Obstacle Avoidance Task Research Articles

Model Predictive Control of a Magnetically Guided Rolling Microrobot

In this work, we present a model predictive control (MPC) approach to the motion control of a nonholonomic rolling microrobot. The mobile microrobot is designed to follow a rotating magnetic field such that it rolls on a supporting surface in a low Reynold’s number liquid environment. The MPC framework takes the nonholonomic and kinematic constraints of the microrobot and the actuation limitations of the system into account for the task of trajectory tracking and obstacle avoidance. We verify our approach by presenting a simulation study and experimental results with an 8-coil electromagnetic actuation system.

Robot Obstacle Avoidance Learning Based on Mixture Models

We briefly surveyed the existing obstacle avoidance algorithms; then a new obstacle avoidance learning framework based on learning from demonstration (LfD) is proposed. The main idea is to imitate the obstacle avoidance mechanism of human beings, in which humans learn to make a decision based on the sensor information obtained by interacting with environment. Firstly, we endow robots with obstacle avoidance experience by teaching them to avoid obstacles in different situations. In this process, a lot of data are collected as a training set; then, to encode the training set data, which is equivalent to extracting the constraints of the task, Gaussian mixture model (GMM) is used. Secondly, a smooth obstacle-free path is generated by Gaussian mixture regression (GMR). Thirdly, a metric of imitation performance is constructed to derive a proper control policy. The proposed framework shows excellent generalization performance, which means that the robots can fulfill obstacle avoidance task efficiently in a dynamic environment. More importantly, the framework allows learning a wide variety of skills, such as grasp and manipulation work, which makes it possible to build a robot with versatile functions. Finally, simulation experiments are conducted on a Turtlebot robot to verify the validity of our algorithms.

The Effect of Gaze Position on Reaching Movements in an Obstacle Avoidance Task.

Numerous studies have addressed the issue of where people look when they perform hand movements. Yet, very little is known about how visuomotor performance is affected by fixation location. Previous studies investigating the accuracy of actions performed in visual periphery have revealed inconsistent results. While movements performed under full visual-feedback (closed-loop) seem to remain surprisingly accurate, open-loop as well as memory-guided movements usually show a distinct bias (i.e. overestimation of target eccentricity) when executed in periphery. In this study, we aimed to investigate whether gaze position affects movements that are performed under full-vision but cannot be corrected based on a direct comparison between the hand and target position. To do so, we employed a classical visuomotor reaching task in which participants were required to move their hand through a gap between two obstacles into a target area. Participants performed the task in four gaze conditions: free-viewing (no restrictions on gaze), central fixation, or fixation on one of the two obstacles. Our findings show that obstacle avoidance behaviour is moderated by fixation position. Specifically, participants tended to select movement paths that veered away from the obstacle fixated indicating that perceptual errors persist in closed-loop vision conditions if they cannot be corrected effectively based on visual feedback. Moreover, measuring the eye-movement in a free-viewing task (Experiment 2), we confirmed that naturally participants’ prefer to move their eyes and hand to the same spatial location.

Stability control during the performance of a simultaneous obstacle avoidance and auditory Stroop task.

Navigation through complex environments requires a greater degree of control and attentional resources from the central nervous system to ensure postural stability and efficient goal completion as compared to quiet standing or unobstructed walking. Furthermore, when a cognitive task is also performed in a dual-task scenario, additional resources may be required. The purpose of the current study was to examine the effects of performing a concurrent cognitive (auditory Stroop task) and complex locomotor task (with a stationary or dynamic obstacle) on frontal plane stability control quantified using a margin of stability (MOS) measure. Fourteen healthy young adults performed 40 dual-task trials (randomized in a balanced design for auditory Stroop congruency and obstacle movement). Results indicated that frontal plane MOS was greatest for the obstacle crossing step and was greater for the dynamic obstacle as compared to the stationary obstacle. Conversely, frontal plane MOS was the smallest for the pre-crossing step, indicating that this point in the obstacle stepping strategy may be the least stable. No effect of cognitive task difficulty was observed for any of the experimental conditions, providing support for a 'posture-first' strategy. These findings suggest that an increase in stability is prioritized for the obstacle crossing step, potentially at the expense of reduced stability in the step immediately preceding the obstacle. These results have implications for better understanding how the CNS controls stability at different events during the obstacle crossing strategy in a complex environment.

