ICRA 2015 Papers

This paper is about opening and closing an unknown drawer using an aerial manipulator. To accommodate practical applications, it is assumed that the direction of motion and mechanical properties of the drawer are not given beforehand. A multirotor combined with a robotic arm is used for the manipulation task. Typical drawers are allowed to move in only one direction, which constrains the motion of the aerial manipulator while operating a drawer. To analyze this interaction, the dynamic characteristics of the aerial manipulator are modeled. Also, configuration of the aerial manipulator for exerting the desired force to a drawer is presented. To handle the uncertainties associated with the mechanism of a drawer, strategies exploiting velocity of the end effector are employed. The proposed approach is validated with experiments including opening and closing a common drawer, which is detected by a camera mounted in the palm of the end effector.

A fundamental aspect of controlling humanoid robots lies in the capability to exploit the whole body to perform tasks. This work introduces a novel whole body control library called OpenSoT. OpenSoT is combined with joint impedance control to create a framework that can effectively generate complex whole body motion behaviors for humanoids according to the needs of the interaction level of the tasks. OpenSoT gives an easy way to implement tasks, constraints, bounds and solvers by providing common interfaces. We present the mathematical foundation of the library and validate it on the compliant humanoid robot COMAN to execute multiple motion tasks under a number of constraints. The framework is able to solve hierarchies of tasks of arbitrary complexity in a robust and reliable way.

Making future autonomous robots capable of accomplishing human-scale manipulation tasks requires us to equip them with knowledge and reasoning mechanisms. We propose Open-EASE, a remote knowledge representation and processing service that aims at facilitating these capabilities. Open-EASE gives its users unprecedented access to the knowledge of leading-edge autonomous robotic agents. It also provides the representational infrastructure to make inhomogeneous experience data from robots and human manipulation episodes semantically accessible, and is complemented by a suite of software tools that enable researchers and robots to interpret, analyze, visualize, and learn from the experience data. Using Open-EASE users can retrieve the memorized experiences of manipulation episodes and ask queries regarding to what the robot saw, reasoned, and did as well as how the robot did it, why, and what effects it caused.

Truly autonomous robots require the capacity to recognise their surroundings by interpreting their sensorimotor stream. We present an online learning algorithm for training a mixture of echo state network experts that can segment a compliant robot's sensorimotor stream. Our method follows a probabilistic approach, using a hidden Markov model to model the switching dynamics between the experts. The algorithm's performance is evaluated on an unsupervised terrain classification problem using a compliant, underactuated, six-strut tensegrity robot. The results show that our model captures the influence of terrain-robot interactions on the robot's complex dynamics and correctly segments the sensorimotor stream. We demonstrate that the activity pattern of the experts can be used to train a highly compliant robot to distinguish between different environments using only noisy internal sensors.

This paper presents an online self-supervised approach to monocular visual odometry and ground classification applied to ground vehicles. We solve the motion and structure problem based on a constrained kinematic model. The true scale of the monocular scene is recovered by estimating the ground surface. We consider a general parametric ground surface model and use the Random Sample Consensus (RANSAC) algorithm for robust fitting of the parameters. The estimated ground surface provides training samples to learn a probabilistic appearance-based ground classifier in an online and self-supervised manner. The appearance-based classifier is then used to bias the RANSAC sampling to generate better hypotheses for parameter estimation of the ground surface model. Thus, without relying on any prior information, we combine geometric estimates with appearance-based classification to achieve an online self-learning scheme from monocular vision. Experimental results demonstrate that online learning improves the computational efficiency and accuracy compared to standard sampling in RANSAC. Evaluations on the KITTI benchmark dataset demonstrate the stability and accuracy of our overall methods in comparison to previous approaches.

We present a framework for planning collision-free paths online for autonomous underwater vehicles (AUVs) in unknown environments. It is composed of three main modules (mapping, planning and mission handler) that incrementally explore the environment while solving start-to-goal queries. We use an octree-based representation of the environment and we extend the optimal rapidly-exploring random tree (RRT*) using concepts of anytime algorithms and lazy collision evaluation, thus including the capability to replan paths according to nearby obstacles perceived during the execution of the mission. To validate our approach, we plan paths for the SPARUS-II AUV, a torpedo-shaped vehicle performing autonomous missions in a 2-dimensional workspace. We demonstrate its feasibility with the SPARUS-II AUV in both simulation and real-world in-water trials.

This paper presents a novel online unsupervised vision system for obstacle detection in field environments which detects many obstacles pathological to appearance- or structure-only obstacle detection systems. Robust obstacle detection in field environments is challenging as it is infeasible to train on all possible obstacles in all conditions, and many obstacles are camouflaged in their appearance or structure. The proposed system combines novelty in structure and appearance cues to detect obstacles, can adapt over time to changes in the environment, and is suitable for long-term operation over changing lighting conditions in various environments. After an initial learning period the method exhibits very few false positives, while successfully detecting most obstacles over both daytime and nighttime datasets including challenging obstacles such as a person lying down in grass.

