ICRA 2015 Papers

This paper investigates the analytical solution of target steady walking speed in 1-DOF limit cycle walking. We introduce an active combined rimless wheel (CRW) model to analyze the target steady walking state when the CRW walks on level ground. The walking speed is determined by the step period because the step length is constant. First we propose a two-period stepwise control system and the target walking period can be generated by solving the equations of boundary conditions. Second we extend this method to (n + 1)-period stepwise control system and generate a general formula for the boundary equation. At last we generate the target steady step period in the continuous control systems by calculating the approximate solution based on discretization of control input. We verify all the results through numerical simulations. If the generated walking gait is single-step-cycle, we can generate the target steady step period by our general formula in most of control systems.

This paper introduces an approximate analytical solution to the otherwise non-integrable double-stance dynamics of the bipedal spring-loaded inverted pendulum (SLIP). Despite the apparent structural simplicity of the SLIP, the exact analytical solution to its stance dynamics cannot be found. Approximate maps have been proposed for the monoped SLIP runner (encompassing a single-stance phase). Still, even in an approximate form, a solution to the double-stance dynamics of the bipedal SLIP walker remained an open problem. We propose a double-stance map that can be readily utilized especially in the design of control systems for active dynamic walking. The accuracy of the derived map over a feasible range of locomotion properties is analyzed numerically, and a control application based on this solution is presented. Simulations for an arbitrary chosen energy level reveals that the devised controller enlarges the stable walking domain of the standard SLIP considerably.

Snakes and worms are capable of travelling along a straight line on a resistance medium using rectilinear locomotion. To apply this kind of gait in snake-inspired robots, some fundamental properties remain elusive. In this paper, we study the directional migration of a three-segment self-propelled system on a rough horizontal surface with Coulomb's dry friction. We elucidate the possibility that the system achieves directional migration through the propagation of extension and contraction wave. The necessary and sufficient conditions are certified to insure that the system is able to start from the state of rest. To analyze the approximately steady-state motion of the system, we provide explicit formulas for the mean velocity of the whole system by method of averaging. It is confirmed that for isotropic friction, the average speed and direction of the motion can be excited controlled.

It is important to understand the mechanism of human standing-up motion to improve the declined physical ability of the elderly people. This study employs the concept of muscle synergies (modular structure of coordinative muscle activation) to understand how humans coordinate their muscles to achieve the standing-up motion. Neuro-musculoskeletal model was developed to represent human body to generate standing-up motion. Using the developed model, forward dynamic simulation was used to analyze how humans utilized the muscle synergies to realize the motion. Results showed that the developed model could generate the standing-up motion with four muscle synergies rather than controlling individual muscles. Moreover, further analysis showed that three different strategies of the standing-up motion could be generated only by changing the start time of the particular muscle synergy.

A miniature robotic device that can fold-up on the spot, accomplish tasks, and disappear by degradation into the environment promises a range of medical applications but has so far been a challenge in engineering. This work presents a sheet that can self-fold into a functional 3D robot, actuate immediately for untethered walking and swimming, and subsequently dissolve in liquid. The developed sheet weighs 0.31 g, spans 1.7 cm square in size, features a cubic neodymium magnet, and can be thermally activated to self-fold. Since the robot has asymmetric body balance along the sagittal axis, the robot can walk at a speed of 3.8 body-length/s being remotely controlled by an alternating external magnetic field. We further show that the robot is capable of conducting basic tasks and behaviors, including swimming, delivering/carrying blocks, climbing a slope, and digging. The developed models include an acetone-degradable version, which allows the entire robot's body to vanish in a liquid. We thus experimentally demonstrate the complete life cycle of our robot: self-folding, actuation, and degrading.

This paper presents the design concept and fabricated prototype of a device that swings the arms for use in gait rehabilitation. The device is designed to be used in conjunction with a body-weight-support treadmill. The device is backdrivable, wearable, capable of assisting the user's arm swing in the sagittal plane, and has unhindered kinematics in the remaining unactuated degrees of freedom. Tests are performed to validate the shoulder-angle prediction equations based on the non-collocated motor-angle sensor measurements, to validate the device's ability to provide adequate torque to induce arm-swing in a passive user, and to investigate whether or not the user's active involvement can be determined by examining sensor data. The results show that the device does provide sufficient torque to move the arms with a factor of safety, but that the model-based shoulder-angle estimates obtained from the motor measurements have non-negligible error with the current prototype. It is shown that the controlled device generates low RMS tracking error and is able to diagnose user-assistance level (i.e., if the user is passive or actively assisting arm swing) online by observing shoulder-angle amplitudes and peak motor torques.

