IROS 2021 论文列表

Disentangling two or more cables often requires many steps to remove crossings between and within cables. We formalize the problem of disentangling multiple cables and present an algorithm, Iterative Reduction Of Non-planar Multiple cAble kNots (IRON-MAN), that outputs robot actions to remove crossings from multi-cable knotted structures. IRON-MAN uses a learned perception system inspired by prior work in single-cable untying to imitate a graph-based supervisor, and operates on RGB image inputs of the workspace. Given a sequence of images as input, the system can disentangle two-cable twists, three-cable braids, and knots of two or three cables, such as overhand, square, carrick bend, sheet bend, crown, and fisherman’s knots. IRON-MAN keeps track of task-relevant keypoints corresponding to cable endpoints and crossings and iteratively disentangles the cables by identifying and undoing crossings that are critical to knot structure. Using a da Vinci surgical robot, we experimentally evaluate the effectiveness of IRON-MAN on disentangling multi-cable knots of types that appear in the training data, as well as generalizing to novel classes of multi-cable knots. Results suggest that IRON-MAN is effective in disentangling knots involving up to three cables with 80.5% success and generalizing to knot types that are not present during training, with cables of identical or distinct colors. Supplementary material and videos can be found at https://tinyurl.com/multi-cable-disentangling.

Robust motion planners for unmanned ground vehicles must minimize risk while obeying vehicle mobility constraints. Algorithms such as the State Lattice (SL) utilize offline computation to generate expressive control sets which form recombinant search spaces, enabling the use of heuristic search to efficiently produce feasible motion plans online. The Adaptive State Lattice (ASL) demonstrated that local optimizations of the continuous states explored by heuristic search can produce lower-cost solutions in less time than more densely sampled unadapted lattices in sufficiently complex environments. However, the computational cost of this online adaptation limits the application of ASL for mobile robot navigation. We present the Efficiently Adaptive State Lattice (EASL), a novel formalism for online discrete ASL adaptation to overcome this limitation. By discretizing the space of states considered during adaptation, EASL limits the set of feasible motions which could arise during search. This permits the precomputation of an approximation of all motions that could be expressed by an ASL. This approximation removes the online trajectory generation component of the ASL while retaining the benefits of lattice adaptation and enables the use of precomputed swaths for evaluating edge costs. Experimental results demonstrate how an EASL-based planner can generate lower-cost paths than a SL-based planner in roughly equal to or less than the same amount of time.

In this paper, we propose a novel algorithm for fusing a sequence of 3D images, named as Direct Bundle Adjustment (DBA). This algorithm simultaneously optimizes the global pose parameters of image frames and the intensity values of the fused global image using the 3D image data directly (without extracting features from the images). This one-step 3D image fusion approach is achieved by formulating the problem as an optimization problem to minimize the intensity differences between the global image and the corresponding points in the different local images. The proposed DBA method is particularly useful in the scenarios where distinct features are not available, such as Transesophageal Echocardiography (TEE) images. We validate the proposed method via simulated and in-vivo 3D TEE images. It is shown that the proposed method is robust to intensity noises and much more accurate than the conventional sequential fusion method.

A recent line of work has shown that end-to-end optimization of Bayesian filters can be used to learn state estimators for systems whose underlying models are difficult to hand-design or tune, while retaining the core advantages of probabilistic state estimation. As an alternative approach for state estimation in these settings, we present an end-to-end approach for learning state estimators modeled as factor graph-based smoothers. By unrolling the optimizer we use for maximum a posteriori inference in these probabilistic graphical models, our method is able to learn probabilistic system models in the full context of an overall state estimator, while also taking advantage of the distinct accuracy and runtime advantages that smoothers offer over recursive filters. We study our approach using two fundamental state estimation problems, object tracking and visual odometry, where we demonstrate a significant improvement over existing baselines. Our work comes with an extensive code release, which includes training and evaluation scripts, as well as Python libraries for Lie theory and factor graph optimization: https://sites.google.com/view/diffsmoothing/.

Traditional decision and planning frameworks for self-driving vehicles (SDVs) scale poorly in new scenarios, thus they require tedious hand-tuning of rules and parameters to maintain acceptable performance in all foreseeable cases. Recently, self-driving methods based on deep learning have shown promising results with better generalization capability but less hand engineering effort. However, most of the previous learning-based methods are trained and evaluated in limited driving scenarios with scattered tasks, such as lane-following, autonomous braking, and conditional driving. In this paper, we propose a graph-based deep network to achieve scalable self-driving that can handle massive traffic scenarios. Specifically, more than 7,000 km of evaluation is conducted in a high-fidelity driving simulator, in which our method can obey the traffic rules and safely navigate the vehicle in a large variety of urban, rural, and highway environments, including unprotected left turns, narrow roads, roundabouts, and pedestrian-rich intersections. Demonstration videos are available at https: //caipeide.github.io/dignet/.

