ICRA 2015 Papers

Safe, intuitive human-robot interaction requires that robots intelligently interface with their environments, ideally sensing and localizing physical contact across their link surfaces. We introduce a stretchable smart skin sensor that provides this function. Stretchability allows it to conform to arbitrary robotic link surfaces. It senses contact over nearly the entire surface, localizes contact position of a typical finger touch continuously over its entire surface (RMSE = 7.02mm for a 14.7cm×14.7cm area), and provides an estimate of the contact force. Our approach exclusively employs stretchable, flexible materials resulting in skin strains of up to 150%. We exploit novel carbon nanotube elastomers to create a two-dimensional potentiometer surface. Finite element simulations validate a simplified polynomial surface model to enable real-time processing on a basic microcontroller with no supporting electronics. Using only five electrodes, the skin can be scaled up to arbitrary sizes without needing additional electrodes. We designed, implemented, calibrated, and tested a prototype smart skin as a tactile sensor on a custom medical robot for sensing unexpected physical interactions. We experimentally demonstrate its utility in collaborative robotic applications by showing its potential to enable safer, more intuitive human-robot interaction.

Series elastic actuator (SEA) has been widely used in rehabilitation robotics, where human-robot interaction is required. Due to its intrinsic compliance, SEA can improve the usage of power for its motor, which leads to a compact and lightweight SEA design. The aim of this paper is to reduce the energy consumption and the power requirements of the motor of the SEA by optimizing the stiffness of its spring. This study is inspired by the biomechanics of a human ankle joint, which stores elastic energy during the first phases of the walking process and releases the stored energy in the next gait phases to propel the human body forward. Power analysis and optimization procedure are conducted on complete SEA models with different complexity, including inertia, damping and stiffness, and with open loop and closed loop control strategies. Simulation results demonstrate that a reduction of 56.6% of the peak motor power can be achieved with the optimized spring stiffness.

This paper focuses on detecting and classifying pole-like objects from point clouds obtained in urban areas. To achieve our goal, we propose a system consisting of three stages: localization, segmentation and classification. The localization algorithm based on slicing, clustering, pole seed generation and bucket augmentation takes advantage of the unique characteristics of pole-like objects and avoids heavy computation on the feature of every point in traditional methods. Then, the bucket-shaped neighborhood of the segments is integrated and trimmed with region growing algorithms, reducing the noises within candidate's neighborhood. Finally, we introduce a representation of six attributes based on the height and five point classes closely related to the pole categories and apply SVM to classify the candidate objects into 4 categories, including 3 pole categories light, utility pole and sign, and the non-pole category. The performance of our method is demonstrated through comparison with previous works on a large-scale urban dataset.

This paper presents the design, implementation, and validation of balance control and point-to-point motion control of a single-wheeled (unicycle) robot. The robot consists of a wheel, a body, a reaction wheel, and a turntable. The wheel is used to move the robot forward and backward to obtain longitudinal stability. The reaction wheel is used for obtaining lateral stability. The turntable provides steering torque about the vertical axis of the robot. The dynamic model of the system is derived using Euler-Lagrange formulation. By retaining the predominant nonlinear terms and neglecting the high-order coupling terms, the system model is simplified to three decoupled systems. Sliding mode control is then used to design the balance and steering controllers for the simplified model. By combining the obtained balance control and steering control, a point-to-point motion control strategy is proposed. The experimental results are presented to verify the effectiveness of the designed control schemes.

Development of intuitive user interface of a dexterous forceps with wrist joint for minimally-invasive surgery (MIS) is important for reducing the surgeon's confusion and fatigue. In this paper, a wearable robotic forceps manipulator is proposed. By measuring the wrist attitude of the surgeon by a motion sensor and driving the forceps tip by actuators, intuitive operation is realized, compared with fully mechanical forceps. A compact and lightweight system is achieved since we adopt pneumatic actuators to drive the manipulator. In order to help the force sensing of the surgeon, the external force is measured from the air pressure and displayed in a visual way. Experimental comparison with a conventional handheld forceps and a teleoperation system is conducted and its results show the effectiveness of the proposed handheld forceps.

In this work, we examine two design modifications to a tethered motor-driven flapping-wing system. Previously, we had demonstrated a simple mechanism utilizing a linear transmission for resonant operation and direct drive of the wing flapping angle for control. The initial two-wing system had a weight of 2.7 grams and a maximum lift-to-weight ratio of 1.4. While capable of vertical takeoff, in open-loop flight it demonstrated instability and pitch oscillations at the wing flapping frequency, leading to flight times of only a few wing strokes. Here the effect of vertical wing offset as well as an alternative multi-wing layout is investigated and experimentally tested with newly constructed prototypes. With only a change in vertical wing offset, stable open-loop flight of the two-wing flapping system is shown to be theoretically possible, but difficult to achieve with our current design and operating parameters. Both of the new two and four-wing systems, however, prove capable of flying to the end of the tether, with the four-wing system prototype eliminating disruptive wing beat oscillations.