Various robotic applications including surgical instruments, wearable robots and autonomous mobile robots are often constrained with strict design requirements on high degrees of freedom (DoF) and minimal volume and weight. An intuitive design to meet these contradictory requirements is to embed locking mechanism in under actuated robotic manipulators to direct the actuation from a single and remote source to drive different joints on demand. Mechanical clutches do serve such purposes but often are bulky and require auxiliary mechanism making it difficult to justify the high cost adding the additional DoF, especially in cm scale. Here, we introduce an under-actuated robotic arm with shape memory polymer (SMP) joints. Through controlling the temperature, the stiffness of the joints can be adjusted and selected joints will be activated while the rest are fixed in their position. The presented prototype can control the joints independently with a coupled actuation from two stepper motors. Since we have redundant DoFs in the arm, there can be more than one configuration to reach a given position. We use a probabilistic technique to determine the optimum configuration with the minimum number of active joints that can yield the desired posture. In this paper, we report on the performance of the proposed design for the hardware and the configuration planner.

Reliable precision grasping is a pre-condition for manipulation tasks e.g. in assembly and packaging tasks. Especially for small and light objects robust grasping is extremely challenging since even slight errors in the object pose or dimensions lead to irreparable failures caused by unintended finger-object contacts. State of the art techniques address the problem of grasping in the presence of uncertainty only at the grasp planning stage. We regard grasping as a dynamic process that needs closed loop control to be robustly executed, and propose an approach to robustly perform precision grasps despite of the significant perceptual and actuation uncertainties we have to face in reality. We conduct extensive real world experiments with light and fragile objects of unknown dimensions. The result confirms that our uncertainty-aware closed-loop approach significantly improves the robustness.

This paper presents a path planning concept for trucks with trailers with kingpin hitching. This system is nonholonomic, has no flat output and is not stable in backwards driving direction. These properties are major challenges for path planning. The presented approach concentrates on the loading bay scenario. The considered task is to plan a path for the truck-trailer system from a start to a specified target configuration corresponding to the loading bay. Thereby, close distances to obstacles and multiple driving direction changes have to be handled. Furthermore, a so-called jackknife position has to be avoided. In a first step, an initial path is planned from the target to the start configuration using a tree-based path planner. Afterwards this path is refined locally by solving an optimal control problem. Due to the local nature of the planner, heuristic rules for direction changes are formulated. The performance of the proposed path planner is evaluated in simulation studies.

This paper presents a fast tight-tolerance threading technique for string and rope. Instead of relying on simulations of these deformable objects to plan a path or compute control actions, we control the movement of the string with a virtual magnetic vector field emanating from the narrow openings we wish to thread through. We compute an approximate Jacobian to move the tip of the string through the vector field and propose a method to promote alignment of the head of the string to the opening. We also propose a method for re-grasping the string based on the relationship between the string's configuration, the orientation of the opening, and direction of gravity. This re-grasping method in conjunction with our controller can be used to thread the string through a sequence of openings. We evaluated our method in simulation (with simulated sensor noise) and on the Da Vinci surgical robot. Our results suggest that our method is quite robust to errors in sensing, and is capable of real-world threading tasks with the da Vinci robot, where the diameter of the string (3.5mm) and opening (4.9mm) differ by only 1.4 mm.

For agile, accurate autonomous robotics, it is desirable to plan motion in the presence of uncertainty. The Partially Observable Markov Decision Process (POMDP) provides a principled framework for this. Despite the tremendous advances of POMDP-based planning, most can only solve problems with a small and discrete set of actions. This paper presents General Pattern Search in Adaptive Belief Tree (GPS-ABT), an approximate and online POMDP solver for problems with continuous action spaces. Generalized Pattern Search (GPS) is used as a search strategy for action selection. Under certain conditions, GPS-ABT converges to the optimal solution in probability. Results on a box pushing and an extended Tag benchmark problem are promising.

In this paper, we introduce an Iterative Kalman Smoother (IKS) for tracking the 3D motion of a mobile device in real-time using visual and inertial measurements. In contrast to existing Extended Kalman Filter (EKF)-based approaches, smoothing can better approximate the underlying nonlinear system and measurement models by re-linearizing them. Additionally, by iteratively optimizing over all measurements available, the IKS increases the convergence rate of critical parameters (e.g., IMU-camera clock drift) and improves the positioning accuracy during challenging conditions (e.g., scarcity of visual features). Furthermore, and in contrast to existing inverse filters, the proposed IKS's numerical stability allows for efficient 32-bit implementations on resource-constrained devices, such as cell phones and wearables. We validate the IKS for performing vision-aided inertial navigation on Google Glass, a wearable device with limited sensing and processing, and demonstrate positioning accuracy comparable to that achieved on cell phones. To the best of our knowledge, this work presents the first proof-of-concept real-time 3D indoor localization system on a commercial-grade wearable computer.