It is the flash which appears, the movement will follow: Investigating the relation between spatial attention and obstacle avoidance.

Obstacles are represented in the attentional landscape. However, it is currently unclear what the exclusive contribution of attention is to the avoidance response. This is because in earlier obstacle avoidance designs, it was impossible to disentangle an effect of attention from the changing features of the obstacle (e.g., its identity, size, or orientation). Conversely, any feature manipulation could be interpreted as an attentional as well as an obstacle effect on avoidance behavior. We tested the possible tuning of avoidance responses by a spatial cue in two experiments. In both experiments, spatial and nonspatial cues were separately given as go cues for an obstacle avoidance task. Participants had to reach past two obstacles in Experiment 1, and past a single obstacle in Experiment 2. We found that when the right obstacle was flashed, participants veered away more and produced more-variable trajectories over trials than in conditions with nonspatial and left spatial cues, regardless of the presence or absence of another obstacle. Therefore, we concluded that the tuning of avoidance responses can be influenced by spatial cues. Moreover, we speculated that a flashed obstacle receives more attentional weighting in the attentional landscape and prompts a stronger repulsion away from the obstacle.

Improved gait adjustments after gait adaptability training are associated with reduced attentional demands in persons with stroke

After stroke, the ability to make step adjustments during walking is reduced and requires more attention, which may cause problems during community walking. The C-Mill is an innovative treadmill augmented with visual context (e.g., obstacles and stepping targets), which was designed specifically to practice gait adaptability. The objective of this study was to determine whether C-Mill gait adaptability training can help to improve gait adjustments and associated attentional demands. Sixteen community-ambulating persons in the chronic stage of stroke (age: 54.8 ± 10.8 years) received ten sessions of C-Mill training within 5-6 weeks. Prior to and after the intervention period, participants performed an obstacle-avoidance task with and without a secondary attention-demanding auditory Stroop task to assess their ability to make gait adjustments (i.e., obstacle-avoidance success rates) as well as the associated attentional demands (i.e., Stroop success rates, stratified for pre-crossing, crossing, and post-crossing strides). Obstacle-avoidance success rates improved after C-Mill training from 52.4 ± 16.3 % at pretest to 77.0 ± 16.4 % at posttest (p < 0.001). This improvement was accompanied by greater Stroop success rates during the obstacle-crossing stride only (pretest: 62.9 ± 24.9 %, posttest: 77.5 ± 20.4 %, p = 0.006). The observed improvements in obstacle-avoidance success rates and Stroop success rates were strongly correlated (r = 0.68, p = 0.015). The ability to make gait adjustments and the associated attentional demands can be successfully targeted in persons with stroke using C-Mill training, which suggests that its underlying assumptions regarding motor control are appropriate. This study lends support and guidance for designing a randomized controlled trial to further examine the potential of C-Mill training for improving safe community ambulation after stroke.

Coordination control of multiple Euler–Lagrange systems for escorting mission

SummaryIn this paper, a motion control problem of multi‐agent systems for escorting a target is investigated by employing nonsingular fast terminal sliding mode control and adaptive control associated with kinematic control. The proposed control law is robust to model uncertainty and disturbances, and it guarantees all the agents to scatter around the target evenly and escort it with a fixed distance while avoiding obstacles (or collisions) in p‐dimensional case ( is a positive integer). Finite‐time convergence of the position errors and velocity errors is proved rigorously by a Lyapunov‐based approach and finite‐time control techniques. Simulation results in both two‐dimensional and three‐dimensional space are provided to illustrate the effectiveness and high‐precision performance of the control algorithm compared with the traditional adaptive sliding mode control, showing that all the agents can move into suitable positions on the surface of the sphere in the escort mission, and the formation can be reconfigured automatically when the obstacle (or collision) avoidance task is active. Copyright © 2014 John Wiley &amp; Sons, Ltd.

