ICRA 2017 Papers

This paper presents the static and dynamic modeling for a recently discovered artificial muscle-twisted and coiled actuator (TCA). This actuator can generate large force and displacement; moreover, it is low-cost, easy to fabricate, and customizable. Since the discovery of TCA, it has been widely adopted for various robotic applications. Nevertheless, theoretical models to describe the static performance and dynamic response are underexplored. In this paper, we aim to model the statics and dynamics for TCA from physics perspective. Specifically, the developed model utilizes parameters related to the working principle and material properties of the actuator. Experiments are conducted to verify the proposed model, and the results demonstrate that the proposed model can predict the static performance and dynamic response for the actuator. Given the wide applications of TCA in robotics, the developed model will enable closed-loop control of robotic systems with TCAs to achieve precise motion.

We present a novel method for extrinsically calibrating a camera and a 2-D Laser Rangefinder (LRF) whose beams are invisible from the camera image. We show that point-to-plane constraints from a single observation of a V-shaped calibration pattern composed of two non-coplanar triangles suffice to uniquely constrain the relative pose between two sensors. Next, we present an approach to obtain solutions using point-to-plane constraints from single or multiple observations. Along the way, we also show that previous solutions, in contrast to our method, have inherent ambiguities and therefore must rely on a good initial estimate. Real and synthetic experiments validate our method and show that it achieves better accuracy than previous methods.

A critical question to be answered to improve robotic rehabilitation is what is the optimal rehabilitation environment for a subject that will facilitate maximum recovery during therapy? Studies suggest that task variability, nature and degree of assistance or error-augmentation and type of feedback play a critical role in motor (re)-learning. In this work, we present a framework for robot-assisted motor (re)-learning that provides subject-specific training by allowing for simultaneous adaptation of task, assistance and feedback based on the performance of the subject on the task. We model a continuous and coordinated multi-joint task using a learning-from-demonstration approach, which allows the task to be modeled in a generative manner such that the challenge-level of the task could be modulated in an online manner. To train the subjects for dexterous manipulation, we present a torque-based task that requires the subject to dynamically regulate their joint torques. Finally, we carry out a pilot study with healthy human subjects using our previously developed hand exoskeleton to test a hypothesis and the results suggest that training under simultaneous adaptation of task, assistance and feedback positively affects motor learning.

This paper deals with the design of a new robust adaptive controller for parallel manipulators based on sliding mode and modelbased adaptive control. More precisely, the proposed controller relies on continuous finite-time terminal sliding mode (TSM) control and the linear-in-the-parameters property of the inverse dynamics of the manipulator. The main motivation behind the proposed scheme is to improve the tracking performance of fast and accurate parallel manipulators while guaranteeing the closed-loop system's robustness. Based on the linear-in-the-parameters property of the inverse dynamics of the manipulator, an adaptive law is proposed in order to estimate in real-time the dynamic parameters of the manipulator. The proposed controller has the advantage of relying on the desired reference trajectories instead of measured ones which can improve its robustness and efficiency. To demonstrate the effectiveness of the proposed controller, real-time experiments are conducted on a four-degree-of-freedom parallel manipulator called Veloce.

How to design a robotic hand for reproducing the move characteristics of human hand joints is a big challenge in robotics. In this paper, we present an approach to determine the actuation configuration based on the statistical results of hand joint angle in different grasps. A relationship between the basic statistical metrics and actuation configuration strategies is built. In this case, a novel actuation configuration is proposed and the joints of four fingers are arranged into five actuation modules. For the mechanical implementation, the motion of human four finger joints is decomposed to proportion motion, differential motion and chain proportion motion, mechanically implemented by pulley, planetary gear differential module and gear transmission chain. Finally, the implemented mechanism is embedded in palm, and the mechanical implementation to the human hand move characteristics is verified by the measured joint angles of the robotic hand when actuators separately move along PC1 and PC2. Meanwhile, the robotic hand can grasp different objects with a versatile grasp function.

This paper reports a new design topology of 4 pole hybrid electromagnetic levitation system in conjunction with linear synchronous motor. The mover, 4 pole hybrid electromagnet, is levitated underneath of the linear motor with zero power control algorithm. Propulsion and one-dimensional planar motion of the hybrid electromagnet is provided by the vector controlled linear motor. The new design of 4 pole hybrid electromagnet has several improvements over older systems, such as weight reduction and less material usage. Diagonal drive integration of mover and linear motor is proposed and experimental results are presented.