Modular robots are built of many basic robotic modules that can be connected into robots of various shapes. These robots are able to recover from a failure by ejecting and replacing damaged modules. Although this type of failure recovery is usually suggested in literature, it may be precluded due inability of the broken modules to cooperate or if no spare modules are available. In such a case, locomotion of a damaged robot should be adapted to allow the robot to reach a repair station or even to finish its task without the need to exchange the broken modules. In this paper, we investigate how to adapt motions of modular robots with respect to failures using the concept of motion planning with motion primitives. The ability of the proposed system to recover from failures is verified in a simulation and also in a HW experiment.

Biped walking control based on simplified models relies much on online feedback stabilizers to compensate the zero-moment point (ZMP) error which partially comes from the model inconsistency of pattern generation. Inspired by the fact that human improves the performance by practicing a task for multiple times, this paper presents an online learning control framework for improving the robustness during the dominant repetitive phases of walking. The key idea is to learn a compensative feedforward ZMP term from previous ZMP error trajectories in order to achieve better ZMP tracking. Based on the iterative learning control theory, the learning process is conducted online continuously with minimal iteration of two footsteps, which can practically run in parallel with state-of-the-art walking controllers. A varying forgetting factor is designed to reduce the influence of the landing impact. Convergence of the learning control algorithm and improved ZMP tracking performance is verified both in dynamics simulation and experiment on the DLR humanoid robot TORO.

This paper discusses an online dynamic motion generation scheme for nonprehensile object manipulation by using a set of predefined motions and a trajectory deformation algorithm capable of incorporating positional and velocity boundary constraints. By creating optimal trajectories offline and deforming them online, computational complexity during execution is reduced considerably. As tight convex hulls of the deformed trajectories can be found, possible obstacles or workspace boundaries can be circumnavigated precisely without collision. The approach is verified through experiments on an inclined planar air-table for volleyball scenario using two 3-DoF robots.

A framework for online simultaneous localization, mapping and self-calibration is presented which can detect and handle significant change in the calibration parameters. Estimates are computed in constant-time by factoring the problem and focusing on segments of the trajectory that are most informative for the purposes of calibration. A novel technique is presented to detect the probability that a significant change is present in the calibration parameters. The system is then able to re-calibrate. Maximum likelihood trajectory and map estimates are computed using an asynchronous and adaptive optimization. The system requires no prior information and is able to initialize without any special motions or routines, or in the case where observability over calibration parameters is delayed. The system is experimentally validated to calibrate camera intrinsic parameters for a nonlinear camera model on a monocular dataset featuring a significant zoom event partway through, and achieves high accuracy despite unknown initial calibration parameters. Self-calibration and re-calibration parameters are shown to closely match estimates computed using a calibration target. The accuracy of the system is demonstrated with SLAM results that achieve sub-1% distance-travel error even in the presence of significant re-calibration events.

We introduce CHAMP, an algorithm for online Bayesian changepoint detection in settings where it is difficult or undesirable to integrate over the parameters of candidate models. CHAMP is used in combination with several articulation models to detect changes in articulated motion of objects in the world, allowing a robot to infer physically-grounded task information. We focus on three settings where a changepoint model is appropriate: objects with intrinsic articulation relationships that can change over time, object-object contact that results in quasi-static articulated motion, and assembly tasks where each step changes articulation relationships. We experimentally demonstrate that this system can be used to infer various types of information from demonstration data including causal manipulation models, human-robot grasp correspondences, and skill verification tests.

This paper investigates the calibration of a microphone array based robot audition system, namely calibration of microphone array Transfer Functions (TFs). There are mainly two methods to obtain TFs: geometrical calculation and measurement. The geometrical calculation has difficulty in simulating robot- and room-acoustics such as diffraction and reflection properties of a robot's body and a room, and the measurement is accurate but time-consuming and requires expertise on acoustics. Thus, we propose fast and simple on-the-spot calibration of TFs including robot- and room-acoustics. The proposed approach first estimates microphone location and clock-difference by Simultaneous Localization And Mapping (SLAM) using hand clap acoustic signals while a human is walking around a robot. Second, TFs with robot- and room-acoustics are estimated by hand clap acoustic signals and interpolated so that the TFs can be roundly arranged at regular intervals in an online manner. In the evaluation, we calibrated TFs only by 20 hand claps (took only 20 seconds), and the TFs showed considerable improvements in sound source localization and separation compared to geometrically calculated TFs and achieved comparable performance towards the measured TFs which are calibrated by approximately 60 minute recordings.

In tele-robotics, time delay is a significant problem. When video feedback is delayed, operators adopt inefficient move-wait strategies, so system performance decreases. Predictive display (PD) is an effective solution to compensate for delays by graphics rendering of predicted visual feedback. Using advanced computer vision technology, we implemented a PD system based-on online real-time 3D reconstruction from monocular video. This paper describes the client-server system architecture. Experimental results indicate it can capture 3D models and render the predicted image in realistic applications covering outdoor rover operation on earth, Canadian Space Agency's (CSA) Mars analogue environment, UAV operation.