Deficits in memory-guided limb movements impair obstacle avoidance locomotion in Alzheimer's disease mouse model.

Memory function deficits induced by Alzheimer's disease (AD) are believed to be one of the causes of an increased risk of tripping in patients. Working memory contributes to accurate stepping over obstacles during locomotion, and AD-induced deficits of this memory function may lead to an increased risk of contact with obstacles. We used the triple transgenic (3xTg) mice to examine the effects of memory deficits in terms of tripping and contact with obstacles. We found that the frequency of contact of the hindlimbs during an obstacle avoidance task increased significantly in 10–13 month-old 3xTg (Old-3xTg) mice compared with control mice. However, no changes in limb kinematics during unobstructed locomotion or successful obstacle avoidance locomotion were observed in the Old-3xTg mice. Furthermore, we found that memory-based movements in stepping over an obstacle were impaired in these mice. Our findings suggest that working memory deficits as a result of AD are associated with an increased risk of tripping during locomotion.

Virtual reality-based navigation task to reveal obstacle avoidance performance in individuals with visuospatial neglect.

Persons with post-stroke visuospatial neglect (VSN) often collide with moving obstacles while walking. It is not well understood whether the collisions occur as a result of attentional-perceptual deficits caused by VSN or due to post-stroke locomotor deficits. We assessed individuals with VSN on a seated, joystick-driven obstacle avoidance task, thus eliminating the influence of locomotion. Twelve participants with VSN were tested on obstacle detection and obstacle avoidance tasks in a virtual environment that included three obstacles approaching head-on or 30 (°) contralesionally/ipsilesionally. Our results indicate that in the detection task, the contralesional and head-on obstacles were detected at closer proximities compared to the ipsilesional obstacle. For the avoidance task collisions were observed only for the contralesional and head-on obstacle approaches. For the contralesional obstacle approach, participants initiated their avoidance strategies at smaller distances from the obstacle and maintained smaller minimum distances from the obstacles. The distance at detection showed a negative association with the distance at the onset of avoidance strategy for all three obstacle approaches. We conclusion the observation of collisions with contralesional and head-on obstacles, in the absence of locomotor burden, provides evidence that attentional-perceptual deficits due to VSN, independent of post-stroke locomotor deficits, alter obstacle avoidance abilities.

Biosonar navigation above water II: exploiting mirror images.

As in vision, acoustic signals can be reflected by a smooth surface creating an acoustic mirror image. Water bodies represent the only naturally occurring horizontal and acoustically smooth surfaces. Echolocating bats flying over smooth water bodies encounter echo-acoustic mirror images of objects above the surface. Here, we combined an electrophysiological approach with a behavioral experimental paradigm to investigate whether bats can exploit echo-acoustic mirror images for navigation and how these mirrorlike echo-acoustic cues are encoded in their auditory cortex. In an obstacle-avoidance task where the obstacles could only be detected via their echo-acoustic mirror images, most bats spontaneously exploited these cues for navigation. Sonar ensonifications along the bats' flight path revealed conspicuous changes of the reflection patterns with slightly increased target strengths at relatively long echo delays corresponding to the longer acoustic paths from the mirrored obstacles. Recordings of cortical spatiotemporal response maps (STRMs) describe the tuning of a unit across the dimensions of elevation and time. The majority of cortical single and multiunits showed a special spatiotemporal pattern of excitatory areas in their STRM indicating a preference for echoes with (relative to the setup dimensions) long delays and, interestingly, from low elevations. This neural preference could effectively encode a reflection pattern as it would be perceived by an echolocating bat detecting an object mirrored from below. The current study provides both behavioral and neurophysiological evidence that echo-acoustic mirror images can be exploited by bats for obstacle avoidance. This capability effectively supports echo-acoustic navigation in highly cluttered natural habitats.

