ICRA 2019 Papers

We extend the ability of cameras to perceive obstacles for aircraft safety systems by enabling 3d sensing of free hanging wires. Our algorithm exploits the specialized 2d and 3d structure of wires to exceed state of the art performance in 2d sensing and 3d location estimation of wire obstacles. In 2d, a new neural network architecture, Deep Wire CNN, directly predicts the location of wire line segments in the image. In 3d, the detections are tracked and triangulated as the aircraft flies in order to estimate the wire's location. Our triangulation uses a new formulation of wire reconstruction as the estimation of the wire's vertical plane. Together these advancements enable real-time detections of wire hazards at ranges of over 1km. The system performance is evaluated on prior image level wire detection datasets and we introduce a new public dataset in order to evaluate full system results on over 40 approaches to power lines from a manned helicopter.

A key aspect to controlling and reducing the effects invasive insect species have on agriculture is to obtain knowledge about the migration patterns of these species. Current state-of-the-art methods of studying these migration patterns involve a mark-release-recapture technique, in which insects are released after being marked and researchers attempt to recapture them later. However, this approach involves a human researcher manually searching for these insects in large fields and results in very low recapture rates. In this paper, we propose an automated system for detecting released insects using an unmanned aerial vehicle. This system utilizes ultraviolet lighting technology, digital cameras, and lightweight computer vision algorithms to more quickly and accurately detect insects compared to the current state of the art. The efficiency and accuracy that this system provides will allow for a more comprehensive understanding of invasive insect species migration patterns. Our experimental results demonstrate that our system can detect real target insects in field conditions with high precision and recall rates.

In order to solve robotic bin-picking problem in many industrial applications, accurate 6D object pose estimation is one of fundamental technologies. This paper presents a method for accurate 6D pose estimation from a single RGB image for texture-less industrial parts. These objects are common but still challenging to deal with, due to the fact that poor surface texture and brightness makes difficult to compute discriminative local appearance descriptors. The proposed method mainly consists of two stages, which ranges from the detection stage to the optimization stage. Firstly, all known objects in the RGB image are detected with 2D bounding box via a tiny convolutional neural network. Then, the second stage will optimize the 6D pose in the Edge image given several coarse initializations. These coarse initializations are generated from the Edge image via a hypothesis-evaluation scheme. Furthermore, the proposed method is validated by achieving state-of-the-art results of texture-less industrial parts for RGB input. According to practical experiments, the proposed method is accurate and robust enough to be applied on the robotic manipulation platform to complete a simple assembly task.

Epidural anesthesia, despite being a relatively common medical procedure, remains quite demanding in terms of skills as it is mostly blind and thus heavily reliant on the haptic sensations. Although some training support solutions exist, anesthetists consider them mostly inefficient or impractical. A few attempts at creating a simulator for this particular procedure exist but each one lacks one of the important requirements of the procedure. This article introduces a haptic simulator featuring a more complete and realistic simulation of the procedure than we could observe in existing simulators. The simulator is composed of a generic electrical haptic interface coupled with a pneumatic cylinder.

In this paper, we present an autonomous AI system designed for a Human-Robot Interaction (HRI) study, set around a dice game scenario. We conduct a case study to answer our research question: Does a robot with a socially engaged personality lead to a higher acceptance than a competitive personality? The flexibility of our proposed system allows us to construct and attribute two different personalities to a humanoid robot: a socially engaged personality that maximizes its user interaction and a competitive personality that is focused on playing and winning the game. We evaluate both personalities in a user study, in which the participants play a turn-taking dice game with the robot. Each personality is assessed with four different evaluation tools: 1) the Godspeed Questionnaire, 2) the Mind Perception Questionnaire, 3) a custom questionnaire concerning the overall HRI experience, and 4) a Convolutional Neural Network analyzing the emotions on the participants' facial feedback throughout the game. Our results show that the socially engaged personality evokes stronger emotions among the participants and is rated higher in likability and animacy than the competitive one. We conclude that designing the robot with a socially engaged personality contributes to a higher acceptance within an HRI scenario.

In this paper, we design novel liquid time-constant recurrent neural networks for robotic control, inspired by the brain of the nematode, C. elegans. In the worm's nervous system, neurons communicate through nonlinear time-varying synaptic links established amongst them by their particular wiring structure. This property enables neurons to express liquid time-constants dynamics and therefore allows the network to originate complex behaviors with a small number of neurons. We identify neuron-pair communication motifs as design operators and use them to configure compact neuronal network structures to govern sequential robotic tasks. The networks are systematically designed to map the environmental observations to motor actions, by their hierarchical topology from sensory neurons, through recurrently-wired interneurons, to motor neurons. The networks are then parametrized in a supervised-learning scheme by a search-based algorithm. We demonstrate that obtained networks realize interpretable dynamics. We evaluate their performance in controlling mobile and arm robots, and compare their attributes to other artificial neural network-based control agents. Finally, we experimentally show their superior resilience to environmental noise, compared to the existing machine learning-based methods.

