ICRA 2019 Papers

This paper presents a nonlinear, backstepping depth and pitch controller for a dual-bladder buoyancy engine actuated by gear pumps. Flow-rate feedback is obtained using a custom flow sensor comprised of a differential pressure sensor and a small, 3D-printed attachment. The controller is simulated using a model of the CephaloBot, our in-house developed autonomous underwater vehicle (AUV). Its depth control capability is also experimentally validated using a single-bladder buoyancy engine on-board a smaller-scale test cylinder. Lyapunov stability analysis shows global, asymptotic stability, which is exhibited in our simulation. Our experiments verify that this buoyancy engine is a feasible and effective depth controller for AUVs.

This paper addresses the problem of cooperative object transportation for multiple Underwater Vehicle Manipulator Systems (UVMSs) in a constrained workspace involving static obstacles. We propose a Nonlinear Model Predictive Control (NMPC) approach for a team of UVMSs in order to transport an object while avoiding significant constraints and limitations such as: kinematic and representation singularities, obstacles within the workspace, joint limits and control input saturations. More precisely, by exploiting the coupled dynamics between the robots and the object, and using certain load sharing coefficients, we design a distributed NMPC for each UVMS in order to cooperatively transport the object within the workspace's feasible region. Moreover, the control scheme adopts load sharing among the UVMSs according to their specific payload capabilities. Additionally, the feedback relies on each UVMS's locally measurements and no explicit data is exchanged online among the robots, thus reducing the required communication bandwidth. Finally, real-time simulation results conducted in UwSim dynamic simulator running in ROS environment verify the efficiency of the theoretical finding.

As the demand for robotic applications in unconstrained and dynamic environments rises, so does the benefit of advancing the state of the art in soft robotic technologies. However, the complex capabilities of soft robots elicited by their high-dimensional, non-linear characteristics simultaneously yield difficult challenges in control and sensing. Moreover, embedding tactile sensing capabilities in soft materials is often expensive and difficult to fabricate. In recent years, however, the invention of small-scale depth-sensing cameras introduced a promising channel for soft tactile sensor design. In this work, we propose a novel soft device inspired by the human fingertip that not only utilizes a small depth camera as the perception mechanism, but also possesses compliance-modulating capabilities. We demonstrate its ability to accurately estimate contact regions upon interaction with an external obstacle, and show that the estimation sensitivity can be modulated via internal fluid states. In addition, we determine an empirical model of the device's force-deformation characteristics under simplifying assumptions, and validate its performance with real-time force matching control experiments.

This paper presents a three-tendon actuated continuum robot with an origami backbone to assist the elderly and physically impaired individuals in performing activities of daily living. The proposed design solution is an inherently safe and cost-effective alternative to current assistive robots. The origami backbone based on a variation of the Yoshimura pattern provides controlled deployment of the robot and enables length variation (15 cm - 56 cm) in order to increase the reachable workspace. A pneumatic stiffness mechanism was implemented, increasing the weight bearing capabilities of the continuum robot to 500 g. This new stiffness modulation approach was assessed with the use of several testing rigs. Additionally, the robot is joypad controlled and is easily transportable due to its high packing efficiency of 73% and light weight of 1.3 kg for the main body (including the actuation system). For demonstration of usability studies, the robot was successfully tested at a simulated kitchen terminal and also performed pick and place tasks.

The sensing modalities available to individual agents in a multi-robot team have a significant effect on what the team can accomplish. Previous work on infinitesimal shape-similarity has shown that maintaining relative angles between robots equipped with bearing-only sensors can render a formation of these robots invariant to translation, rotation, and uniform scaling; however, previous work has not proposed decentralized control strategies for exploiting this invariance. To address this deficiency, this paper proposes a decentralized formation control strategy for assembled triangulations, a class of infinitesimally shape-similar formations. Heterogeneous in terms of sensing and control, a decentralized formation control strategy is developed in which one robot sets the position of the formation, a robot capable of measuring bearings and distances controls the scale and heading, and the remaining robots maintain the assembled triangulation. The asymptotic controllers that compose the formation control strategy of this work are implemented on a team of differential-drive robots.

In this paper, we propose a semi-empirical approach for simulating robot locomotion on granular media. We first develop a contact model based on the stick-slip behavior between rigid objects and granular grains, which is then learned through running extensive experiments. The contact model represents all possible contact wrenches that the granular substrate can provide as a convex volume, which our method formulates as constraints in an optimization-based contact force solver. During simulation, granular substrates are treated as rigid objects that allow penetration and the contact solver solves for wrenches that maximize frictional dissipation. We show that our method is able to simulate plausible interaction response with several granular media at interactive rates.

