ICRA 2022 论文列表

Slipping may cause a vehicle out of control with serious accident potential. However, a kind of car slipping named “drifting” can be seen in professional contests. So, it is reasonable to apply drift maneuvers in autonomous driving. This article proposes a controller for the particular driving skill—drifting based on MPC(Model Predictive Control). Firstly, we analyze drift cornering mechanisms of the 3-state vehicle model and linearize the model at a drifting equilibrium point to build a fast controller. Then, the predicted lateral errors are brought into the MPC cost function to realize trajectory tracking. The algorithm makes use of the expert strategy in car drifting to adjust the steering angle and longitudinal drive force according to lateral errors during the trajectory following. Finally, it is experimentally verified that the proposed controller can track the reference trajectory in real time while maintaining the drifting state.

ComOpT is an open-source research tool for coverage-driven testing of autonomous driving systems, focusing on planning and control. Starting with (i) a meta-model characterizing discrete conditions to be considered and (ii) constraints specifying the impossibility of certain combinations, ComOpT first generates constraint-feasible abstract scenarios while maximally increasing the coverage of k-way combinatorial testing. Each abstract scenario can be viewed as a conceptual equivalence class, which is then instantiated into multiple concrete scenarios by (1) randomly picking one local map that fulfills the specified geographical condition, and (2) assigning all actors accordingly with parameters within the range. Finally, ComOpT evaluates each concrete scenario against a set of KPIs and performs local scenario variation via spawning a new agent that might lead to a collision at designated points. We use ComOpT to test the Apollo 6 autonomous driving software stack. ComOpT can generate highly diversified scenarios with limited test budgets while uncovering problematic situations such as inabilities to make simple right turns, uncomfortable accelerations, and dangerous driving patterns. ComOpT participated in the 2021 IEEE AI Autonomous Vehicle Testing Challenge and won first place among more than 110 contending teams.

When multiple quadrotors fly in a cluttered environment, collision-free flight must be assured. In this paper, we propose a novel elastic safety clearance based model predictive control (ESC-MPC) for multiple maneuverable quadrotors to avoid collisions in the presence of disturbance. This is accomplished through leveraging tube based model predictive control to maintain the quadrotor in a tube of trajectories. Exponential control barrier function (ECBF) is integrated to realize the elastic safety clearance mechanism which offers a dynamic safety margin in maneuverable flight. We validate the superiority of our approach with laboratory experiments.

In this paper, we deal with the problem of creating globally consistent pose graphs in a centralized multi-robot SLAM framework. For each robot to act autonomously, individual onboard pose estimates and maps are maintained, which are then communicated to a central server to build an optimized global map. However, inconsistencies between onboard and server estimates can occur due to onboard odometry drift or failure. Furthermore, robots do not benefit from the collaborative map if the server provides no feedback in a computationally tractable and bandwidth-efficient manner. Motivated by this challenge, this paper proposes a novel collaborative mapping framework to enable accurate global mapping among robots and server. In particular, structural differences between robot and server graphs are exploited at different spatial scales using graph spectral analysis to generate necessary constraints for the individual robot pose graphs. The proposed approach is thoroughly analyzed and validated using several real-world multi-robot field deployments where we show improvements of the onboard system up to 90%.

Retinal surgery is a complex medical procedure that requires exceptional expertise and dexterity. For this purpose, several robotic platforms are currently under development to enable or improve the outcome of microsurgical tasks. Since the control of such robots is often designed for navigation inside the eye in proximity to the retina, successful trocar docking and insertion of the instrument into the eye represents an additional cognitive effort, and is therefore one of the open challenges in robotic retinal surgery. For this purpose, we present a platform for autonomous trocar docking that combines computer vision and a robotic setup. Inspired by the Cuban Colibri (hummingbird) aligning its beak to a flower using only vision, we mount a camera onto the endeffector of a robotic system. By estimating the position and pose of the trocar, the robot is able to autonomously align and navigate the instrument towards the Trocar Entry Point (TEP) and finally perform the insertion. Our experiments show that the proposed method is able to accurately estimate the position and pose of the trocar and achieve repeatable autonomous docking. The aim of this work is to reduce the complexity of the robotic setup prior to the surgical task and therefore, increase the intuitiveness of the system integration into clinical workflow.

