IROS 2020 论文列表

This paper investigates the specific scenario of high-intensity confrontations between humans and robots, to understand how robots can defuse the conflict. It focuses on the effectiveness of using five different affective expression modalities as main drivers for defusing the conflict. The aim is to discover any strengths or weaknesses in using each modality to mitigate the hostility that people feel towards a poorly performing robot. The defusing of the situation is accomplished by making the robot better at acknowledging the conflict and by letting it express remorse. To facilitate the tests, we used a custom affective robot in a simulated conflict situation with 105 test participants. The results show that all tested expression modalities can successfully be used to defuse the situation and convey an acknowledgment of the confrontation. The ratings were remarkably similar, but the movement modality was different (ANON p<; .05) than the other modalities. The test participants also had similar affective interpretations on how impacted the robot was of the confrontation across all expression modalities. This indicates that defusing a high-intensity interaction may not demand special attention to the expression abilities of the robot, but rather require attention to the abilities of being socially aware of the situation and reacting in accordance with it.

In smart home environments, central interfaces that take commands from users and give orders to each relevant device appropriately are increasingly important. We investigated the type of central interface that consumers are more willing to adopt and whether these interfaces enhance the evaluation of services provided by smart home devices. This study confirms that speaker interfaces are preferred over social robots, speaker interfaces are perceived by users as more persuasive, and the adoption of central interfaces increases the overall service evaluation.

Modular robotic systems with self-repair or self-replication capabilities have been presented as a robust, low cost solution to extraterrestrial or Arctic exploration. This paper explores using ice as the sole structure element to build robots. The ice allows for increased flexibility in the system design, enabling the robotic structure to be designed and built post deployment, after tasks and terrain obstacles have been better identified and analyzed. However, ice presents many difficulties in manufacturing. The authors explore a structure driven approach to examine compatible manufacturing processes with an emphasis on conserving process energies. The energy analysis shows the optimal manufacturing technique depends on the volume of the final part relative to the volume of material that must be removed. Based on experiments three general design principles are presented. A mobile robotic platform made from ice is presented as a proof of concept and first demonstration.

Robotic packaging of fresh leafy produce such as herbs and salads generally involves picking out a target mass from a pile or crate of plant material. Typically, for low-complexity parallel grippers, the weight picked can be controlled by varying the opening aperture. However, often individual strands of plant material get entangled with each other, causing more to be picked out than desired. This paper presents a simple spread-and-pick approach that significantly reduces the degree of entanglement in a herb pile when picking. Compared to the traditional approach of picking from an entanglement-free point in the pile, the proposed approach results in a decrease of up to 29.06% of the variance in for separate homogeneous piles of fresh herbs. Moreover, it shows good generalisation with up to 55.53% decrease in picked weight variance for herbs previously unseen by the system.

Understanding spatial relations of objects is critical in many robotic applications such as grasping, manipulation, and obstacle avoidance. Humans can simply reason object's spatial relations from a glimpse of a scene based on prior knowledge of spatial constraints. The proposed method enables a robot to comprehend spatial relationships among objects from RGB-D data. This paper proposed a neural-logic learning framework to learn and reason spatial relations from raw data by following logic rules on spatial constraints. The neural-logic network consists of three blocks: grounding block, spatial logic block, and inference block. The grounding block extracts high-level features from the raw sensory data. The spatial logic blocks can predicate fundamental spatial relations by training a neural network with spatial constraints. The inference block can infer complex spatial relations based on the predicated fundamental spatial relations. Simulations and robotic experiments evaluated the performance of the proposed method.

We propose a model-free algorithm for learning efficient policies capable of returning table tennis balls by controlling robot joints at a rate of 100Hz. We demonstrate that evolutionary search (ES) methods acting on CNN-based policy architectures for non-visual inputs and convolving across time learn compact controllers leading to smooth motions. Furthermore, we show that with appropriately tuned curriculum learning on the task and rewards, policies are capable of developing multi-modal styles, specifically forehand and backhand stroke, whilst achieving 80% return rate on a wide range of ball throws. We observe that multi-modality does not require any architectural priors, such as multi-head architectures or hierarchical policies.

Cell manipulation is a critical procedure in related biological applications such as embryo biopsy and intracytoplasmic sperm injection (ICSI), where the biological cell is required to be oriented to the desired position. To bridge the gap between the techniques and the clinical applications, a robotic micromanipulation method, which utilizes friction forces to rotate the cell with standard micropipettes, is presented in this paper. Force models for both in-plane and out-of-plane rotations are well established and analyzed for the rotation control. For better controllability, calibration steps are also designed for adjusting the orientation of the micropipette with a more efficient way. A cell orientation recognition algorithm based on the superpixel segmentation and spectral clustering is reported and achieved high validation accuracy (96%) for estimating the orientation of the oocyte. The extracted visual information further facilitates the feedback control of cell rotation. Experimental results show that the overall success rate for the cell rotation control was about 95% with orientation precision of ±1°.

