions of the prismatic joints, not the coupled motions of the end effector. Accordingly, the position and orientation of the end effector can be obtained via the forward kinematics. This indirect method aims to eliminate the influence of the coupled motions of the end effector on the measurement system, and to guarantee the accuracies and effectiveness of the sensors' outputs. An XYΘ flexure-based manipulator is developed following this method. The design and kinematics modeling is presented, and the effectiveness of this method is computationally verified. The future work will focus on the experimental verification of this method.

Automatic robot / Computed Tomography (CT) scanner registration is an important feature for robot-assisted percutaneous needle placement under CT-scanner. This registration can be done using 3D images, but for fast, low X-ray radiation it is interesting to be able to perform the registration with a single slice. In this paper, a new marker is proposed, which allows to estimate the pose of a device using a single slice. This marker, called ZCM, consists of three circles or ellipses arranged in a Z-shape configuration. It is shown that it provides a larger workspace (i.e. a larger set of visible configurations) for pose measurement than the standard Brown-Roberts-Wells, while maintaining a good accuracy. A closed-form method is proposed for solving the pose estimation with this marker. Simulations and experimental tests using a mock-up patient-mounted robot are presented and confirm the theoretical analysis.

Ocean sampling for highly temporal phenomena, such as harmful algal blooms, necessitates a vehicle capable of fast aerial travel interspersed with an aquatic means of acquiring in-situ measurements. Vehicle platforms with this capability have yet to be widely adopted by the oceanographic community. Several animal examples successfully make this aerial/aquatic transition using a flapping foil actuator, offering an existence proof for a viable vehicle design (Fig. 1).We discuss a preliminary realization of a flapping wing actuation system for use in both air and water. The wing employs an active in-line motion degree of freedom to generate the large force envelope necessary for propulsion in both fluid media.

In this work, we present the SmartBridge, a novel bridge section model equipped with actively controlled arrays of flaps aiming at mitigating wind-induced vibrations of long-span bridges. The active model, as well as its support structure, is described in detail and the key design choices are motivated. Finally, the capabilities of the system and the active control of the bridge section model with moving flaps was validated by wind tunnel experiments. In spite of the relative simplicity of a, manually tuned, control law, the results are encouraging and show a significant damping of the pitch vibration of the deck.

This paper proposes a novel approach to generate trajectories that generalize given demonstrations according to optimality criteria. By formulating the problem as a quadratic program we can efficiently incorporate constraints to adapt to new desired motion requirements while achieving the main goal of matching the acceleration profile of the demonstration. This makes our method particularly suited for the imitation and generalization of trajectories such as hitting movements, where it is crucial to maintain the dynamic traits of the demonstration while respecting strict requirements for the goals position, velocity and time. Our method draws inspiration from the Dynamical Movement Primitives (DMPs) framework, preserving its desirable properties of flexibility and rejection of disturbances during execution. Moreover, it offers an higher degree of control on the generated solution, allowing for example i) to limit the instantaneous positions, velocities and accelerations during the whole trajectory, and ii) to add intermediate way points that were not present in the demonstration. With current state-of-the-art solvers of quadratic programs, a problem with hundreds of parameters can be solved in tens of milliseconds in a standard computer, allowing practical applications. Our methodology results in trajectories with a very good approximation of the shape traits of the demonstration, with additional flexibility in specifying constraints of the generated trajectory.

Recent work [1], [2], [3] has shown promising results in learning from demonstrations for the manipulation of deformable objects. Their approach finds a non-rigid registration between points in the demonstration scene and points in the test scene. This registration is then extrapolated and applied to the gripper motions in the demonstration scene to obtain the gripper motions for the test scene. If more than one demonstration is available, a quality score for the non-rigid registration is used to determine the best matching training scene. For many manipulation tasks, however, the gripper's direction of approach with respect to the objects' surface normals is important in order to succeed at the task. This prior work only registers points across scenes and does not register the surface normals, often leading to warps between scenes that are inappropriate for transfer of manipulation primitives. The main contributions of this paper are (i) An algorithm for non-rigid registration that considers both points and normals, and (ii) An evaluation of this registration approach in the context of learning from demonstrations for robotic manipulation. Our experiments, which consider an insertion task in simulation and also knot-tying and towel-folding executions in a PR2, show that incorporating normals results in improved performance and qualitatively better grasps.

