Hand Movements During Obstacle Avoidance in Real and Natural Environments, 2021-2025

The data collected as part of this grant investigates human obstacle avoidance behaviour under a range of different conditions. The grant consists of four separate but connected projects that will be presented separately below.

1. Obstacle avoidance of physical, stereoscopic, and pictorial objects
(published: Giesel, M., Ruseva, D., & Hesse, C. (2025). Obstacle avoidance of physical, stereoscopic, and pictorial objects. Virtual Reality, 29(1), 1-17)

In this project, we investigated in 4 experiments (Experiment1: N= 27, Experiment 2: N=18, Experiment 3: N= 21, Experiment 4, N=21) the visual cues that are required to ensure natural hand movements in virtual settings. Participants were asked to reach towards a target position without colliding with obstacles of varying height that were placed in the movement path. Using a pre-test post-test design, we tested obstacle avoidance for 2D and 3D images of obstacles both before and after exposure to the physical obstacles. Consistent with previous findings, we found that participants initially underestimated the magnitude differences between the obstacles, but after exposure to the physical obstacles, avoidance performance for the 3D images became similar to performance for the physical obstacles. No such change was found for 2D images. Our findings highlight the importance of disparity cues for naturalistic motor actions in personal space.

All associated data files and explanations can be found on OSF: https://osf.io/6tf9r

2. From Discomfort to Danger: Exploring how affective obstacle properties influence avoidance in stepping
(published: Chee, Z. J., Giesel, M., & Hesse, C. (2025). From Discomfort to Danger: Exploring how affective obstacle properties influence avoidance in stepping. Perception (in press).

In this project, we investigated in two experiments how visual uncertainty and perceived unpleasantness and danger affects obstacle avoidance strategies in stepping. In Experiment 1 (N=20), we found that lead minimum foot clearance (MFC) was initially higher under monocular vision but decreased to binocular levels over trials. While obstacle unpleasantness did not systematically affect MFC or crossing step length, perceived unpleasantness ratings correlated weakly with crossing step length. However, because dangerousness and painfulness ratings were not collected, it remained unclear whether unpleasantness directly influenced avoidance behaviour or served as a proxy for perceived risk. To address this, Experiment 2 (N=22) introduced obstacles covered with metal stud spikes or smooth surfaces, with additional ratings of dangerousness and painfulness. Results showed that MFC was higher for spiky than smooth obstacles. Crucially in this experiment, ratings of perceived dangerousness, not unpleasantness, correlated positively with crossing step length, after controlling for other perceptual ratings. These findings suggest that perceptual affective properties modulate avoidance parameters. However, the nature of those modulations is stimulus specific and highly depends on task demands.

All associated data files and explanations can be found on OSF: https://osf.io/mpk93/

3. Effects of light level, material appearance, and virtuality on hand movements
(in preparation for publication)

In this project, we investigated in four experiments, if and how visual uncertainty and assumptions about the consequences of actions (e.g., collision with an obstacle) shape motor behaviour. Within each experiment, visual uncertainty was varied by using three different light levels. Assumptions about the consequences of collisions with obstacles were manipulated by using obstacles differing in perceived fragility. Between experiments, we varied whether the obstacles were real or virtual objects by using a mirror setup. In Experiment 1 (N=22), the obstacles were real objects, in Experiment 2 (N=20), they were virtual objects, and in Experiments 3 (N=20) and 4 (N=20), real and virtual obstacles were presented simultaneously at the same location. In all experiments, participants moved their hands around an obstacle to pick up an object. We hypothesised that lower light levels would result in larger safety margins (i.e., distance between hand and obstacle), but the overall magnitude of the safety margins would be smaller for virtual obstacles than for real obstacles. Regarding perceived fragility, we assumed that safety margins would be larger for the more fragile obstacle but only if it was a real object. Consistent with our expectations, we found in all experiments that safety margins increased with decreasing light level. The effect of perceived fragility on safety margins was weaker and less consistent but was, contrary to our hypotheses, presented both for real and virtual obstacles. Overall, in these tasks, obstacle avoidance and grasping behaviour were very similar in real and virtual environments.

The trajectory data, explanations, and methods information for these experiments are available on Reshare in the folder entitled “DinnerExperiments.zip”

4. The role of 3D Information, feedback, light levels, and contrast in Virtual Obstacle Avoidance
(ongoing experiments)

In this project, we investigated if and how obstacle avoidance in virtual environments is influenced by 3D information, feedback, light levels, and contrast. In many VR applications, we do not perceive the movement of our actual hand but that of a virtual representation that is interacting with virtual objects of varying degrees of complexity (e.g., 2D or 3D representations). Frequently, some form of feedback (visual, auditory, and/or haptic) is used to signal contact with targets or collisions with obstacles. In four experiments, we investigated the effects of perceptual uncertainty (varying contrast) and visual feedback on obstacle avoidance movements when visual information about the moving hand is provided by a sparse virtual representation (see README for details). We varied the visual complexity of the obstacle between experiments: 2D image in Experiment 1 (N=22), and a virtual 3D objects in Experiment 2 to 4 (Experiment 2: N=24, Experiment 3: N= 24, Experiment 4: N=22). Experiments were conducted in a mirror setup (see README for details). We found that when the obstacle was a 2D image, the effects of perceptual uncertainty and feedback on safety margins were negligible. Adding the virtual 3D object in Experiment 2 resulted in behaviour similar to that observed in real environments highlighting again the importance of 3D information for natural movements. Data analysis from Experiments 3-4 is currently ongoing.

The trajectory data, explanations, and methods information for these experiments are available on Reshare in the folder entitled “VirtualHandObstacleAvoidance.zip”