This paper investigates the free motion of a sphere with initial velocity and angular velocity on a plane under sliding and rolling friction. The sphere will first slide along a parabolic trajectory (with a constant direction of its contact velocity), and then roll along a straight trajectory. Such a curved trajectory can be utilized for obstacle avoidance in path planning. A one-to-one correspondence exists between trajectories connecting two specified locations and pairs of sliding and rolling directions within some 2-dimensional region, called the trajectory space, which is then used for planning. A plane sweep algorithm operating in this space is presented to find all collision-free trajectories in the presence of cylindrical obstacles that are vertically oriented.

This paper presents a method that applies connected component analysis to plan routes that keep robots continuously illuminated and on traversable slopes while reaching one or more goal locations. Such routes promise to extend the lifespan, range, and scientific return of solar-powered robots exploring environments with changing but predictable lighting conditions, particularly those of the Moon and Mercury. Maps of lighting and ground slope that describe these constraints in position and time are computed, and all distinct interconnected regions that have both direct sunlight and safe slope are found using connected component analysis. These three-dimensional connected components are pruned of roots that violate time constraints and branches that dead-end in discontinuous routes. Each component is the basis for a graph that includes all feasible routes from the initial time to the final time of that component. The shortest feasible route between a pair of start and goal positions within the same component is found using A* search and is characterized by its total length and average speed. Malapert Peak and Shackleton Crater, both near the Moon's South Pole, serve as examples throughout this paper due to their highly-relevant, dynamic, and predictable lighting caused by the Moon's motion relative to the Sun.

Object search is a fundamental ability for a service robot to provide higher level services. We focus on object search in an environment with limited free space to place objects and constrained viewpoints to observe the environment, such as a shelf or a cupboard. We propose an object search planner based on A* search algorithm with tree node sampling. The proposed approach also combines visual and arm manipulation search. In other words, the robot searches occluded target object by either repositioning one of the accessible object with its arm or moving its platform to view the environment from a different pose. We evaluate the proposed approach with experiment performed by real robot in the scenario which objects may occlude or block access to one another.

There is strong evidence that robustness in human and robotic grasping can be achieved through the deliberate exploitation of contact with the environment. In contrast to this, traditional grasp planners generally disregard the opportunity to interact with the environment during grasping. In this paper, we propose a novel view of grasp planning that centers on the exploitation of environmental contact. In this view, grasps are sequences of constraint exploitations, i.e. consecutive motions constrained by features in the environment, ending in a grasp. To be able to generate such grasp plans, it becomes necessary to consider planning, perception, and control as tightly integrated components. As a result, each of these components can be simplified while still yielding reliable grasping performance. We propose a first implementation of a grasp planner based on this view and demonstrate in real-world experiments the robustness and versatility of the resulting grasp plans.

Follow-the-leader based approaches have been popular for the control of multi-robot teams for their ability to drive many with few. Typically, in these methods you select a single leader, generate a plan for it, while all other agents follow this leader using their individual controllers. However, there are many scenarios where this approach can lead to highly suboptimal behavior or even failure in the presence of clutter. In this work, we present a planning approach that automatically figures out when to switch leaders on the way to the goal while minimizing a given cost function that penalizes leader switching and deviations from the desired formation. To deal with the increased dimensionality of the problem we show how a recently developed algorithm, MHA* (multi-heuristic A*) can be extended to support planning for a team of robots. We also provide explicit cost minimization and guarantee that paths found are within a user- chosen factor of optimality with respect to the graph modeling the planning problem. Experimentally, we found that allowing for dynamic leader-switching leads to a significant increase in finding feasible plans for multi-robot teams ranging up to 21 robots.

This paper presents a probabilistic framework for synthesizing control policies for general multi-robot systems that is based on decentralized partially observable Markov decision processes (Dec-POMDPs). Dec-POMDPs are a general model of decision-making where a team of agents must cooperate to optimize a shared objective in the presence of uncertainty. Dec-POMDPs also consider communication limitations, so execution is decentralized. While Dec-POMDPs are typically intractable to solve for real-world problems, recent research on the use of macro-actions in Dec-POMDPs has significantly increased the size of problem that can be practically solved. We show that, in contrast to most existing methods that are specialized to a particular problem class, our approach can synthesize control policies that exploit any opportunities for coordination that are present in the problem, while balancing uncertainty, sensor information, and information about other agents. We use three variants of a warehouse task to show that a single planner of this type can generate cooperative behavior using task allocation, direct communication, and signaling, as appropriate. This demonstrates that our algorithmic framework can automatically optimize control and communication policies for complex multi-robot systems.