For self-organized multi-robot systems, one of the widely studied task domains is object clustering, which involves gathering randomly scattered objects into a few piles. Earlier studies have pointed out that environmental boundaries influence the cluster formation process, generally causing clusters to form around the perimeter rather than centrally within the workspace. But it is usually central clusters that are desired in robotic clustering systems. In this paper, we derive general conditions that prevent the problem of boundaries causing perimeter clusters. We develop a mathematical model to explain how sets of clusters evolve into a single cluster without any boundary cluster being formed. Through analysis of the model, we show that time-averaged spatial densities of the robots play a significant role in producing conditions that ensure a single central cluster emerges. Thus, local densities of robots can be considered a system-level control parameter to achieve this task. We further investigate how the physical packing of clusters affects clustering dynamics. To do this, we introduce a measure of scaled compactness and show that the lifetime of clusters is well predicted by this descriptor.

This paper introduces the Manipulation Task Model (MTM), an algebraic approach to describe manipulation tasks for robotic systems. The model is based on dual quaternion algebra to describe poses, linear and angular velocities, and forces and moments, all of which are joined to form the algebraic group of actions. We then use this group together with process algebra to describe temporal and logical relations between actions composing manipulation tasks. A simulation example of a cleaner robot shows that the proposed model is comprehensive and expressive enough, and that it provides robustness in relation to perturbations at the task level. Future developments will include merging the model with underlying controllers in order to provide a unified means for description, planning, and execution of manipulation tasks for robotic systems.

To control exoskeletons for walking gait assistance, it is of primary importance to control them to act synchronously with the gaits of users. To effectively estimate the gait cycle (or the phase within a stride) of users, we propose a new adaptive frequency oscillator (AFO). While previous AFOs successfully estimated the walking frequency from joint angles as inputs, the new AFO, called particularly-shaped adaptive oscillator (PSAO) can estimate gait cycle from the same inputs, which would have required foot contact sensors in previous approaches. To predict the effects of PSAO-based gait assistance on human walking, it has been tested with neuromuscular walking simulation. In the simulation, the gait assistance system reduced the metabolic cost of walking for some assistance patterns. The walk ratio (step length per step rate) also changed as assistance patterns shifted in phase, which is meaningful because metabolic cost of walking in general is minimal at specific walk ratio. For a prototype exoskeleton we developed, the effect of gait assistance was experimented on a human subject walking on level ground and inclining slopes to verify the predictions from the simulation: (1) physiological cost index computed from heart rate significantly decreased indicating reduction in metabolic energy expenditure; (2) walk ratio was in fact controllable to an extent.

Can you move along a surface in space without propellant or mechanical contact? This paper presents a new actuator, the induction coupler, which generates an eddy-current force between a robotic inspector and the conductive exterior of its target. This force allows the inspector to crawl along the surface of a target without mechanical contact. Sets of induction couplers composed of spinning arrays of permanent magnets can exert control force and torque in all six rigid-body degrees of freedom by strategically repelling and shearing across the surface of the target. This paper uses an analytical model of eddy-current forces to simulate the set of maneuvers necessary to produce the control force and torque that can move and orient a robotic orbital inspector. Experiments on a low-friction test bed demonstrate a successful implementation of the actuator and verify the feasibility of this approach.

A multilayer architecture to control multirotor UAVs equipped with a servo robot arm is proposed in this paper. The main purpose is to control the aerial platform taking into account the presence of the moving manipulator. Three layers are considered in this work. First, a novel mechanism is proposed considering a moving battery to counterweight the statics of the robotic arm. Then, in order to overcome the mechanical limitations of the previous layer, the residual of the arm static effects on the UAV is computed and compensated through the given control thrust and torques. Finally, an estimator of external forces and moments acting on the aerial vehicle is considered and the estimations are fed back to the controller to compensate neglected aerodynamic effects and the arm dynamics. The performance of the proposed architecture has been experimentally evaluated.

Tumor localization in traditional lung resection surgery requires manual palpation of the deflated lung through a thoracotomy. It is a painful procedure that is not suitable for many patients. Therefore, a multisensory mechatronic device was designed to localize tumors using a minimally invasive approach. The device is sensorized with tactile, ultrasound and position sensors in order to obtain multimodal data of soft tissue in real time. This paper presents the validation of the efficiency and efficacy of this device via an ex vivo experimental study. Tumor pathology was simulated by embedding iodine-agar phantom tumors of varying shapes and sizes into porcine liver tissue. The device was then used to palpate the tissue to localize and visualize the simulated tumors. Markers were then placed on the location of the tumors and fluoroscopic imaging was performed on the tissue in order to determine the localization accuracy of the device. Our results show that the device localized 87.5% of the tumors with an average deviation from the tumor center of 3.42 mm.

