A road surface reconstruction dataset for autonomous driving.

Recent developments in intelligent robot systems, especially autonomous vehicles, put forward higher requirements for safety and comfort. Road conditions are crucial factors affecting the comprehensive performance of ground vehicles. Nonetheless, existing environment perception datasets for autonomous driving lack attention to road surface areas. In this paper, we introduce the road surface reconstruction dataset, providing multi-modal, high-resolution, and high-precision data collected by real-vehicle platform in diverse driving conditions. It covers common road types containing approximately 16,000 pairs of stereo images, point clouds, and ground-truth depth/disparity maps, with accurate data processing pipelines to ensure its quality. Preliminary evaluations reveal the effectiveness of our dataset and the challenge of the task, underscoring substantial opportunities of it as a valuable resource for advancing computer vision techniques. The reconstructed road structure and texture contribute to the analysis and prediction of vehicle responses for motion planning and control systems.

Data-Driven Linear Quadratic Optimization for Controller Synthesis With Structural Constraints.

For various typical cases and situations where the formulation results in an optimal control problem, the linear quadratic regulator (LQR) approach and its variants continue to be highly attractive. In certain scenarios, it can happen that some prescribed structural constraints on the gain matrix would arise. Consequently then, the algebraic Riccati equation (ARE) is no longer applicable in a straightforward way to obtain the optimal solution. This work presents a rather effective alternative optimization approach based on gradient projection. The utilized gradient is obtained through a data-driven methodology, and then projected onto applicable constrained hyperplanes. Essentially, this projection gradient determines a direction of progression and computation for the gain matrix update with a decreasing functional cost; and then the gain matrix is further refined in an iterative framework. With this formulation, a data-driven optimization algorithm is summarized for controller synthesis with structural constraints. This data-driven approach has the key advantage that it avoids the necessity of precise modeling which is always required in the classical model-based counterpart; and thus the approach can additionally accommodate various model uncertainties. Illustrative examples are also provided in the work to validate the theoretical results.

Sparse R-CNN: An End-to-End Framework for Object Detection.

Object detection serves as one of most fundamental computer vision tasks. Existing works on object detection heavily rely on dense object candidates, such as k anchor boxes pre-defined on all grids of an image feature map of size H×W. In this paper, we present Sparse R-CNN, a very simple and sparse method for object detection in images. In our method, a fixed sparse set of learned object proposals ( N in total) are provided to the object recognition head to perform classification and localization. By replacing HWk (up to hundreds of thousands) hand-designed object candidates with N (e.g., 100) learnable proposals, Sparse R-CNN makes all efforts related to object candidates design and one-to-many label assignment completely obsolete. More importantly, Sparse R-CNN directly outputs predictions without the non-maximum suppression (NMS) post-processing procedure. Thus, it establishes an end-to-end object detection framework. Sparse R-CNN demonstrates highly competitive accuracy, run-time and training convergence performance with the well-established detector baselines on the challenging COCO dataset and CrowdHuman dataset. We hope that our work can inspire re-thinking the convention of dense prior in object detectors and designing new high-performance detectors.

Real-time time-optimal continuous multi-axis trajectory planning using the trajectory index coordination method.

Real-time time-optimal trajectory planning exists in a wide range of applications such as computer numerical control (CNC) manufacturing, robotics and autonomous vehicles. Generally, the methods to generate time-optimal trajectory can be categorized as non-real-time methods and real-time methods. Non-real-time methods such as direct optimization method tend to generate time-optimal trajectory through nonlinear or linear programming while it is computationally prohibitive for high frequency real-time applications. Current real-time methods are computationally efficient but either deal with the sparse waypoint trajectories or sacrifice the time optimality a lot. This paper innovatively proposed a time-optimal switching trajectory index coordination (TOS-TIC) framework to solve the real-time time-optimal planning problem for continuous multi-axis trajectories. The proposed method is able to generate time-optimal trajectory for continuous geometric paths while considering the axial velocity and acceleration constraints. The time-optimality of the trajectory planned by TOS-TIC is nearly the same as the offline planned optimal results. Meanwhile, the proposed method is computationally efficient for even 5kHz real-time applications. The main idea of TOS-TIC is coordinating several one-axis time-optimal switching controls to generate a modified control that decreases the state deviation from the desired trajectory. Several comparative experiments are carried out on an industrial biaxial linear motor stage. And the experimental results consistently verify that the proposed TOS-TIC real-time planner generates faster trajectory compared with the real-time lookahead method. In addition, the trajectory running time and final tracking error of the proposed method are nearly the same as the offline direct optimization method.

Autonomous Ground Vehicle Lane-Keeping LPV Model-Based Control: Dual-Rate State Estimation and Comparison of Different Real-Time Control Strategies.