Handling of fabric is a crucial step in the manufacturing of garments. This task is typically performed by trained workers who manipulate one sheet at a time, thus introducing a bottleneck in the automation of the textile industry. This paper seeks to address the challenge of picking fabric up by proposing a new method of achieving ply-separation. Our approach relies on a finger-tip sized (2 cm2) electroadhesive skin to lift fabric up. A pinch-type grasp is then used to securely hold the separated sheet of fabric, enabling easy manipulation thereafter. The ability to successfully pick up and manipulate a variety of commercial fabrics with diverse materials, shapes, sizes and textures is demonstrated. The ability to handle fabrics 100s of times larger than the electroadhesive skin is unique to our approach. Additionally, we demonstrate the manipulation of non-flat fabrics, a challenge that has not been previously addressed by electroadhesive approaches. We believe that this method introduces a smarter way of handling flexible and limp materials, showing great potential towards automation of garment manufacturing.

Embryos/oocytes vitrification is an essential cryopreservation technique in IVF (in vitro fertilization) clinics. The reliable and effective transferring of embryos/oocytes is crucial to the subsequent steps in the whole procedure of vitrification. After each transferring, the straw needs to be replaced with a new one. Due to the uncertainties in the fabrication and installation, the exact knowledge of the kinematic model of the straw is usually unknown, and the relationship between the microscope and the straw is also unknown without calibration beforehand. In such situation, automatically transferring the oocytes from micropipette to the narrow tip of straw (0.7mm) is very challenging. In this paper, a new vision-guided robotic system is developed to automate the transferring of the oocyte without calibration. To this end, the unknown depth information is estimated then compensated by constructing a deep vision network through microscope image, and an approximate Jacobian control algorithm is also proposed to servo control the end tip of the uncalibrated straw to contact the micropipette with the vision feedback. After that, the oocyte is automatically transferred from the micropipette to the straw to finalize the task. The stability of the closed-loop control system is rigorously proved with Lyapunov methods, and the effectiveness of the developed robot is validated in experiments.

The skin is an important organ which enables humans to interact with the unstructured environment around. It is perfectly soft and covers the entire body providing immediate feedback even when that part is not directly in the field of vision. With the human skin as an inspiration, in this paper, we develop a novel completely soft robot skin for tactile sensing. The skin utilizes a new type of material called as Permanent Magnet Elastomer (PME) to replace the traditionally used hard permanent magnet for hall effect based tactile sensors. PME is formed by mixing Neodymium particles in a polymer base and using strong magnetization (up to 6 T) for anisotropy and to achieve strong and complete magnetization. The 6-axis soft PME is a perfect replacement for powerful hard magnets. We also do a thorough analysis of this material by infusing it in different types of silicone and as a result the most suitable combinations are selected. Performance tests show that the sensor can detect minute forces like 0.1 N. Moreover, the hysteresis test is carried out and the hysteresis error for our skin is found to be only 1.402%. An overloading test is also performed by loading the skin up to 64 N to check the robustness. In conclusion, the skin can produce reliable Triaxial force measurements and we present two models of it for smaller and large force range measurements respectively.

The bones, muscles, tendons and connective tissues form a continuous tension network throughout human body. This heterogeneous mixture presents the characteristics of tensegrity, providing the body with structurally integrity, stability and flexibility. Inspired by this, this paper proposes a novel soft robotic gripper based on tensegrity structures. Firstly, the design and working principle of the tensegrity-based robotic gripper is introduced, which is composed of a series of discrete rigid segments connected with tensegrity joints by means of tensional cables. Then, the kinematics of the robotic gripper is analyzed using force density method to obtain the relationship between the pose of the gripper and the tension of cables. Finally experiments on the developed prototype demonstrates that the robotic gripper is able to grasp various objects of different sizes, shapes, and materials. Additional desirable properties are derived from using tensegrity structures in the robotic gripper: light weight, high compliance, inherent safety, low cost, and waterproof and dustproof performance. It is suggested that tensegrity structures have great potential to be an effective alternative to the development of soft robotic grippers.