An underwater glider named ROUGHIE (Research Oriented Underwater Glider for Hands-on Investigative Engineering) is designed and manufactured to provide a test platform and framework for experimental underwater automation. This paper presents an efficient multi-level motion controller that can be used to enhance underwater glider control systems or easily modified for additional sensing, computing, or other requirements for advanced automation design testing. The ultimate goal is to have a fleet of modular and inexpensive test platforms for addressing the issues that currently limit the use of autonomous underwater vehicles (AUVs). Producing a low-cost vehicle with maneuvering capabilities and a straightforward expansion path will permit easy experimentation and testing of different approaches to improve underwater automation.

Concentric tube robots are thin, tentacle-like devices that can move along curved paths and can potentially enable new, less invasive surgical procedures. Safe and effective operation of this type of robot requires that the robot's shaft avoid sensitive anatomical structures (e.g., critical vessels and organs) while the surgeon teleoperates the robot's tip. However, the robot's unintuitive kinematics makes it difficult for a human user to manually ensure obstacle avoidance along the entire tentacle-like shape of the robot's shaft. We present a motion planning approach for concentric tube robot teleoperation that enables the robot to interactively maneuver its tip to points selected by a user while automatically avoiding obstacles along its shaft. We achieve automatic collision avoidance by precomputing a roadmap of collision-free robot configurations based on a description of the anatomical obstacles, which are attainable via volumetric medical imaging. We also mitigate the effects of kinematic modeling error in reaching the goal positions by adjusting motions based on robot tip position sensing. We evaluate our motion planner on a teleoperated concentric tube robot and demonstrate its obstacle avoidance and accuracy in environments with tubular obstacles.

A model predictive controller is developed for following the position of a human dancer in robot ballroom dancing. The control design uses a dynamic model of a dancer, based on a variant of the so-called 3D Linear Inverted Pendulum Mode that includes also the swing foot. This model serves as a basis for a Kalman predictor of the human motion during the single-support phase, while a simpler kinematic technique is used during the double-support phase. The output of the prediction filter enables to design a Model Predictive Control (MPC) law, by recursively solving on line and within a preview window a convex linear-quadratic optimization problem, constrained by differential kinematic bounds on robot commands. Two different control strategies, either at the velocity or at the acceleration level, are proposed and compared in simulations and in actual experiments. Accurate and reactive behaviors are obtained by the ballroom robot follower, confirming the benefit of the predictive/filtering nature of a MPC approach to handle uncertainty of human intentions and noisy signals.

A semi-autonomous mobile self-leveling landing platform designed to launch, recover and re-launch VTOL UAVs without the need for human intervention is described. The landing platform is rugged, lightweight and inexpensive, making it ideal for civilian applications that require a base station from which a rotorcraft UAV can be launched and/or recovered on terrain that is normally unsuitable for UAV take-off and landing. This landing platform is capable of autonomously self-leveling on rough terrain and inclines up to 25°, and can operate in isolated remote locations for extended periods of time using large onboard lithium batteries and wireless communication. The unique design aspects of this landing platform are that it is mobile, self-leveling, and man-portable. A fully-operational prototype has been designed, constructed and evaluated. Design details and experimental results are presented to demonstrate the landing platform's functionality, and that all primary design requirements have been met.

Transanal Endoscopic Microsurgery (TEM) is a minimally invasive oncological resection procedure that utilises a natural orifice approach rather than the traditional abdominal or open approach. However, TEM has a significant recurrence rate due to incomplete excisions, which can possibly be attributed to the absence of intraoperative image guidance. The use of real-time histological data could allow the surgeons to assess the surgical margins intraoperatively and adjust the procedure accordingly. This paper presents the integration of endomicroscopy and ultrasound imaging through a robotically actuated instrument. Endomicroscopy can provide high resolution images at a surface level while ultrasound provides depth resolved information at a macroscopic level. Endomicroscopy scanning is achieved with a novel scanning approach featuring a passive force adaptive mechanism. The instrument is manipulated across the surgical workspace through an articulated flexible shaft. This results in the ability to perform large area mosaics coupled with ultrasound scanning. In addition, the use of endoscopic tracking is demonstrated, allowing three-dimensional reconstruction of the ultrasound data displayed onto the endoscopic view. An ex vivo study on porcine colon tissue has been performed, demonstrating the clinical applicability of the instrument.