In this contribution, we suggest two proposals to achieve fast, real-time lane-keeping control for Autonomous Ground Vehicles (AGVs). The goal of lane-keeping is to orient and keep the vehicle within a given reference path using the front wheel steering angle as the control action for a specific longitudinal velocity. While nonlinear models can describe the lateral dynamics of the vehicle in an accurate manner, they might lead to difficulties when computing some control laws such as Model Predictive Control (MPC) in real time. Therefore, our first proposal is to use a Linear Parameter Varying (LPV) model to describe the AGV's lateral dynamics, as a trade-off between computational complexity and model accuracy. Additionally, AGV sensors typically work at different measurement acquisition frequencies so that Kalman Filters (KFs) are usually needed for sensor fusion. Our second proposal is to use a Dual-Rate Extended Kalman Filter (DREFKF) to alleviate the cost of updating the internal state of the filter. To check the validity of our proposals, an LPV model-based control strategy is compared in simulations over a circuit path to another reduced computational complexity control strategy, the Inverse Kinematic Bicycle model (IKIBI), in the presence of process and measurement Gaussian noise. The LPV-MPC controller is shown to provide a more accurate lane-keeping behavior than an IKIBI control strategy. Finally, it is seen that Dual-Rate Extended Kalman Filters (DREKFs) constitute an interesting tool for providing fast vehicle state estimation in an AGV lane-keeping application.

Design of a Passive, Variable Stiffness Exoskeleton for Triceps Deficiency Mitigation.

Individuals with neurological impairment, particularly those with cervical level spinal cord injuries (SCI), often have difficulty with daily tasks due to triceps weakness or total loss of function. More demanding tasks, such as sit-skiing, may be rendered impossible due to their extreme strength demands. Design of exoskeletons that address this issue by providing supplemental strength in arm extension is an active field of research but commercial devices are not yet available for use. Most current designs employ electric motors that necessitate the addition of bulky power sources and extraneous wiring, rendering the devices impractical in daily life. The possibility of powering an upper extremity exoskeleton passively has been explored, but to date, none have delivered sufficient function or strength to provide useful assistance for sit-skiing. We seek to rectify this with the design of a passively actuated exoskeletal arm brace capable of operating in two, adjustable-strength modes: one for low level gravity compensation to aid in active range of motion, and the other for more stringent weight bearing activities. The mechanism developed through this paper allows for an affordable, lightweight, modular device that can be adjusted and customized for the needs of each individual patient.

Development of a knee joint assist suit with a velocity-based mechanical safety device: Control method and experiments

The aim of this study is to examine the proposed control method of the assist suit with a Velocity-Based Mechanical Safety Device (VBMSD) for patients with difficulty moving their lower legs by themselves. The proposed control method for the assist suit assists the patients as if the patients move their knee joint under zero gravity. A physical simulation model is used to examine whether the gravitational torque of the subject's lower leg and foot was canceled by the torque generated by the assist suit. Experimental results indicated that the gravitational torque of the subject's lower leg and foot is canceled by the torque generated by the assist suit. The control of the assist suit was not adversely influenced by the VBMSD. That is, the VBMSD did not prevent the control of the assist suit. The proposed control method makes the assist suit assist the patient in moving his/her knee joint in a zero gravity-like environment. However, a weight of 3 kg was used instead of an actual patient in the experiment. Experiments with actual patients should be conducted to verify the effectiveness of the proposed control method in clinical use. Furthermore, it will be necessary to take into consideration the patients' general conditions and symptoms

No disponible

Hybrid Control Interface of a Semi-soft Assistive Glove for People with Spinal Cord Injuries.

Active assistive devices have been designed to augment the hand grasping capabilities of individuals with spinal cord injuries (SCI). An intuitive bio-signal of wrist extension has been utilized in the device control, which imitates the passive grasping effect of tenodesis. However, controlling these devices in this manner limits the wrist joint motion while grasping. This paper presents a novel hybrid control interface and corresponding algorithms (i.e., a hybrid control method) of the Semi-soft Assistive Glove (SAG) developed for individuals with C6/C7-SCI. The secondary control interface is implemented to enable/disable the grasp trigger signal generated by the primary interface detecting the wrist extension. A simulation study reveals that the hybrid control method can facilitate grasping situations faced in daily activities. Empirical results with three healthy subjects suggest that the proposed method can assist the user to reach and grasp objects with the SAG naturally.

A Remote Control Strategy for an Autonomous Vehicle with Slow Sensor Using Kalman Filtering and Dual-Rate Control.

This work presents a novel remote control solution for an Autonomous Vehicle (AV), where the system structure is split into two sides. Both sides are assumed to be synchronized and linked through a communication network, which introduces time-varying delays and packet disorder. An Extended Kalman Filter (EKF) is used to cope with the non-linearities that appear in the global model of the AV. The EKF fuses the data provided by the sensing devices of the AV in order to estimate the AV state, reducing the noise effect. Additionally, the EKF includes an h-step-ahead state prediction stage, which, together with the consideration of a packet-based control strategy, enables facing the network-induced delays. Since the AV position is provided by a camera, which is a slow sensing device, a dual-rate controller is required to achieve certain desired (nominal) dynamic control performance. The use of a dual-rate control framework additionally enables saving network bandwidth and deals with packet disorder. As the path-tracking control algorithm, pure pursuit is used. Application results show that, despite existing communication problems and slow-rate measurements, the AV is able to track the desired path, keeping the nominal control performance.