Learning effective visuomotor policies for robots purely from data is challenging, but also appealing since a learning-based system should not require manual tuning or calibration. In the case of a robot operating in a real environment the training process can be costly, time-consuming, and even dangerous since failures are common at the start of training. For this reason, it is desirable to be able to leverage simulation and off-policy data to the extent possible to train the robot. In this work, we introduce a robust framework that plans in simulation and transfers well to the real environment. Our model incorporates a gradient-descent based planning module, which, given the initial image and goal image, encodes the images to a lower dimensional latent state and plans a trajectory to reach the goal. The model, consisting of the encoder and planner modules, is first trained through a meta-learning strategy in simulation. We subsequently perform adversarial domain transfer on the encoder by using a bank of unlabelled but random images from the simulation and real environments to enable the encoder to map images from the real and simulated environments to a similarly distributed latent representation. By fine tuning the entire model (encoder + planner) with only a few real world expert demonstrations, we show successful planning performances in different navigation tasks.

Modeling individual-specific gait dynamics based on kinematic data could aid development of gait rehabilitation robotics by enabling robots to predict the user's gait kinematics with and without external inputs, such as mechanical or electrical perturbations. Here we address a current limitation of data-driven gait models, which do not yet predict human gait dynamics nor responses to perturbations. We used Switched Linear Dynamical Systems (SLDS) to model joint angle kinematic data from healthy individuals walking on a treadmill during normal gait and during gait perturbed by functional electrical stimulation (FES) to the ankle muscles. Our SLDS models were able to generate joint angle trajectories in each of four gait phases, as well as across an entire gait cycle, given initial conditions and gait phase information. Because the SLDS dynamics matrices encoded significant coupling across joints that differed across indivdiuals, we compared the SLDS predictions to that of a kinematic model, where the joint angles were independent. Joint angle trajectories generated by SLDS and kinematic models were similar over time horizons of a few milliseconds, but SLDS models provided better predictions of gait kinematics over time horizons of up to a second. We also demonstrated that SLDS models can infer and predict individual-specific responses to FES during swing phase. As such, SLDS models may be a promising approach for online estimation and control of and human gait dynamics, allowing robotic control strategies to be tailored to an individual's specific gait coordination patterns.

The paper addresses the static balancing and walking of elastically actuated legged robots. The control is realized by commanding the motor positions only and exploiting the bijective relation between motor and link positions at equilibrium under static external forces. The approach is formulated in a quite general framework first. The main implementation contribution is the definition of a body coordinate system and of an appropriate set of constraints, which leads to a fully determined system of equations. In addition to the desired COM and the vertical foot positions, which are defined by the walking task and the terrain, the imposed constraints are related to distances between individual legs. The controller is experimentally validated on a compliantly actuated quadruped.

Modern lightweight dual-arm robots bring the physical capabilities to quickly take over tasks at typical industrial workplaces designed for workers. Low setup times - including the instructing/specifying of new tasks - are crucial to stay competitive. We propose a constraint programming approach to simultaneous task allocation and motion scheduling for such industrial manipulation and assembly tasks. Our approach covers the robot as well as connected machines. The key concept are Ordered Visiting Constraints, a descriptive and extensible model to specify such tasks with their spatiotemporal requirements and combinatorial or ordering constraints. Our solver integrates such task models and robot motion models into constraint optimization problems and solves them efficiently using various heuristics to produce makespan-optimized robot programs. For large manipulation tasks with 200 objects, our solver implemented using Google's Operations Research tools requires less than a minute to compute usable plans. The proposed task model is robot-independent and can easily be deployed to other robotic platforms. This portability is validated through several simulation-based experiments.

Compliant and precise rotary drive units are essential for the design of articulated robots that are capable of safe human-robot collaboration. In this paper, we present a new pneumatic rotary drive unit that combines the compliance of pneumatic systems with the ability to perform high precision positioning. We use pneumatic artificial muscles (PAMs) pulling on a swash plate to avoid the stick-slip phenomenon and to realize adjustable stiffness. Furthermore, the presented drive unit can operate in 360° continuous rotation. These properties make the drive particularly suitable for the later use in human-robot collaboration. We explain the mechanic design as well as the pneumatic and electric control system that we use to operate the drive unit. We derive the equations to calculate the static torque distribution and compare the theoretical results to the data measured on the realized laboratory test stand, depicted in figure 1. The accuracy of the used 16-bit encoder is achieved and adjustable stiffness is realized and measured on the laboratory test stand. The measurements of the reaction to a step response are discussed based on a first and basic control strategy.

In this paper, we study the problem of exploring a translating plume with a team of aerial robots. The shape and the size of the plume are unknown to the robots. The objective is to find a tour for each robot such that they collectively explore the plume. Specifically, the tours must be such that each point in the plume must be visible from the field-of-view of some robot along its tour. We propose a recursive Depth-First Search (DFS)-based algorithm that yields a constant competitive ratio for the exploration problem. The competitive ratio is 2(Sr + Sp)(R+⌊log R⌋)/(Sr + Sp)(R+⌊log R⌋) where R is the number of robots, and Sr and Sp are the robot speed and the plume speed, respectively. We also consider a more realistic scenario where the plume shape is not restricted to grid cells but an arbitrary shape. We show our algorithm has 2(Sr + Sp)(18 R+⌊log R⌋)/(Sr + Sp)(1+⌊log R⌋) competitive ratio under the fat condition. We empirically verify our algorithm using simulations.