The food packaging industry handles an immense variety of food products with wide-ranging shapes and sizes, even within one kind of food. Menus are also diverse and change frequently, making automation of pick-and-place difficult. A popular approach to bin-picking is to first identify each piece of food in the tray by using an instance segmentation method. However, human annotations to train these methods are unreli-able and error-prone since foods are packed close together with unclear boundaries and visual similarity making separation of pieces difficult. To address this problem, we propose a method that trains purely on synthetic data and successfully transfers to the real world using sim2real methods by creating datasets of filled food trays using high-quality 3d models of real pieces of food for the training instance segmentation models. Another concern is that foods are easily damaged during grasping. We address this by introducing two additional methods- a novel adaptive finger mechanism to passively retract when a collision occurs, and a method to filter grasps that are likely to cause damage to neighbouring pieces of food during a grasp. We demonstrate the effectiveness of the proposed method on several kinds of real foods.

There is a gap in holistic urban scene understanding between multi-modal datasets for segmentation and object detection on the one hand and traffic light datasets on the other hand. The role of traffic lights in the former is not labelled, making it difficult to use them for higher-level tasks and leave critical information of an intersection scene blank. Including traffic lights from traffic light specific datasets into the comprehensive semantic data introduces a penalty from the domain shift. We close this gap by providing semantically annotated traffic lights for the Cityscapes dataset. We demonstrate the domain shift penalty by using a traffic light dataset from a similar domain and show superior performance on data labelled in the original domain. We demonstrate an application by training a real-time capable network for semantic segmentation and object detection which can now additionally make sense of traffic lights, delivering an F1- Score of 66.4% on the important class of traffic lights relevant to the ego vehicle. The network is made publicly available at https://github.com/joeda/NNAD and the data at https://github.com/KIT-MRT/cityscapes-t1.

We present CineMPC, an algorithm to autonomously control a UAV-borne video camera in a nonlinear Model Predicted Control (MPC) loop. CineMPC controls both the position and orientation of the camera-the camera extrinsics-as well as the lens focal length, focal distance, and aperture-the camera intrinsics. While some existing solutions autonomously control the position and orientation of the camera, no existing solutions also control the intrinsic parameters, which are essential tools for rich cinematographic expression. The intrinsic parameters control the parts of the scene that are focused or blurred, the viewers' perception of depth in the scene and the position of the targets in the image. CineMPC closes the loop from camera images to UAV trajectory and lens parameters in order to follow the desired relative trajectory and image composition as the targets move through the scene. Experiments using a photo-realistic environment demon-strate the capabilities of the proposed control framework to successfully achieve a full array of cinematographic effects not possible without full camera control.

Practical robot designs must strike a compromise between fabrication/manufacture cost and anticipated execution performance. Compared to parsimonious designs, more capable (and hence more expensive) robots generally achieve their ends with greater efficiency. This paper examines how the roboticist might explore the space of designs to gain an understanding of such trade-offs. We focus, specifically, on design choices that alter the set of actions available to the robot, and model those actions as involving uncertainty. We consider planning problems under the Markov Decision Process (MDP) model, which leads us to examine how to relate the cost of some design to the expected cost of an execution for the optimal policies feasible with that design. The complexity of this problem ─expressed via hardness in the fixed parameter tractability sense─depends on the number of actions to choose from. When that number is not negligible, we give a novel representation and an algorithm utilizing that structure that allows useful savings over naïve enumeration.

A key problem in robotic locomotion is in finding optimal shape changes to effectively displace systems through the world. Variational techniques for gait optimization require estimates of body displacement per gait cycle; however, these estimates introduce error due to unincluded high order terms. In this paper, we formulate existing estimates for displacement, and describe the contribution of low order terms to these estimates. We additionally describe the magnitude of higher (third) order effects, and identify that choice of body coordinate, gait diameter, and starting phase influence these effects. We demonstrate that variation of such parameters on two example systems (the differential drive car and Purcell swimmer) effectively manages third order contributions.