A sEMG Classification Framework with Less Training Data.

Supervised machine learning algorithms, such as Artificial Neural Network (ANN), have been applied to surface electromyograph (sEMG) to classify user's muscular states. This paper introduces a novel framework to design a binary sEMG classifier to distinguish if the user performs a repetitive motion with a dumbbell. This framework enables to reduce the number of tasks required for collecting training data as it utilizes prior knowledge of sEMG. The performance of the proposed classifier is validated experimentally. Experimental results show that the proposed framework enables the design of a classifier which distinguishes the user's state with a 95.7% success rate. This accuracy is comparable to an accuracy of ANN classifier (99.6%), but with less training data. Under the identical training conditions, the accuracy of the proposed framework outperforms the ANN classifier whose accuracy drops to 65.6%.

Clinical impact of gait training enhanced with visual kinematic biofeedback: Patients with Parkinson's disease and patients stable post stroke.

As the world's population ages, falls, physical inactivity, decreased attention and impairments in balance and gait arise as a consequence of decreased sensation, weakness, trauma and degenerative disease. Progressive balance and gait training can facilitate postural righting, safe ambulation and community participation. This small randomized clinical trial evaluated if visual and kinematic feedback provided during supervised gait training would interfere or enhance mobility, endurance, balance, strength and flexibility in older individuals greater than one year post stroke (Gobbi et al., 2009) or Parkinson's disease (PD) (Gobbi et al., 2009). Twenty-four individuals consented to participate. The participants were stratified by diagnosis and randomly assigned to a control (usual gait training (Gobbi et al., 2009)) or an experimental group (usual gait training plus kinematic feedback (Gobbi et al., 2009)). At baseline and 6 weeks post training (18 h), subjects completed standardized tests (mobility, balance, strength, range of motion). Gains were described across all subjects, by treatment group and by diagnosis. Then they were compared for significance using nonparametric statistics. Twenty-three subjects completed the study with no adverse events. Across all subjects, by diagnosis (stroke and PD) and by training group (control and experimental), there were significant gains in mobility (gait speed, step length, endurance, and quality), balance (Berg Balance), range of motion and strength. There were no significant differences in the gain scores between the control and experimental groups. Subjects chronic post stroke made greater strength gains on the affected side than subjects with PD but otherwise there were no significant differences. In summary, during supervised gait training, dynamic visual kinematic feedback from wireless pressure and motion sensors had similar, positive effects as verbal, therapist feedback. A wireless kinematic feedback system could be used at home, to provide feedback and motivation for self correction of gait while simultaneously providing data to the therapist (at a distance).

Initial investigation into the effect of an Active/Passive exoskeleton on hammer curl performance in healthy subjects.

Assistive robotic devices are traditionally constrained by their power source. Entirely passive devices exist, but are limited by their fixed mechanical parameters. This work introduces a new device that can provide active and passive assistance. This device provides assistance in a passive mode, but retains the actively change this passive response. This paper examines the effect different passive parameter settings have on healthy subjects performing hammer curls. Passive parameter settings to either increase or decrease the number of curls a subject could perform were found. An average increase of 84% or a decrease of 33% in curls was produced by varying the passive parameters. These effects were seen across all six subjects. This indicates that there is potential for the Active/Passive framework to provide lightweight, energy efficient assistance.

A mobile gait monitoring system for abnormal gait diagnosis and rehabilitation: a pilot study for Parkinson disease patients.

Conventional gait rehabilitation treatment does not provide quantitative information on abnormal gait kinematics, and the match of the intervention strategy to the underlying clinical presentation may be limited by clinical expertise and experience. Also the effect of rehabilitation treatment may be reduced as the rehabilitation treatment is achieved only in a clinical setting. In this paper, a mobile gait monitoring system (MGMS) is proposed for the diagnosis of abnormal gait and rehabilitation. The proposed MGMS consists of Smart Shoes and a microsignal processor with a touch screen display. It monitors patients' gait by observing the ground reaction force (GRF) and the center of GRF, and analyzes the gait abnormality. Since visual feedback about patients' GRFs and normal GRF patterns are provided by the MGMS, patients can practice the rehabilitation treatment by trying to follow the normal GRF patterns without restriction of time and place. The gait abnormality proposed in this paper is defined by the deviation between the patient's GRFs and normal GRF patterns, which are constructed as GRF bands. The effectiveness of the proposed gait analysis methods with the MGMS has been verified by preliminary trials with patients suffering from gait disorders.