Towards photorealistic and immersive virtual-reality environments for simulated prosthetic vision: integrating recent breakthroughs in consumer hardware and software.

Simulated prosthetic vision (SPV) in normally sighted subjects is an established way of investigating the prospective efficacy of visual prosthesis designs in visually guided tasks such as mobility. To perform meaningful SPV mobility studies in computer-based environments, a credible representation of both the virtual scene to navigate and the experienced artificial vision has to be established. It is therefore prudent to make optimal use of existing hardware and software solutions when establishing a testing framework. The authors aimed at improving the realism and immersion of SPV by integrating state-of-the-art yet low-cost consumer technology. The feasibility of body motion tracking to control movement in photo-realistic virtual environments was evaluated in a pilot study. Five subjects were recruited and performed an obstacle avoidance and wayfinding task using either keyboard and mouse, gamepad or Kinect motion tracking. Walking speed and collisions were analyzed as basic measures for task performance. Kinect motion tracking resulted in lower performance as compared to classical input methods, yet results were more uniform across vision conditions. The chosen framework was successfully applied in a basic virtual task and is suited to realistically simulate real-world scenes under SPV in mobility research. Classical input peripherals remain a feasible and effective way of controlling the virtual movement. Motion tracking, despite its limitations and early state of implementation, is intuitive and can eliminate between-subject differences due to familiarity to established input methods.

Coupling Movement Primitives: Interaction With the Environment and Bimanual Tasks

The framework of dynamic movement primitives (DMPs) contains many favorable properties for the execution of robotic trajectories, such as indirect dependence on time, response to perturbations, and the ability to easily modulate the given trajectories, but the framework in its original form remains constrained to the kinematic aspect of the movement. In this paper, we bridge the gap to dynamic behavior by extending the framework with force/torque feedback. We propose and evaluate a modulation approach that allows interaction with objects and the environment. Through the proposed coupling of originally independent robotic trajectories, the approach also enables the execution of bimanual and tightly coupled cooperative tasks. We apply an iterative learning control algorithm to learn a coupling term, which is applied to the original trajectory in a feed-forward fashion and, thus, modifies the trajectory in accordance to the desired positions or external forces. A stability analysis and results of simulated and real-world experiments using two KUKA LWR arms for bimanual tasks and interaction with the environment are presented. By expanding on the framework of DMPs, we keep all the favorable properties, which is demonstrated with temporal modulation and in a two-agent obstacle avoidance task.

Line-bisectioning and obstacle avoidance: Evidence for separate strategies

Previous studies have frequently applied a combination of line-bisection tasks (in which participants indicate the middle of a line) and obstacle avoidance tasks (in which participants move their hand between two obstacles) with the aim of revealing perception–action dissociations in certain neurological disorders, such as visual form agnosia and optic ataxia. However, valid conclusions about the underlying processing pathways can only be drawn if participants apply the same strategy in both tasks (i.e. finding the middle between the obstacles). Yet, this assumption has never been tested directly. In this experiment, we investigated whether participants perform obstacle avoidance and line-bisectioning using similar strategies by manipulating the position of the obstacles and the start position of the hand relative to the obstacles. Our results indicate that the lateral hand position during obstacle avoidance does not only vary as a function of obstacle location but also strongly depends on the start position. Moreover, participants showed increased sensitivity to obstacle shifts occurring closer to the hand's start position. In contrast, during line-bisectioning the sensitivity to obstacles shifts was unaffected by the hand's start position. The findings suggest that during obstacle-avoidance the need to keep a safe distance from the obstacles is balanced with the requirement to minimise energetic demands. In contrast, the main intention during line-bisectioning is to move to the perceived midpoint as accurately as possible. The fact that very different constraints underlie trajectory planning in both tasks implies that caution has to be taken when interpreting differences in performance levels.