Legged robots have opened their way to more stable and practical mobile robot applications. However, their locomotion strategies are limited to similar patterns, and dynamic running at high speed still is not successfully realized. One of the key technology required for the realization of the dynamic running of the legged robot is to estimate the ground contact force and control it in real-time. This paper tackles this problem in two ways: the derivation of the observer algorithm based on the leg dynamics and the simplification of the observer design using the Rotating Workspace motion description. To this end, two novel coordinate systems are introduced to describe the joint space motion and the workspace differently, and the ground contact force observer is designed in the novel coordinate systems. The performance of the proposed observer is verified through experimental results.

A resilient multi-vehicle system cooperatively performs tasks by exchanging information, detecting, and removing cyber attacks that have the intent of hijacking or diminishing performance of the entire system. In this paper, we propose a framework to: i) detect and isolate misbehaving vehicles in the network, and ii) securely encrypt information among the network to alert and attract nearby vehicles toward points of interest in the environment without explicitly broadcasting safety-critical information. To accomplish these goals, we lever-age a decentralized virtual spring-damper mesh physics model for formation control on each vehicle. To discover inconsistent behavior of any vehicle in the network, we consider an approach that monitors for changes in sign behavior of an inter-vehicle residual that does not match with an expectation. Similarly, to disguise important information and trigger vehicles to switch to different behaviors, we leverage side-channel information on the state of the vehicles and characterize a hidden spring-damper signature model detectable by neighbor vehicles. Our framework is demonstrated in simulation and experiments on formations of unmanned ground vehicles (UGVs) in the presence of malicious man-in-the-middle communication attacks.

Engineering humanoid robot hands with the ability to dexterously grasp objects of different sizes, shapes, mate-rial properties and weights requires sophisticated tactile sensing and intelligent controllers able to interpret sensory information and adapt contact forces with the object to achieve a stable and safe grasp. In this paper, we present a new soft humanoid hand equipped with a multimodal sensor system in each finger and a human-inspired grasp-phases controller that is able to detect the different phases of a grasping and manipulation task, adapt interaction forces with the manipulated object and balance the force distribution in both precision and power grasps based on tactile feedback. To evaluate the controller, we conducted experiments with the hand on the humanoid robot ARMAR-6 and 31 different soft and rigid everyday objects and food items with weights ranging from 4.8 g of a paper cup to 1133.8 g of a bottle, different shapes and material properties. The results show that grasping force can be reduced by 65% compared to a naive grasping approach using maximum force for grasping and manipulating both fragile objects without destruction as well as heavy objects.

Multi-robot task allocation (MRTA) problems involve optimizing the allocation of robots to tasks. MRTA problems are known to be challenging when tasks require multiple robots and the team is composed of heterogeneous robots. These challenges are further exacerbated when we need to account for uncertainties encountered in the real-world. In this work, we address coalition formation in heterogeneous multi-robot teams with uncertain capabilities. We specifically focus on tasks that require coalitions to collectively satisfy certain minimum requirements. Existing approaches to uncertainty-aware task allocation either maximize expected pay-off (risk-neutral approaches) or improve worst-case or near-worst-case outcomes (risk-averse approaches). Within the context of our problem, we demonstrate the inherent limitations of unilaterally ignoring or avoiding risk and show that these approaches can in fact reduce the probability of satisfying task requirements. Inspired by models that explain foraging behaviors in animals, we develop a risk-adaptive approach to task allocation. Our approach adaptively switches between risk-averse and risk-seeking behavior in order to maximize the probability of satisfying task requirements. Comprehensive numerical experiments conclusively demonstrate that our risk-adaptive approach outperforms risk-neutral and risk-averse approaches. We also demonstrate the effectiveness of our approach using a simulated multi-robot emergency response scenario.

This paper presents a mobile UVC disinfection robot designed to mitigate the threat of airborne and surface pathogens. Our system comprises a mobile robot base, a custom UVC lamp assembly, and algorithms for autonomous navigation and path planning. We present a model of UVC disinfection and dosage of UVC light delivered by the mobile robot. We also discuss challenges and prototyping decisions for rapid deployment of the robot during the COVID-19 pandemic. Experimental results summarize a long-term deployment at The Greater Boston Food Bank, where the robot delivers (nightly) UVC dosages of at least 10 mJ/cm2 to a 4000 ft2 area in under 30 minutes. These dosages are capable of neutralizing 99% of coronaviruses, including SARS-CoV-2, on surfaces and in airborne particles. Further simulations present how this mobile UVC disinfection robot may be extended to classic problems in robotic path planning and adaptive multi-robot coverage control.