This paper presents a modeling and control frame-work for multibody flying robots subject to non-negligible aero-dynamic forces acting on the centroidal dynamics. First, aero-dynamic forces are calculated during robot flight in different operating conditions by means of Computational Fluid Dynamics (CFD) analysis. Then, analytical models of the aerodynamics coefficients are generated from the dataset collected with CFD analysis. The obtained simplified aerodynamic model is also used to improve the flying robot control design. We present two control strategies: compensating for the aerodynamic effects via feedback linearization and enforcing the controller robustness with gain-scheduling. Simulation results on the jet-powered humanoid robot iRonCub validate the proposed approach.

This paper studies the complex task of simultaneous multi-object 3D reconstruction, 6D pose and size estimation from a single-view RGB-D observation. In contrast to instance- level pose estimation, we focus on a more challenging problem where CAD models are not available at inference time. Existing approaches mainly follow a complex multi-stage pipeline which first localizes and detects each object instance in the image and then regresses to either their 3D meshes or 6D poses. These approaches suffer from high-computational cost and low performance in complex multi-object scenarios, where occlusions can be present. Hence, we present a simple one- stage approach to predict both the 3D shape and estimate the 6D pose and size jointly in a bounding-box free manner. In particular, our method treats object instances as spatial centers where each center denotes the complete shape of an object along with its 6D pose and size. Through this per- pixel representation, our approach can reconstruct in real- time (40 FPS) multiple novel object instances and predict their 6D pose and sizes in a single-forward pass. Through extensive experiments, we demonstrate that our approach significantly outperforms all shape completion and categorical 6D pose and size estimation baselines on multi-object ShapeNet and NOCS datasets respectively with a 12.6% absolute improvement in mAP for 6D pose for novel real-world object instances.

Accurately predicting the possible behaviors of traffic participants is an essential capability for autonomous vehicles. Since autonomous vehicles need to navigate in dynamically changing environments, they are expected to make accurate predictions regardless of where they are and what driving circumstances they encountered. Therefore, generalization capability to unseen domains is crucial for prediction models when autonomous vehicles are deployed in the real world. In this paper, we aim to address the domain generalization problem for vehicle intention prediction tasks and a causal-based time series domain generalization (CTSDG) model is proposed. We construct a structural causal model for vehicle intention prediction tasks to learn an invariant representation of input driving data for domain generalization. We further integrate a recurrent latent variable model into our structural causal model to better capture temporal latent dependencies from time-series input data. The effectiveness of our approach is evaluated via real-world driving data. We demonstrate that our proposed method has consistent improvement on prediction accuracy compared to other state-of-the-art domain generalization and behavior prediction methods.

As robots move toward more complex environments, imbuing them with a sense of touch similar to humans becomes increasingly important. To fulfill that goal, there has been significant research conducted in the past few decades to develop a tactile sensor that matches human level touch capabilities. Recently, the progress in capacitive touch screens has made capacitive sensing a very appealing option for such a sensor, and therefore many research groups have proposed novel designs of tactile sensors based on capacitive technologies. This technology has the advantage of generating a predictable sensor response with a high degree of sensitivity, but has the drawback of a limited spatial resolution. This paper shows how using mutual capacitance in combination with a microstructured dielectric can lead to a very sensitive sensor that also possesses a high spatial resolution. The response of the sensor in relation to its various components is explored in order to fully comprehend the physical principles of the sensing mechanism and generate a predictable output.

The promising automation of flexible surgical instruments and robots is impeded by the lack of sensory means, which allow for sensing of an instrument's position to the surrounding tissue. This work presents a novel sensory method utilizing capacitive proximity sensing to derive a relative localization of a flexible instrument inside a hollow organ. The method is evaluated by exemplary integration of a sensor in a commercial gastroendoscope and accuracy analysis using a high precision robot. The results show an accuracy of distance sensing from a medical phantom's center of 2%. The method is also evaluated for the irregularly shaped surrounding of ex-vivo tissue in a dynamic scenario. This promising approach holds potential for transfer to clinical scenarios and for further development towards pose estimation of flexible surgical robots and shape sensing of a minimally invasive environment.