Concurrent Performance of a Cognitive and Dynamic Obstacle Avoidance Task: Influence of Dual-Task Training

ABSTRACTThe performance of 2 or more attention demanding tasks simultaneously is poorly understood. The purpose of the study was to investigate optimal practice strategies for performing 2 simultaneous tasks. Eighteen young adults walked and stepped over either a static or dynamic obstacle, while responding to an auditory Stroop test. Participants were randomly assigned to 1 of 3 groups: one that practiced both tasks simultaneously, practiced only the cognitive task, or received no practice. Results indicate that only the dual-task practice group showed significantly more improvement in the locomotor task through reduced variability of gait velocity, obstacle clearance, and takeoff distance. Findings demonstrate that the practice of two concurrent, attention demanding tasks results in the best performance improvement for both tasks.

Gait adaptability training improves obstacle avoidance and dynamic stability in patients with cerebellar degeneration

Balance and gait problems in patients with cerebellar degeneration lead to reduced mobility, loss of independence, and frequent falls. It is currently unclear, however, whether balance and gait capacities can be improved by training in this group of patients. Therefore, the aim of this study was to examine the effects of gait adaptability training on obstacle avoidance and dynamic stability during adaptive gait. Ten patients with degenerative cerebellar ataxia received 10 protocolized gait adaptability training sessions of 1h each during 5 weeks. Training was performed on a treadmill with visual stepping targets and obstacles projected on the belt’s surface. As the primary outcome, we used an obstacle avoidance task while walking on a treadmill. We determined avoidance success rates, as well as dynamic stability during the avoidance manoeuvre. Clinical ratings included the scale for the assessment of ataxia (SARA), 10m walking test, timed up-and-go test, berg balance scale, and the obstacle subtask of the emory functional ambulation profile (EFAP). Following the intervention, success rates on the obstacle avoidance task had significantly improved compared to pre-intervention. For successful avoidance, participants allowed themselves smaller stability margins in the sagittal plane in the (shortened) pre-crossing step. However, in the subsequent steps they returned to baseline stability values more effectively than before training. SARA scores and the EFAP obstacle subtask improved significantly as well. This pilot study provides preliminary evidence of a beneficial effect of gait adaptability training on obstacle avoidance capacity and dynamic stability in patients with cerebellar degeneration.

Echolocation behaviour of the big brown bat (Eptesicus fuscus) in an obstacle avoidance task of increasing difficulty.

Four big brown bats (Eptesicus fuscus) were challenged in an obstacle avoidance experiment to localize vertically stretched wires requiring progressively greater accuracy by diminishing the wire-to-wire distance from 50 to 10 cm. The performance of the bats decreased with decreasing gap size. The avoidance task became very difficult below a wire separation of 30 cm, which corresponds to the average wingspan of E. fuscus. Two of the bats were able to pass without collisions down to a gap size of 10 cm in some of the flights. The other two bats only managed to master gap sizes down to 20 and 30 cm, respectively. They also performed distinctly worse at all other gap sizes. With increasing difficulty of the task, the bats changed their flight and echolocation behaviour. Especially at gap sizes of 30 cm and below, flight paths increased in height and flight speed was reduced. In addition, the bats emitted approach signals that were arranged in groups. At all gap sizes, the largest numbers of pulses per group were observed in the last group before passing the obstacle. The more difficult the obstacle avoidance task, the more pulses there were in the groups and the shorter the within-group pulse intervals. In comparable situations, the better-performing bats always emitted groups with more pulses than the less well-performing individuals. We hypothesize that the accuracy of target localization increases with the number of pulses per group and that each group is processed as a package.

Role of vision in gait stabilization: Local dynamic stability in treadmill walking while blindfolded

Background: Vision plays a fundamental role in human locomotion, especially in orientation and obstacle avoidance tasks. On the other hand, it is known that the stepping pattern is mainly regulated by proprioceptive feedbacks and spinal processing.