Polar linkages have two degrees-of-freedom (DOF) where one input joint angle controls the length of a radial segment while another controls its angle. Considering a theoretical planar robot model, this mapping between joint angles to output motions can be shown to be energetically advantageous over the ubiquitous two-revolute linkage. Since a polar linkage’s typical construction involves a moving prismatic joint, it is cumbersome to implement alongside rotary electromagnetic actuators offsetting any advantage. In this paper, we present a procedure for designing polar linkages using only revolute joints. The procedure starts with a pre-existing single DOF straight line linkage and then finds the dimensions of a three-link attachment to produce the second DOF. In the end, the straight line linkage actuates the polar length and the attachment actuates the polar angle. The design process is framed under optimization with an objective that is both polynomial and invariant to the number of discretization points. This enables the techniques of numerical continuation to efficiently find complete sets of minima. We demonstrate our procedure with an example in which multiple minima are found including the global minimum. This computed design solution is then fabricated in order to validate the designed kinematics.

In the last decade, autonomous vertical take-off and landing (VTOL) vehicles have become increasingly important as they lower mission costs thanks to their re-usability. However, their development is complex, rendering even the basic experimental validation of the required advanced guidance and control (G & C) algorithms prohibitively time-consuming and costly. In this paper, we present the design of an inexpensive small-scale VTOL platform that can be built from off-the-shelf components for less than 1000 USD. The vehicle design mimics the first stage of a reusable launcher, making it a perfect test-bed for G & C algorithms. To control the vehicle during ascent and descent, we propose a real-time optimization-based G & C algorithm. The key features are a real-time minimum fuel and free-final-time optimal guidance combined with an offset-free tracking model predictive position controller. The vehicle hardware design and the G & C algorithm are experimentally validated both indoors and outdoor, showing reliable operation in a fully autonomous fashion with all computations done on-board and in real-time.

Today, many inspection domains utilize the benefits of drones to monitor and inspect infrastructure in an efficient manner. The energy grid is challenged by frequent and thorough inspection to stay operational. So far, drones have already been introduced to solve this challenge. However, the inspection drone still requires manual control and subsequent human examination of the captured photos and videos. This inspection process comes with a high inspection cost and is susceptible to human errors in fault finding. The proposed system builds on top of the authors’ previous research to develop and verify an integrated drone system for autonomous power line inspection, enabling functioning mechanics through pneumatics, extension of operation time through energy harvesting, Artificial Intelligent (AI) fault detection, and system autonomy using navigational algorithms. An advanced drone system has been designed and manufactured for the mission, with the results demonstrating the capability to perform an autonomous inspection mission in conditions up to 30 kts wind speeds, being additionally able to detect faults in real-time at a high rate during drone motion. Furthermore, we demonstrate the ability to recharge the drone battery within 2.4 hours.

One way to create a quadruped galloping robot from scratch is to design a brick-shaped body and utilize relatively simple open-chain leg mechanisms controlled with relatively complex control algorithms. Alternatively, we can look at how nature solved the same task designing fast mammals such as cheetah, and by means of morphological computation, we can design a complex mechanical system that has much of the desired behavior within inherent dynamics and only a little control effort is needed to stabilize or augment the motion.In this paper, we have analyzed a real cheetah motion using video tracking and looked for a way to match the dynamic model parameters of the real cheetah with a galloping robot with an elastic spine. We believe the elastic spine is the essential feature for a fast-running energy-efficient galloping robot. Within this paper, we are focused on the flying stage when the elastic spine affects the motion of the robot’s front and rear bodies. We have found how to optimize mass distribution and elasticity in the spine in order to get the cheetah-like galloping motion of a quadruped robot.

Brain-computer interface (BCI)-based robotic telepresence provides an opportunity for people with disabilities to control robots remotely without any actual physical movement. However, traditional BCI systems usually require the user to select the navigation direction from visual stimuli in a fixed background, which makes it difficult to control the robot in a dynamic environment during the locomotion. In this paper, a novel SSVEP-based BCI stimuli system is proposed for robotic telepresence. The novel system utilized the live video streamed from the robot onboard camera as the input. By altering and flickering the detected objects in the scene with different frequencies predefined based on their relative positions on the screen, the robot can be navigated based on the user’s attention in a dynamic manner. In order to better differentiate multiple objects (more than the number of frequencies predefined), the task-related component analysis (TRCA) model was trained with a priori offline experimental data to select the front objects with priority. Experiments were conducted to validate the proposed system. Using the system, four human subjects are able to control a humanoid robot to navigate through multiple objects to reach the desired goal. The success rate reaches 87.5% in average.