Aerial robots have been widely used as sensor carrying platforms in a wide range of application, mainly because using this type of systems for physical interaction seems to be an unsuitable operation. This is not only due to the risk of collision and damage of the platform, but also because it is unclear whether a consumer-grade UAV can withstand physical contact with the environment. In this paper, we address the issue of performing physical interaction with the environment by a multirotor UAV implementing a basic cascaded position-attitude controller, typical of most of consumer-grade multirotor systems. Precisely, we identify mathematically the boundaries where the system can safely be used to perform physical interaction with the environment. The theoretical approach is finally validated through experiments showing that physical contact can only be achieved within a predefined region of control inputs.

Recently, vehicle re-identification methods based on deep learning constitute remarkable achievement. However, this achievement requires large-scale and well-annotated datasets. In constructing the dataset, assigning globally available identities (Ids) to vehicles captured from a great number of cameras is labour-intensive, because it needs to consider their subtle appearance differences or viewpoint variations. In this paper, we propose camera-tracklet-aware contrastive learning (CTACL) using the multi-camera tracklet information without vehicle identity labels. The proposed CTACL divides an unlabelled domain, i.e., entire vehicle images, into multiple camera-level subdomains and conducts contrastive learning within and beyond the subdomains. The positive and negative samples for contrastive learning are defined using tracklet Ids of each camera. Additionally, the domain adaptation across camera networks is introduced to improve the generalisation performance of learnt representations and alleviate the performance degradation resulted from the domain gap between the subdomains. We demonstrate the effectiveness of our approach on video-based and image-based vehicle Re-ID datasets. Experimental results show that the proposed method outperforms the recent state-of-the-art unsupervised vehicle Re-ID methods. The source code for this paper is publicly available on https://github.com/andreYoo/CTAM-CTACL-VVReID.git.

Task-relevant grasping is critical for industrial assembly, where downstream manipulation tasks constrain the set of valid grasps. Learning how to perform this task, however, is challenging, since task-relevant grasp labels are hard to define and annotate. There is also yet no consensus on proper representations for modeling or off-the-shelf tools for performing task-relevant grasps. This work proposes a framework to learn task-relevant grasping for industrial objects without the need of time-consuming real-world data collection or manual annotation. To achieve this, the entire framework is trained solely in simulation, including supervised training with synthetic label generation and self-supervised, hand-object interaction. In the context of this framework, this paper proposes a novel, object-centric canonical representation at the category level, which allows establishing dense correspondence across object instances and transferring task-relevant grasps to novel instances. Extensive experiments on task-relevant grasping of densely-cluttered industrial objects are conducted in both simulation and real-world setups, demonstrating the effectiveness of the proposed framework. Code and data are available at https://sites.google.com/view/catgrasp.

Multi-beam LiDAR sensors are increasingly used in robotics, particularly with autonomous cars for localization and perception tasks, both relying on the ability to build a precise map of the environment. For this, we propose a new real-time LiDAR-only odometry method called CT-ICP (for Continuous-Time ICP), completed into a full SLAM with a novel loop detection procedure. The core of this method, is the introduction of the combined continuity in the scan matching, and discontinuity between scans. It allows both the elastic distortion of the scan during the registration for increased precision, and the increased robustness to high frequency motions from the discontinuity. We build a complete SLAM on top of this odometry, using a fast pure LiDAR loop detection based on elevation image 2D matching, providing a pose graph with loop constraints. To show the robustness of the method, we tested it on seven datasets: KITTI, KITTI-raw, KITTI-360, KITTICARLA, ParisLuco, Newer College, and NCLT in driving and high-frequency motion scenarios. Both the CT-ICP odometry and the loop detection are made available online. CT-ICP is currently first, among those giving access to a public code, on the KITTI odometry leaderboard, with an average Relative Translation Error (RTE) of 0.59% and an average time per scan of 60ms on a CPU with a single thread.