Temporal Coupling Is More Robust Than Spatial Coupling: An Investigation of Interlimb Coordination After Stroke

ABSTRACT Interlimb coordination obtained through temporal and spatial coupling is a significant feature of human motor control. To understand the robustness of this capability the authors introduced a method to quantify interlimb coordination strength and compare individuals with asymmetric effector ability poststroke to nondisabled controls. Quantitative analyses determined the relative strength of interlimb coupling with an asymmetric obstacle avoidance task. Participants performed bimanual discrete, multijoint aiming movements in the frontal plane with a vertical barrier positioned midway to the target for one limb. To quantify coupling strength between limbs and groups, we regressed individual participant nonbarrier limb movement time or maximum vertical displacement separately, on barrier limb performance. Temporal and spatial interlimb coupling strength varied across participants in both groups. Barrier limb performance predicted nonbarrier limb behavior; however, interlimb coupling was significantly stronger for the nondisabled compared to the stroke group. In the stroke group, deficits in interlimb coordination affected spatial coupling more than temporal coupling. The decreased coupling strength detected, even in the presence of mild hemiparesis, demonstrates the measure's sensitivity. The authors propose this metric as a powerful assessment of the effectiveness of rehabilitation interventions and to monitor the recovery of bimanual coordination poststroke.

Action strategies used by children to avoid two vertical obstacles in non-confined space

Information used to determine the action strategies necessary to successfully pass through apertures is based on the dimensions of the individual and the mover's action capabilities (Warren in J Exp Psychol 10:683-703, 1984; Warren and Whang in J Exp Psychol 13:371-383, 1987). Previous research has demonstrated that when children must pass through small spaces, they will produce a shoulder rotation at apertures 1.6 times their shoulder width and smaller (i.e., critical point) and their avoidance strategies are based more on dynamic than geometric measures (Snapp-Childs and Bingham in Exp Brain Res 198:527-533, 2009; Wilmut and Barnett in Exp Brain Res 210:185-194, 2011). The question remains as to whether similar strategies exist when children are given a choice in their obstacle avoidance strategy. The current study aimed to determine the action strategies employed by children when confronted with a non-confined obstacle avoidance task. Specifically, the study intended to identify the aperture width that elicited a change in action (e.g., a shoulder rotation or a change in travel path). Children (N = 12, mean age = 7.1 years, ±0.2) were instructed to walk along a 10-m path toward a visible goal located at the end of the pathway and avoid colliding with the two vertical obstacles placed halfway (5 m) down the path on either side of the midline. The space between the obstacles ranged between 0.6 and 1.8 times the participant's shoulder width (presented in increments of 0.2). Results revealed that when the aperture was too small for straight passage, children choose to circumvent the obstacles rather than rotate their shoulders the majority of the time. However, unlike young and older adults (Hackney and Cinelli in Gait Posture 37:93-97, 2013a, Exp Brain Res 225:419-429, 2013b), this strategy was not used consistently. Instead, changes in travel path were highly variable both across participants and within trials. This variability suggests that a true critical point cannot be established for children in this non-confined task. Variable actions at the time of crossing were significantly correlated with the medial-lateral center of mass variability during the approach to the obstacles. These results further support the idea that children's actions may be more affected by dynamic factors than geometric measures.

Dyna-QUF: Dyna-Q based univector field navigation for autonomous mobile robots in unknown environments

A novel approach was presented to solve the navigation problem of autonomous mobile robots in unknown environments with dense obstacles based on a univector field method. In an obstacle-free environment, a robot is ensured to reach the goal position with the desired posture by following the univector field. Contrariwise, the univector field cannot guarantee that the robot will avoid obstacles in environments. In order to create an intelligent mobile robot being able to perform the obstacle avoidance task while following the univector field, Dyna-Q algorithm is developed to train the robot in learning moving directions to attain a collision-free path for its navigation. Simulations on the computer as well as experiments on the real world prove that the proposed algorithm is efficient for training the robot in reaching the goal position with the desired final orientation.