Keypoint matching is an important operation in computer vision and its applications such as visual simultaneous localization and mapping (SLAM) in robotics. This matching operation heavily depends on the descriptors of the keypoints, and it must be performed reliably when images undergo condition changes such as those in illumination and viewpoint. Previous research in keypoint description has pursued three classes of descriptors: hand-crafted, those from trained convolutional neural networks (CNN), and those from pre-trained CNNs. This paper provides a comparative study of the three classes of keypoint descriptors, in terms of their ability to handle conditional changes. The study is conducted on the latest benchmark datasets in computer vision with challenging conditional changes. Our study finds that (a) in general CNN-based descriptors outperform hand-crafted descriptors, (b) the trained CNN descriptors perform better than pre-trained CNN descriptors with respect to viewpoint changes, and (c) pre-trained CNN descriptors perform better than trained CNN descriptors with respect to illumination changes. These findings can serve as a basis for selecting appropriate keypoint descriptors for various applications.

In this paper we present a novel method of categorizing naturalistic human arm motions during activities of daily living using clustering techniques. While many current approaches attempt to define all arm motions using heuristic interpretation, or a combination of several abstract motion primitives, our unsupervised approach generates a hierarchical description of natural human motion with well recognized groups. Reliable recommendation of a subset of motions for task achievement is beneficial to various fields, such as robotic and semi-autonomous prosthetic device applications. The proposed method makes use of well-known techniques such as dynamic time warping (DTW) to obtain a divergence measure between motion segments, DTW barycenter averaging (DBA) to get a motion average, and Ward's distance criterion to build the hierarchical tree. The clusters that emerge summarize the variety of recorded motions into the following general tasks: reach-to-front, transfer-box, drinking from vessel, on-table motion, turning a key or door knob, and reach-to-back pocket. The clustering methodology is justified by comparing against an alternative measure of divergence using Bezier coefficients and K-medoids clustering.

Hamilton-Jacobi (HJ) reachability analysis has been developed over the past decades into a widely-applicable tool for determining goal satisfaction and safety verification in nonlinear systems. While HJ reachability can be formulated very generally, computational complexity can be a serious impediment for many systems of practical interest. Much prior work has been devoted to computing approximate solutions to large reachability problems, yet many of these methods may only apply to very restrictive problem classes, do not generate controllers, and/or can be extremely conservative. In this paper, we present a new method for approximating the optimal controller of the HJ reachability problem for control-affine systems. While also a specific problem class, many dynamical systems of interest are, or can be well approximated, by control-affine models. We explicitly avoid storing a representation of the reachability value function, and instead learn a controller as a sequence of simple binary classifiers. We compare our approach to existing grid-based methodologies in HJ reachability and demonstrate its utility on several examples, including a physical quadrotor navigation task.

Robotic manipulation is a very active field of research nowadays; however, pick-and-place operations constitute the majority of today's industrial robotic applications. In order to adopt a robotic solution for an industrial setting, proper evaluation processes should be defined to assess the system's performance. A number of benchmarks have been proposed in the literature focusing mainly on individual components needed to perform the task, like grasping, perception and motion planning; thus, they do not provide enough information on the performance of the entire robotic system. To address this, we propose a benchmarking framework for a pick-and-place task inspired by a use case for picking fruits and vegetables in an industrial setting. To foster reproducible research and comparison of different robotic systems, the benchmarking framework uses surrogate objects with instructions on how to build them, an easy-to-reproduce environment, and guidelines for object placement. The proposed benchmark is applied to evaluate the performance of two variants of a robotic system with different end-effectors.

Loss of functional motor skills are common and often require patients to undergo rehabilitation so that they have a chance at motor recovery. Advancement in technology has seen to a rise in the use of robotic technology in conducting rehabilitative exercises that are traditionally carried out by physiotherapists. In recent years, soft robotic exoskeletons, using pneumatic-based actuation in particular, have gained much interest due to their compliant characteristics and safe operating conditions. In order to carry out complex task-based rehabilitative exercises, these soft pneumatic actuators must ideally be able to move with multiple degrees of freedom or minimally, in a bidirectional motion. Majority of the research covering soft actuators can only achieve finger flexion with some providing passive finger extension. Non-invasive intent detection in the control of these exoskeletons is also lacking in sensing both finger flexion and extension. In this paper we present our work on a fold-based bidirectional 3D printed intent-sensing soft pneumatic actuator (ISPA) that can achieve bidirectional motion and provide intent detection for finger flexion and extension for application in upper limb rehabilitative exoskeletons.

This work demonstrates 3D printed soft actuators with complex shapes and remote actuation using an external magnetic field. Instead of embedding magnetic particles in a polymeric matrix, we fabricated a novel ferrofluid-based actuator, in which the fluid can be moved to different locations in the actuator to affect actuator response. We studied the effect of both the ferrofluid and the 3D printed material on the motion of simple actuators using 3D printed tubes. In addition, we 3D printed more complex actuators mimicking a human hand and a worm to demonstrate more complex motion.