This paper proposes the method for mechanical characterization of floating cells on a microfluidic chip. We measured the reactive force of cells on the chip by applying mechanical deformation to the cells. Particularly we focused on the method to improve the sensing resolution of the probe position by using the phase detection of moiré fringe. This method realized approximately ten times higher resolution compared to the previous method. The detailed data suggest some unique characteristics of cells. We discussed the phenomenon seen on the data with a mechanical model, which is based on Heltzian contact theory in the light of stiff nucleus.

In this paper, we address the pursuit/evasion problem of capturing an omnidirectional evader using a Differential Drive Robot (DDR) in an obstacle-free environment. The goal of the evader is to keep the pursuer farther than the capture distance for as long as possible and for the pursuer the goal is to capture the evader as soon as possible. In [1] an open-loop time-optimal strategy is proposed for this pursuit/evasion problem. In [2] a state feedback-based time-optimal motion policy for the DDR is provided. The time-optimal strategies obtained in [1] are in Nash equilibrium, meaning that any unilateral deviation of a player from the optimal strategies does not provide it a benefit in its payoff. However, Nash equilibrium does not tell if one player deviates from its optimal policy then, does there exist a new strategy for the other player that can take advantage of such deviation? If so, which is the required information to improve the payoff compared with the worst case scenario? In this paper we address those questions, analysing the scenario in which the players deviate from their optimal controls. We show that when the evader deviates from its optimal speed there are cases where there exists a new pursuer motion strategy that reduces the time to capture the evader. The shown cases where the time to capture the evader is reduced require more information about the evader's state. Nevertheless, there are also cases in which despite the availability of new information, the pursuer must stick to the worst case strategy, otherwise it cannot capture the evader.

The stiffness ellipsoid, i.e. the locus of task-space forces obtained corresponding to a deformation of unit norm in different directions, has been extensively used as a powerful representation of robot interaction capabilities. The size and shape of the stiffness ellipsoid at a given end-effector posture are influenced by both joint control parameters and - for redundant manipulators - by the chosen redundancy resolution configuration. As is well known, impedance control techniques ideally provide control parameters which realize any desired shape of the Cartesian stiffness ellipsoid at the end-effector in an arbitrary non-singular configuration, so that arm geometry selection could appear secondary. This definitely contrasts with observations on how humans control their arm stiffness, who in fact appear to predominantly use arm configurations to shape the stiffness ellipsoid. To understand this discrepancy, we provide a more complete analysis of the task-space force/deformation behavior of redundant arms, which explains why arm geometry also plays a fundamental role in interaction capabilities of a torque controlled robot. We show that stiffness control of realistic robot models with bounds on joint torques can't indeed achieve arbitrary stiffness ellipsoids at any given arm configuration. We first introduce the notion of maximum allowable Cartesian force/displacement (“stiffness feasibility”) regions for a compliant robot. We show that different robot configurations modify such regions, and explore the role of different configurations in defining the performance limits of Cartesian stiffness controllers. On these bases, we design a stiffness control method that suitably exploits both joint control parameters and redundancy resolution to achieve desired task-space interaction behavior.

As the expansion of the field of robotics has continued, the physical interaction between robots and humans has become an increasingly important area of study. Many of these physical interactions can be seen as a cooperative task conducted by both the robot and the human. Often, when two humans are interacting, one of them will act as the leader of some aspect of the task and the other will act as a follower. This cooperation may require the switching of roles between leader and follower. This can be further complicated by the fact that different participants may be the leaders of different aspects of the task. Previous research in human-robot cooperation focused on the switching of only a single role. In this paper, we investigate a novel method for the simultaneous switching of two roles between a robot and a human participant. This switching method was examined using both fixed and adaptive parameters that control role switching. Overall, human-robot cooperation was successful in the task 85% of the time when using a non-adaptive method and 95% when using an adaptive control method.

In this paper we investigate the monotonicity of various optimality criteria during the exploration phase of an active SLAM algorithm. Optimality criteria such as A-opt, D-opt or E-opt are used in active SLAM to account for uncertainty in the map or the robot's pose, and these criteria are usually part of utility functions which help active SLAM algorithms decide where the robot should move next. The monotonicity of the optimality criteria is of utmost importance. During the exploration phase, i.e. when the robot is traversing new territory or cannot perform a loop closure, the most common way of estimating the pose of the robot is through dead-reckoning. Correctly accounting for the uncertainty is important for an active SLAM algorithm and in particular for a dead-reckoning scenario, where by definition the uncertainty in the robot's pose grows. If monotonicity does not hold in this scenario, active SLAM algorithms can execute actions under the false belief that the uncertainty has reduced. We show analytically and experimentally some conditions in which the A-opt and E-opt criteria lose monotonicity in a dead-reckoning scenario, where the propagation of the robot's pose is done using a linearized framework. We also show analytically and experimentally that under the same conditions the D-opt does not lose monotonicity and, in general for the linearized framework under consideration, D-opt does not break monotonicity.