The ability to accurately predict human motion is imperative for any human-robot interaction application in which the human and robot interact in close proximity to one another. Although a variety of human motion prediction approaches have already been developed, they are often designed for specific types of tasks or motions, and thus do not generalize well. Furthermore, it is not always obvious which of these methods is appropriate for a given task, making human motion prediction difficult to implement in practice. We address this problem by introducing a multiple-predictor system (MPS) for human motion prediction. In our approach, the system learns directly from task data in order to determine the most favorable parameters for each implemented prediction method and which combination of these predictors to use. Our implementation consists of three complementary methods: velocity-based position projection, time series classification, and sequence prediction. We describe the process of forming the MPS and our evaluation of its performance against the individual methods in terms of accuracy of predictions of human position over a range of look-ahead time values. We report that our method leads to a reduction in mean error of 18.5%, 28.9%, and 37.3% when compared with the three individual methods, respectively.

This paper describes a novel class of robots specifically adapted to climb periodic lattices, which we call “Relative Robots”. These robots use the regularity of the structure to simplify the path planning, align with minimal feedback, and reduce the number of degrees of freedom (DOF) required to locomote. They can perform vital inspection and repair tasks within the structure that larger truss construction robots could not perform without modifying the structure. We detail a specific type of relative robot designed to traverse a cuboctahedral (CubOct) cellular solids lattice, show how the symmetries of the lattice simplify the design, and test these design methodologies with a CubOct relative robot that traverses a 76.2 mm (3 in.) pitch lattice, MOJO (Multi-Objective JOurneying robot). We perform three locomotion tasks with MOJO: vertical climbing, horizontal climbing, and turning, and find that, due to changes in the orientation of the robot relative to the gravity vector, the success rate of vertical and horizontal climbing is significantly different.

The compliance of soft actuators makes manipulation safer and simplifies control. But their high flexibility also makes sensorization challenging. From the large space of possible deformations not all are equally important. We present a method for sensorization of soft actuators that, for a given application, finds an effective layout from a set of sensors. It starts from a redundant sensor layout and iteratively reduces the number of sensors. Applying the method to the PneuFlex actuators of the RBO Hand 2, we identify a layout of four liquid metal strain sensors and one pressure sensor to predict actuator deformation in three dimensions: flexional, lateral, and twist. Finally, the layout is used to build a sensorized RBO Hand 2. It can detect passive shape adaptation while grasping and reveals failure cases during manipulation, e.g. slipping fingers while opening a door.

This paper presents preliminary results of the design, development, and evaluation of a hand rehabilitation glove fabricated using lobster-inspired hybrid design with rigid and soft components for actuation. Inspired by the bending abdomen of lobsters, hybrid actuators are built with serially jointed rigid shells actuated by pressurized soft chambers inside to generate bending motions. Such bio-inspiration absorbs features from the classical rigid-bodied robotics with preciselydefined motion generation, as well as the emerging soft robotics with light-weight, physically safe, and adaptive actuation. The fabrication procedure is described, followed by experiments to mechanically characterize these actuators. Finally, an open-palm glove design integrated with these hybrid actuators are presented for a qualitative case study. A hand rehabilitation system is developed by learning patterns of the sEMG signals from the users forearm to train the assistive glove for hand rehabilitation exercises.

Shared control is a key technology for various robotic applications in which a robotic system and a human operator are meant to collaborate efficiently. In order to achieve efficient task execution in shared control, it is essential to predict the desired behavior for a given situation or context in order to simplify the control task for the human operator. This prediction is obtained by exploiting Learning from Demonstration (LfD), which is a popular approach for transferring human skills to robots. We encode the demonstrated behavior as trajectory distributions and generalize the learned distributions to new situations. The goal of this paper is to present a shared control framework that uses learned expert distributions to gain more autonomy. Our approach controls the balance between the controller's autonomy and the human preference based on the distributions of the demonstrated trajectories. Moreover, the learned distributions are autonomously refined from collaborative task executions, resulting in a master-slave system with increasing autonomy that requires less user input with an increasing number of task executions. We experimentally validated that our shared control approach enables efficient task executions. Moreover, the conducted experiments demonstrated that the developed system improves its performances through interactive task executions with our shared control.

Over the last decade, automated vehicles have been widely researched and their massive potential has been verified through several milestone demonstrations. However, there are still many challenges ahead. One of the biggest challenges is integrating them into urban environments in which dilemmas occur frequently. Conventional automated driving strategies make automated vehicles foolish in dilemmas such as making lane-change in heavy traffic, handling a yellow traffic light and crossing a double-yellow line to pass an illegally parked car. In this paper, we introduce a novel automated driving strategy that allows automated vehicles to tackle these dilemmas. The key insight behind our automated driving strategy is that expert drivers understand human interactions on the road and comply with mutually-accepted rules, which are learned from countless experiences. In order to teach the driving strategy of expert drivers to automated vehicles, we propose a general learning framework based on maximum entropy inverse reinforcement learning and Gaussian process. Experiments are conducted on a 5.2 km-long campus road at Seoul National University and demonstrate that our framework performs comparably to expert drivers in planning trajectories to handle various dilemmas.