Visual robot perception has been challenging to successful robot manipulation in noisy, cluttered and dynamic environments. While some perception systems fail to provide an adequate semantics of the scene, others fail to present appropriate learning models and training data. Another major issue encountered in some robot perception systems is their inability to promptly respond to robot control programs whose realtimeness is crucial.This paper proposes an architecture to robot vision for manipulation tasks that addresses the three issues mentioned above. The architecture encompasses a generator of training datasets and a learnable scene describer, coined as RobotVQA for Robot Visual Question Answering. The architecture leverages the power of deep learning to predict and photo-realistic virtual worlds to train. RobotVQA takes as input a robot scene's RGB or RGBD image, detects all relevant objects in it, then describes in realtime each object in terms of category, color, material, shape, openability, 6D-pose and segmentation mask. Moreover, RobotVQA computes the qualitative spatial relations among those objects. We refer to such a scene description in this paper as scene graph or semantic graph of the scene. In RobotVQA, prediction and training take place in a unified manner. Finally, we demonstrate how RobotVQA is suitable for robot control systems that interpret perception as a question answering process.

In this paper, the 2D robot-to-robot relative pose (position and orientation) estimation problem based on ego-motion and noisy distance measurements is considered. We address this problem using an optimization-based method, which does not require complicated numerical analysis while yields no inferior relative localization (RL) results compared to existing approaches. In particular, we start from a state-of-the-art method named square distances weighted least square (SD-WLS), and reformulate it as a non-convex quadratically constrained quadratic programming (QCQP) problem. To handle its non-convex nature, a semidefinite programming (SDP) relaxation optimization-based method is proposed, and we prove that the relaxation is tight when measurements are free from noise or just corrupted by small noise. Further, to obtain the optimal solution of the relative pose estimation problem in the sense of maximum likelihood estimation (MLE), a theoretically optimal WLS method is developed to refine the estimate from the SDP optimization. Comprehensive simulations and well-designed experiments are presented for validating the tightness of the SDP relaxation, and the effectiveness of the proposed algorithm is highlighted by comparing it to the existing approaches.

One out of eight women will get breast cancer during their lifetime. A biopsy, a procedure in which a tissue sample is acquired from the lesion, is required to confirm the diagnosis. A biopsy is preferably executed under ultrasound (US) guidance because it is simple, fast, and cheap, gives real-time image feedback and causes little patient discomfort. However, Magnetic Resonance (MR)-detected lesions may be barely or not visible on US and difficult to find due to deformations of the breast. This paper presents a robotic setup and workflow that assists the radiologist in targeting MR-detected breast lesions under US guidance, taking into account deformations and giving the radiologist robotic accuracy. The setup consists of a seven degree-of-freedom robotic serial manipulator equipped with an end-effector carrying a US transducer and a three degree-of-freedom actuated needle guide. During probe positioning, the US probe is positioned on the patient's skin while the system tracks skin contact and tissue deformation. During the intervention phase, the radiologist inserts the needle through the actuated guide. During insertion, the tissue deformation is tracked and the needle path is adjusted accordingly. The workflow is demonstrated on a breast phantom. It is shown that lesions with a radius down to 2.9 mm can be targeted. While MRI is becoming more important in breast cancer detection, the presented robot-assisted approach helps the radiologist to effectively and accurately confirm the diagnosis utilizing the preferred US-guided method.

We explore the interpretation of sound for robot decision making, inspired by human speech comprehension. While previous methods separate sound processing unit and robot controller, we propose an end-to-end deep neural network which directly interprets sound commands for visual-based decision making. The network is trained using reinforcement learning with auxiliary losses on the sight and sound networks. We demonstrate our approach on two robots, a TurtleBot3 and a Kuka-IIWA arm, which hear a command word, identify the associated target object, and perform precise control to reach the target. For both robots, we show the effectiveness of our network in generalization to sound types and robotic tasks empirically. We successfully transfer the policy learned in simulator to a real-world TurtleBot3.

Mobile robots operating in public environments require the ability to navigate among humans and other obstacles in a socially compliant and safe manner. This work presents a combined imitation learning and deep reinforcement learning approach for motion planning in such crowded and cluttered environments. By separately processing information related to static and dynamic objects, we enable our network to learn motion patterns that are tailored to real-world environments. Our model is also designed such that it can handle usual cases in which robots can be equipped with sensor suites that only offer limited field of view. Our model outperforms current state-of-the-art approaches, which is shown in simulated environments containing human-like agents and static obstacles. Additionally, we demonstrate the real-time performance and applicability of our model by successfully navigating a robotic platform through real-world environments.

Previous work has shown that interacting with a human adversary can significantly improve the efficiency of the learning process in robot grasping. However, people are not consistent in applying adversarial forces; instead they may alternate between acting antagonistically with the robot or helping the robot achieve its tasks. We propose a physical framework for robot learning in a mixed adversarial/collaborative setting, where a second agent may act as a collaborator or as an antagonist, unbeknownst to the robot. The framework leverages prior estimates of the reward function to infer whether the actions of the second agent are collaborative or adversarial. Integrating the inference in an adversarial learning algorithm can significantly improve the robustness of learned grasps in a manipulation task.