In this paper, we consider the problem of tracking a reference trajectory for a simplified car model based on unicycle kinematics, whose position only is measured, and where the control input and the measurements are corrupted by independent Gaussian noises. To tackle this problem we devise a novel observer-controller: the invariant Linear Quadratic Gaussian controller (ILQG). It is based on the Linear Quadratic Gaussian controller, but the equations are slightly modified to account for, and to exploit, the symmetries of the problem. The gain tuning exhibits a reduced dependency on the estimated trajectory, and is thus less sensitive to misestimates. Beyond the fact the invariant approach is sensible (there is no reason why the controller performance should depend on whether the reference trajectory is heading west or south), we show through simulations that the ILQG outperforms the conventional LQG controller in case of large noises or large initial uncertainties.

This paper presents a framework to embody a user (e.g. disabled persons) into a humanoid robot controlled by means of brain-computer interfaces (BCI). With our framework, the robot can interact with the environment, or assist its user. The low frequency and accuracy of the BCI commands is compensated by vision tools, such as objects recognition and mapping techniques, as well as shared-control approaches. As a result, the proposed framework offers intuitive, safe, and accurate robot navigation towards an object or a person. The generic aspect of the framework is demonstrated by two complex experiments, where the user controls the robot to serve him a drink, and to raise his own arm.

Dynamic Movement Primitives (DMPs) are a common method for learning a control policy for a task from demonstration. This control policy consists of differential equations that can create a smooth trajectory to a new goal point. However, DMPs only have a limited ability to generalize the demonstration to new environments and solve problems such as obstacle avoidance. Moreover, standard DMP learning does not cope with the noise inherent to human demonstrations. Here, we propose an approach for robot learning from demonstration that can generalize noisy task demonstrations to a new goal point and to an environment with obstacles. This strategy for robot learning from demonstration results in a control policy that incorporates different types of learning from demonstration, which correspond to different types of observational learning as outlined in developmental psychology.

Visual servoing based on photometric data is of great interest since it does not necessitate any image processing or visual tracking steps. A vital issue in such methods is the change in the image resulting from the appearance and disappearance of portions of the scene from the camera field-of-view during the visual servo. In this paper, we propose a spatial weighting scheme to counter this problem. A general model for photometric moments that allows for inclusion of spatial weights is presented. A custom weighting function is then chosen and specific properties of its derivative are exploited in order to develop the interaction matrix in analytical form. The resultant effects on the invariance properties are discussed. Experiments have been performed to validate the visual servoing with the proposed weighting scheme on a 4 dof robot arm.

This paper examines an instance of a collaborative perception problem across aerial and ground robots, specifically in the context of transferring learned appearance classifiers from one to the other. This problem is extremely challenging, as the two robots differ significantly in target resolution, pixels on target, and perspectives. We empirically explore a set of state of the art features used to distinguish the target and ask the question: What features are the most transferable between the aerial and ground robots for the purposes of target handoff, where a target must be distinguished from `confuser' objects? We show that while these features are successful on each individual robot, they are not able to be transferred naively. However, we show that using feature alignment techniques, where sparse features are mapped from one robot to the other using a shared context, we are able to successfully transfer classifiers that use color and texture features. We demonstrate these results on an extremely challenging outdoor data set simultaneously collected using an aerial vehicle and sensor-mounted ground vehicle viewing the same scene and target.

In this paper we present the results of our recent study on comparing the emotion expression recognition abilities of children diagnosed with high functioning Autism (ASD) with those of typically developing (TD) children through use of a humanoid robot, Zeno. In our study we investigated the effect of incorporating gestures to the emotion expression prediction accuracy of both child groups. Although the idea that ASD individuals suffer from general emotion recognition deficits is widely assumed [1], we found no significant impairment in the general emotion prediction. However, a specific deficit in correctly identifying Fear was found for children with Autism when compared to the TD children. Furthermore, we found that gestures can significantly impact the prediction accuracy of both ASD and TD children in a negative or positive manner depending on the specific expression. Thus, the use of gestures for conveying emotional expressions by a humanoid robot in a social skill therapy setting is relevant. The methodology and experimental protocol are presented and additional discussion of the Zeno R-50 robot used is given.

In this paper, we introduce an efficient technique that addresses the deployment problem of a team of heterogeneous robots. For multiple scattered tasks in arbitrary space, the objective to be solved is to minimize time for robots to reach all the task locations. In order to reach the assigned task locations efficiently, the team composed of one carrier robot and several servant robots with different motion capabilities is considered. The carrier robot is responsible for transporting and deployment of the servant robots. The servant robot, deployed by the carrier robot, moves to the assigned task location and executes the given task. Then, the overall time to reach all the task locations is determined by considering the path of the carrier that is made by deployment points, and by the maximum time traveled by the servant robots. To find efficient deployment points, the set of all the tasks are iteratively divided into two subsets. Then, an optimal deployment point corresponding to two subsets is calculated by considering the speed of the carrier robot and servant robot, the unloading time for deployment, and few dynamics of the carrier robot. The overall time is computed and compared to the previous result so that the time it takes is continuously reduced. The proposed method has been tested on simulations.