Increasing demands in endovascular intervention have motivated technical skill training and competency-based measures of performance. However, there are no well-established online metrics for technical skill assessment; few studies have explored operator behavioral patterns from catheter motion and operator hand motions. This paper proposes a platform for active online training and objective assessment of endovascular skills, through learning optimum catheter motions from multiple demonstrations. An ungrounded hand-held haptic device for providing intuitive haptic guidance to novice users based on this learnt information is also proposed. Statistical models are implemented to extract the underlying catheter motion patterns, and utilize them for performance evaluation and haptic guidance. The results show significant improvements in endovascular navigation for inexperienced operators. Finer catheter motions were achieved with the provided haptic guidance. The results suggest that the proposed platform can be integrated into current clinical training setups, and motivate the improvement of endovascular training platforms with better realism.

Many autonomous systems require the ability to perceive and understand motion in a dynamic environment. We present a novel algorithm that estimates this motion from raw LIDAR data in real-time without the need for segmentation or model-based tracking. The sensor data is first used to construct an occupancy grid. The foreground is then extracted via a learned background filter. Using the filtered occupancy grid, raw scene flow between successive scans is computed. Finally, we incorporate these measurements in a filtering framework to estimate temporal scene flow. We evaluate our method on the KITTI dataset.

We focus on a mobile robot that must learn another robot's motion model from observations to track it in a given map. This problem has several real-world applications such as self-driving cars being electronically towed by other cars and for telepresence robots. Our context is a nested particle filter, a generalization of the traditional particle filter, that allows both self-localization and tracking of another robot simultaneously. While the robot's observations are used to weight nested particles, the problem arises during the propagation step of the nested particles during which a motion model is needed. We introduce a novel layered hidden Markov model for this problem and present an on-line algorithm which learns the HMM parameters from observations gathered during the run. We demonstrate significantly improved tracking accuracy on using this new model to predict the motion of a leading mobile robot, in comparison to pre-defined and random motion models as previously used in literature.

Rigid actuators that accompany traditional robotic systems allow them to perform tasks with unmatched precision and efficiency. However, such systems are inapt when flexibility and suppleness are required. Inversely, soft robotic actuators are constructed out of naturally compliant materials. This compliance allows them to execute complex motions while their inherent lack of hardness removes the dangers that commonly accompany traditional robots. One archetype of the soft robot is the pneumatic bending actuator. The actuators convert the flow of pressurized air into bending outputs; however, most bending actuators are only limited to one bending trajectory. A new bending actuator, capable of active reconfiguration, is realized through the combination of a non-inflating, flexible, plastic spine and multiple, inflatable, fabric modules. The spine acts as the primary conduit and is responsible for distributing the pressurized flow to the modules. Modules can be removed and replaced seamlessly. The combination of varying module geometries can produce a heterogenic bending trajectory. Blockages located along the spine can separate flow, allowing more customization of the trajectory. A physical model predicting the pressure-bending relationship was conceptualized and verified. When pressurized to 100kPa, the 5-module and 10-module actuators achieved bending angles of 48.76° and 137.65° respectively. The actuators behave as a first-order system when supplied with a step input.

The robotic grasp detection is a great challenge in the area of robotics. Previous work mainly employs the visual approaches to solve this problem. In this paper, a hybrid deep architecture combining the visual and tactile sensing for robotic grasp detection is proposed. We have demonstrated that the visual sensing and tactile sensing are complementary to each other and important for the robotic grasping. A new THU grasp dataset has also been collected which contains the visual, tactile and grasp configuration information. The experiments conducted on a public grasp dataset and our collected dataset show that the performance of the proposed model is superior to state of the art methods. The results also indicate that the tactile data could help to enable the network to learn better visual features for the robotic grasp detection task.

As technology develops, manufacture process becomes more and more automated using robots. There is demand for high performance tactile sensor which can support robotic grippers in manipulation tasks especially for unstructured flexible objects. Despite the efforts that have been spent, the fabrication process of those functional sensor remains complicated due to their requirement of specialized materials and equipment. The proposed multimodal sensor overcomes the difficulty by enhancing the electrical and mechanical design therefore simplifying the manufacture steps. In this version, static and dynamic sensing are integrated in the same layer of capacitive sensor with direct written microstructured dielectric. This structure allows it to have large range of force sensing as well as the ability of detecting contact events such as slippage or losing of contact.

In this paper, we present a precise robot-aided single-cell surgery system to perform single-cell biopsy for cells <25 μm in diameter. A microfluidic chip is designed to arrange upto 100 individual cells in an array. A micropipette mounted onto a 3-DOF micromanipulator and a computer mouse-operated high-precision XY stage is developed to perform high-precision and high-throughput single-cell biopsy. The system is evaluated experimentally by extracting two organelles from adherent cells patterned in a microfluidic chip. The fluorescent-labeled nucleus and mitochondria of human foreskin fibroblast cells are biopsied to demonstrate the capability of the proposed system. The survival rate of the semi-automated biopsy is 73% and 45% for mitochondrial and nucleus biopsies, respectively.