This paper describes a strategy to lift up a deformable 2D object resting on a horizontal plane. Inspired by human hand lifting behavior, the strategy plans grasping trajectories of two fingertips using modified Rapidly-exploring Random Trees (RRT). Compared to a straight squeeze, a planned finger movement not only enlarges the graspable region, but also minimizes the work. Improvements on control complexity and robustness are made through modifications to RRT. The introduced strategy is applicable to both hollow and solid 2D objects, and is extendable to 3D ones.

We present a framework for dynamic quadrupedal locomotion over challenging terrain, where the choice of appropriate footholds is crucial for the success of the behaviour. We build a model of the environment on-line and on-board using an efficient occupancy grid representation. We use Any-time-Repairing A* (ARA*) to search over a tree of possible actions, choose a rough body path and select the locally-best footholds accordingly. We run a n-step lookahead optimization of the body trajectory using a dynamic stability metric, the Zero Moment Point (ZMP), that generates natural dynamic whole-body motions. A combination of floating-base inverse dynamics and virtual model control accurately executes the desired motions on an actively compliant system. Experimental trials show that this framework allows us to traverse terrains at nearly 6 times the speed of our previous work, evaluated over the same set of trials.

This paper presents a trajectory generator for humanoid robots that can simultaneously consider 3D position tracking of the center-of-mass (COM), reactive stepping (foot positions are non-constant), zero-moment-point (ZMP) constraints as well as position constraints. The variable COM height leads to nonlinear ZMP constraints. This work demonstrates that depending on how strict the constraints are, it is possible to generate 1.6 second long trajectories in less than 5 ms on a 3.4 Ghz CPU. This is done by expressing the problem as a quadratically constrained quadratic program (QCQP) and solving it via sequential quadratic programming (SQP). Computation time is improved by providing the analytical derivatives of all constraints. In particular the nonlinear ZMP constraint is expressed in a quadratic form and its derivative is provided.

In this paper we experimentally validate and compare three different methods for estimating the 3D parameters of a planar scene from a (possibly time-varying) set of feature points acquired by a moving monocular camera. The first method, based on the classical decomposition of the homography matrix, is meant to serve as a baseline condition classically used in many previous works. The other two methods exploit an active Structure from Motion (SfM) scheme for either extracting the plane from the reconstructed 3D position of all the tracked points, or for directly estimating the plane parameters by considering a set of discrete image moments as visual input. The possible loss/gain of point features during the camera motion is considered in all three methods by, in particular, introducing a suitable weighting strategy for the image moment case. Finally, the results of an experimental validation are presented with a comparative discussion of the pros/cons of the three methods.

We consider the problem of learning user preferences over robot trajectories for environments rich in objects and humans. This is challenging because the criterion defining a good trajectory varies with users, tasks and interactions in the environment. We represent trajectory preferences using a cost function that the robot learns and uses it to generate good trajectories in new environments. We design a crowdsourcing system - PlanIt, where non-expert users label segments of the robot's trajectory. PlanIt allows us to collect a large amount of user feedback, and using the weak and noisy labels from PlanIt we learn the parameters of our model. We test our approach on 122 different environments for robotic navigation and manipulation tasks. Our extensive experiments show that the learned cost function generates preferred trajectories in human environments. Our crowdsourcing system is publicly available for the visualization of the learned costs and for providing preference feedback: http://planit.cs.cornell.edu

Self-folding robot is usually modeled as rigid origami, a class of origami whose entire surface remains rigid during folding except at crease lines. In this work, we focus on finding valid folding motion that brings the origami from the unfolded state continuously to the folded state. Although recent computational methods allow rapid simulation of folding process of certain rigid origami, these methods can fail even when the input crease pattern is extremely simple but requires implicit folding orders. Moreover, due to the rigidity requirement, the probability of generating a valid configuration via uniform sampling is zero, which greatly hinders the applicability of traditional sampling-based motion planners. We propose a novel sampling strategy that samples in the discrete domain. Our experimental results show that the proposed method could efficiently generate valid configurations. Using those configurations, the planner successfully folds several types of rigid origami that the existing methods fail to fold and could discover multiple folding paths in different homotopies.

Modeling the effects of actions based on the state of the world enables robots to make intelligent decisions in different situations. However, it is often infeasible to have globally accurate models. Task performance is often hindered by discrepancies between models and the real world, since the true outcome of executing a plan may be significantly worse than the expected outcome used during planning. Furthermore, expectations about the world are often stochastic in robotics, making the discovery of model-world discrepancies non-trivial. We present an execution monitoring framework capable of finding statistically significant discrepancies, determining the situations in which they occur, and making simple corrections to the world model to improve performance. In our approach, plans are initially based on a model of the world that is only as faithful as computational and algorithmic limitations allow. Through experience, the monitor discovers previously unmodeled modes of the world, defined as regions of a feature space in which the experienced outcome of a plan deviates significantly from the predicted outcome. The monitor may then make suggestions to change the model to match the real world more accurately. We demonstrate this approach on the adversarial domain of robot soccer: we monitor pass interception performance of potentially unknown opponents to try to find unforeseen modes of behavior that affect their interception performance.