Deploying robot learning frameworks in unconstrained environments requires robustness and tractability. We must not only equip the robot with a sufficient range of sensing capabilities, but also provide training data in a sample-efficient manner. To this end, we identify and address a need specifically in robot learning from demonstration (LfD) literature to account for not only end-effector pose and wrench signals, but also tactile signals for contact. While traditional pose and wrench signals have proven to be sufficient for robots to learn basic position and force-control behaviors, they are inherently too constraining for the learning of general manipulation tasks. In particular, useful manipulation tasks often rely on the geometry of the contact interaction. To explore the value of geometry-based tactile signals, we utilize a LfD framework built upon hidden Markov models and Gaussian mixture regression, adapt it to our robotic system equipped with a soft tactile sensor, and validate its performance with an edge-following task and a manipulation task involving different object geometries.

Chinese calligraphy is a unique art form with great artistic value but difficult to master. In this paper, we formulate the calligraphy writing problem as a trajectory optimization problem, and propose an improved virtual brush model for simulating the real writing process. Our approach is inspired by pseudospectral optimal control in that we parameterize the actuator trajectory for each stroke as a Chebyshev polynomial. The proposed dynamic virtual brush model plays a key role in formulating the objective function to be optimized. Our approach shows excellent performance in drawing aesthetically pleasing characters, and does so much more efficiently than previous work, opening up the possibility to achieve real-time closed-loop control.

This paper presents a novel autonomous surface vessel (ASV), called Roboat II for urban transportation. Roboat II is capable of accurate simultaneous localization and mapping (SLAM), receding horizon tracking control and estimation, and path planning. Roboat II is designed to maximize the internal space for transport, and can carry payloads several times of its own weight. Moreover, it is capable of holonomic motions to facilitate transporting, docking, and inter-connectivity between boats. The proposed SLAM system receives sensor data from a 3D LiDAR, an IMU, and a GPS, and utilizes a factor graph to tackle the multi-sensor fusion problem. To cope with the complex dynamics in the water, Roboat II employs an online nonlinear model predictive controller (NMPC), where we experimentally estimated the dynamical model of the vessel in order to achieve superior performance for tracking control. The states of Roboat II are simultaneously estimated using a nonlinear moving horizon estimation (NMHE) algorithm. Experiments demonstrate that Roboat II is able to successfully perform online mapping and localization, plan its path and robustly track the planned trajectory in the confined river, implying that this autonomous vessel holds the promise on potential applications in transporting humans and goods in many of the waterways nowadays.

Precomputed roadmaps can enable effective multi-query motion planning: a roadmap can be built for a robot as if no obstacles were present, and then after edges invalidated by obstacles observed at query time are deleted, path search through the remaining roadmap returns a collision-free plan. However, large roadmaps are memory intensive to store, and can be too slow for practical use. We present an algorithm for compressing a large roadmap so that the collision detection phase fits into a computational budget, while retaining a high probability of finding high-quality paths. Our algorithm adapts work from graph theory and data mining by treating roadmaps as unreliable networks, where the probability of edge failure models the probability of a query-time obstacle causing a collision. We experimentally evaluate the quality of the resulting roadmaps in a suite of four motion planning benchmarks.

Robots encounter many risks that threaten the success of practical locomotion tasks. Legs break, electrical components overheat, and feet can unexpectedly slip. When all risks cannot be completely avoided, how does a robot decide its best action? We present a method for planning robot motions by reasoning about risk-of-failure probabilities instead of applying cost-penalty functions or inflexible path constraints. This work develops a risk-constrained formulation that can be straightforwardly included in existing motion planning optimizations. The risk constraints scale tractably with many risk sources, and in some cases, only add linear constraints to the optimization problem and are therefore compatible with model-predictive control techniques. We present a toy "Puck World" proof-of-concept example and a practical implementation on a planar monopod robot that runs at 3.2 m/s when permitted to take high-risk maneuvers. We believe this risk approach can be used to optimize robot behaviors under numerous conflicting task pressures and model risk-conscious behaviors in animals.

This paper presents a novel online framework for safe crowd-robot interaction based on risk-sensitive stochastic optimal control, wherein the risk is modeled by the entropic risk measure. The sampling-based model predictive control relies on mode insertion gradient optimization for this risk measure as well as Trajectron++, a state-of-the-art generative model that produces multimodal probabilistic trajectory forecasts for multiple interacting agents. Our modular approach decouples the crowd-robot interaction into learning-based prediction and model-based control, which is advantageous compared to end-to-end policy learning methods in that it allows the robot's desired behavior to be specified at run time. In particular, we show that the robot exhibits diverse interaction behavior by varying the risk sensitivity parameter. A simulation study and a real-world experiment show that the proposed online framework can accomplish safe and efficient navigation while avoiding collisions with more than 50 humans in the scene.