This paper presents the design of a soft ankle-foot orthosis (AFO) exosuit to aid natural gait restoration for individuals suffering from foot drop. The sock-like AFO is comprised of soft actuators made from fabric-based, thermally-bonded nylon and designed to be worn over the users shoes. The system assists dorsiflexion during swing phase of the gait cycle utilizing a contracting soft actuator, and provides ankle joint proprioception during stance with a variable stiffness soft actuator. A computational model is developed using finite element analysis to optimize the performance characteristics of the fabric actuators prior to fabrication, maximize contraction, and minimize overall volume. The dorsiflexion actuator is able to achieve a linear tensile force of 197 N at 200 kPa. The variable stiffness actuator generates up to 1. 2 Nm of torque at the same pressure. The computational model and soft AFO are experimentally validated and with a healthy participant through kinematic and electromyography studies. When active the AFO is capable of reducing by 13.3% the activity of the muscle responsible for ankle dorsiflexion during the swing phase.

This paper presents the development of a tendon-driven continuum robot segment with a modular design, simple construction and significant lifting capabilities. The segment features a continuous flexible core combined with rigid interlocking vertebrae evenly distributed along its length. This design allows bending in two degrees of freedom while minimising torsional movement. The segment is actuated by two antagonistic tendon pairs, each of which is driven by a single geared DC motor. Modularity is achieved by embedding these motors in one end of the segment, avoiding the need for a bulky actuation unit and allowing variable numbers of segments to be connected. The design features a large hollow central bore which could be used as a vacuum channel for suction-assisted gripping or to allow ingress and egress of fluids. The design process goes through four iterations, the final two of which are subjected to quantitative experiments to evaluate workspace, lifting capabilities and torsional rigidity. All iterations are fabricated using multi-material 3D printing, which allows the entire structure to be printed as a pre-assembled unit with the rigid vertebrae fused to the flexible core. Assembly is then a simple case of inserting the motors and connecting the tendons. This unconventional manufacturing approach is found to be efficient, effective and relatively cheap.

Virtually every mechanized form of transportation, power generation system, industrial equipment, and robotic system has rotating shafts. As the shaft is often the main means of mechanical power transmission, measuring the torque, speed, vibration, and bending of the shaft can be used in many cases to access device performance and health and to implement controls. This paper proposes a shaft sensor that measures all of these phenomena with reasonable accuracy while having a low cost and simple installation process. This sensor transfers strain from the shaft and amplifies it to increase sensitivity. Furthermore, this sensor requires no components to be in the stationary reference frame, allowing the entire device to rotate with the shaft. A prototype is presented. Experimental results illustrate the effectiveness of the proposed system.

Cell microinjection plays an important role in genetics, transgenics, and other biomedical fields. As compared with manual cell microinjection and position-based robotic cell microinjection, force-assisted robotic cell microinjection can improve the success rate and survival rate of the injected cells. In this paper, a novel force-sensing cell injector is designed with piezoresistive force sensor embedded in soft materials. The soft sensors act as fixed-guided beams, which are introduced to achieve the force measurement with high sensitivity in pure one-degree-of-freedom (1-DOF) direction. The injector is developed by considering the installation and replacement issues of the micropipette as well as the connection convenience between the micropipette and tube of compressed air. A prototype of the cell injector with the force sensor is fabricated. Experimental study is conducted to verify its performance in practice.

We present a low-cost, three-degree-of-freedom (3-DOF) prismatic-spherical-revolute (PSR) parallel mechanism used as a testing platform for an unmanned aerial vehicle (UAV) tethered to an unmanned surface vehicle (USV). The mechanism has three actuated linear rails kinematically linked to a platform which replicates boat motion up to 2.5 m vertical heave (sea state 4, Douglas Sea Scale). A lookup table relating relative slider heights to platform roll and pitch was developed numerically leveraging geometric constraints. A design parameter study optimized the arm length, platform size, and ball joint mounting angle relative to the overall radius to maximize the workspace. For this design, a maximum roll and pitch range from -32° to 32° and -25° to 35°, respectively, is achievable. A prototype was manufactured to carry the tethered UAV winch payload. Experimental testing confirmed the workspace and demonstrated boat motion replication, validated using an inertial measurement unit (IMU).

This paper presents a novel light-weight back-drivable inherently-safe robotic mechanism for delivering ankle rehabilitation therapies. The robot is designed to be used as the ankle module of a multi-purpose lower-limb rehabilitation robot. A novel friction-based safety feature has been introduced that enables mechanical adjustment of the maximum amount of allowable transfer forces and torques to the patient's limb. The design procedure, mathematical modeling and experimental validations are provided to demonstrate the performance of the proposed system.

Autonomous robotic weeding systems in precision farming have demonstrated their full potential to alleviate the current dependency on herbicides or pesticides by introducing selective spraying or mechanical weed removal modules, thus reducing the environmental pollution and improving the sustainability. However, most previous works require fast weed detection system to achieve real-time treatment. In this paper, a novel computer vision based weeding control system is presented, where a non-overlapping multi-camera system is introduced to compensate the indeterminate classification delays, thus allowing for more complicated and advanced detection algorithms, e.g. deep learning based methods. The suitable tracking and control strategies are developed to achieve accurate and robust in-row weed treatment, and the performance of the proposed system is evaluated in different terrain conditions in the presence of various delays.

This project aims at developing a bio-inspired climbing robotic technology for cable inspection on the cable-stayed bridge. The design and implementation of a palm-based cable climbing robot: CCRobot-II with mass 25 kg, maximal payload 30kg and maximal length 1.1 m are described. CCRobot-II consists of several novel design features, including the palm-based gripping module, the alternating-sliding frame specialized for high-speed climbing, and the following wheels. With its carefully designed mechanism, climbing gait and trajectory algorithm, CCRobot-II is able to crawl along a bridge cable with a maximal speed 5.2 m/min. This speed has been hardly achieved by a existing climbing robot with such a large scale, and it is nearly the twice as the climbing speed of CCRobot-I which is designed previously. CCRobot-II also works effectively even if the bridge cable surface is attached with small obstacles. Experiments have been conducted, the results show that CCRobot-II has potential engineering applications on the cable-stayed bridge for fieldwork.

Autonomous vehicles apply pertinent planning and control algorithms under different driving conditions. The mode switch between these algorithms should also be autonomous. On top of the nominal planners, a safe fallback routine is needed to stop the vehicle at a safe position if nominal operational conditions are violated, such as for a system failure. This paper describes the design and formal verification of a supervisor to manage all requirements for mode switching between nominal planners, and additional requirements for switching to a safe stop trajectory planner that acts as the fallback routine. The supervisor is designed via a model-based approach and its abstraction is formally verified by model checking. The supervisor is implemented and integrated with the Research Concept Vehicle, an experimental research and demonstration vehicle developed at the KTH Royal Institute of Technology. Simulations and experiments show that the vehicle is able to autonomously drive in a safe manner between two parking lots and can successfully come to a safe stop upon GPS sensor failure.

We present a transformable catheter head structure for endoscopic catheter allowing the simultaneous use of a camera module and a large tool channel introduced through a small incision. At the site of interest, the head with a camera can be expanded from the initial straight configuration, which opens a window for advancing a tool that is located behind the camera. Two different designs were proposed and prototyped. One option has flexure joints directly fabricated at the distal end of a polymer catheter by laser micro-machining, while another design employs a hinged metal head assembled at the tip of the same type of catheter. The kinematic behavior of each head was evaluated during the head-up and tip steering motions, and compared with each other to draw a selection guideline between them. Experimental results prove the feasibility of the proposed head structure for smarter endoscopic catheters.

Aquatic-aerial multimodal vehicle is a new concept aircraft that can freely shuttle between water and air. Some of the natural organisms provide the inspiration to realize this multi-locomotion. Most of current prototypes use rigid link mechanisms or hinges to morph the structure thus to adapt to the aquatic-aerial environment, which is commonly complicated and bulky. In this paper, we present a novel prototype with pneumatically-driven soft fins and arms that can fold and spread just like the flying squid. The fins and arms can augment the lift force during flying by spreading and reduce drag force during swimming by folding. The performance of the morphable structures was investigated in wind and water tunnel. The results explain the tradeoff strategies of multimodal-locomotion between water and air, and verify the feasibility of the novel aquatic-aerial vehicle with soft morphable structures.

Presented in this paper are the design and experiments for a transformable Vertical Take Off and Landing (VTOL) UAV. This work demonstrates shape-shifting transformation, building upon the conceptual designs put forth in [1] and [2], along with hardware prototyping and component testing from [3]. A deterministic model is presented to characterize the flight of the MIST-UAV in simulation. Experimental results from the platform demonstrate for the first time successful in-air transformation from multi-rotor, tail-sitter, and fixed-wing operation. Experiments also validated transformation repeatability, successfully testing multiple sequential transformations.

Haptic joysticks for man-machine interaction used in aerospace flight control have highly demanding requirements of reliability, force density, and high dynamics that can hardly be met with conventional electromagnetic actuators. This work explores the potential of using an alternative actuation strategy based on hyper-redundant MR clutches that modulate the force of a tendon-driven 2-degree-of-freedom spherical gimbal. A system design and its closed-loop force control scheme are proposed. Experimental results for an open-loop characterization, static force control and dynamic force control are set out and compared with typical requirements for such devices from the literature. Results show that the proposed architecture leads to one of the lightest systems reported in the literature that has the potential to meet reliability requirements by providing a jam-free design with duplex fault tolerance, and yet, can generate high force levels while providing enough force resolution. The approach is promising and can extend to high-performance collaborative robot applications.