Limits...
Adjustments of speed and path when avoiding collisions with another pedestrian.

Huber M, Su YH, Krüger M, Faschian K, Glasauer S, Hermsdörfer J - PLoS ONE (2014)

Bottom Line: When walking in open space, collision avoidance with other pedestrians is a process that successfully takes place many times.Currently, the literature is not yet conclusive of how humans adjust these two parameters in the presence of an interferer.Overall, the results were incompatible with predictions from existing models of locomotor collision avoidance.

View Article: PubMed Central - PubMed

Affiliation: Center for Sensorimotor Research, Institute for Clinical Neuroscience, Ludwig-Maximilians-University, Munich, Germany.

ABSTRACT
When walking in open space, collision avoidance with other pedestrians is a process that successfully takes place many times. To pass another pedestrian (an interferer) walking direction, walking speed or both can be adjusted. Currently, the literature is not yet conclusive of how humans adjust these two parameters in the presence of an interferer. This impedes the development of models predicting general obstacle avoidance strategies in humans' walking behavior. The aim of this study was to investigate the adjustments of path and speed when a pedestrian is crossing a non-reactive human interferer at different angles and speeds, and to compare the results to general model predictions. To do so, we designed an experiment where a pedestrian walked a 12 m distance to reach a goal position. The task was designed in such a way that collision with an interferer would always occur if the pedestrian would not apply a correction of movement path or speed. Results revealed a strong dependence of path and speed adjustments on crossing angle and walking speed, suggesting local planning of the collision avoidance strategy. Crossing at acute angles (i.e. 45° and 90°) seems to require more complex collision avoidance strategies involving both path and speed adjustments than crossing at obtuse angles, where only path adjustments were observed. Overall, the results were incompatible with predictions from existing models of locomotor collision avoidance. The observed initiations of both adjustments suggest a collision avoidance strategy that is temporally controlled. The present study provides a comprehensive picture of human collision avoidance strategies in walking, which can be used to evaluate and adjust existing pedestrian dynamics models, or serve as an empirical basis to develop new models.

Show MeSH

Related in: MedlinePlus

Illustration of model predictions about the influence of the obstacle distance on the avoidance strategy.The dashed arrows symbolize a fast walking speed of the pedestrian, the solid arrow a slower speed. A: The obstacle is represented as a repulsive potential. The fast speed allows the pedestrian to climb up the potential higher before the repulsive force leads to a significant change of the path as compared to a slow walking speed. B: A fixed spatial horizon (dashed and solid circle) specifies when an obstacle (grey circle) is taken into account. The horizon is not dependent on the speed (dashed and solid arrows) of the pedestrian [3][7]. C: The horizon is speed dependent. Therefore, the obstacle is taken into account at a greater distance for a higher walking speed. This is comparable to a time-to collision mechanism [15].
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3935867&req=5

pone-0089589-g001: Illustration of model predictions about the influence of the obstacle distance on the avoidance strategy.The dashed arrows symbolize a fast walking speed of the pedestrian, the solid arrow a slower speed. A: The obstacle is represented as a repulsive potential. The fast speed allows the pedestrian to climb up the potential higher before the repulsive force leads to a significant change of the path as compared to a slow walking speed. B: A fixed spatial horizon (dashed and solid circle) specifies when an obstacle (grey circle) is taken into account. The horizon is not dependent on the speed (dashed and solid arrows) of the pedestrian [3][7]. C: The horizon is speed dependent. Therefore, the obstacle is taken into account at a greater distance for a higher walking speed. This is comparable to a time-to collision mechanism [15].

Mentions: Inconsistency of human collision avoidance strategies exists not only in the empirical data, but also in the models that attempt to describe this behavior. These models differ not only in their basic assumptions but also in their predictions about the deployment of path or speed adjustments in collision avoidance. To exemplify this, let us assume a scenario without considerable spatial constraints, and with only one moving obstacle (another person) walking at different speeds. Furthermore, the crossing person (i.e. the obstacle) does not react to the pedestrian by any means. A navigation model based on the heuristics [7] proposes that, as a first rule, a pedestrian chooses a walking direction that allows for the most direct path towards the goal, taking into account the obstacles in between. This model [7] uses a default maximum “horizon distance” of a pedestrian, for which obstacles are taken into account. Critically, this default “horizon distance” leads to adjustments of the path at a fixed distance to the obstacle, independently of the pedestrian’s walking speed (see Fig. 1B). A second rule of this model describes that the pedestrian tries to keep a minimal safety distance to the obstacles, which becomes relevant only within small distances to the obstacle or in the presence of spatial constraints. Given enough space to navigate, together the two rules converge to an adjustment of the path rather than of the speed. Other models inspired by Newton’s laws of motion (e.g., the “social force model” [11], or a modification of it [12]) describe the pedestrian’s motion as a combination of a driving force pointing towards the goal position, and repulsive forces originating from the obstacles. Due to the repulsive and driving forces, adjustments of the walking path are expected rather than adjustments of the walking speed. These models predict that path and speed adjustments are initiated closer to the interfering person when the pedestrian moves at a higher speed (see Fig. 1A). A third approach, based on the theories of optimal control, suggests that the smoothness of the movement is optimized [13]. In an improved version of this model, a penalty for speed changes was introduced [14], favoring path adjustments rather than speed adjustments in collision avoidance. As another approach, Fajen and Warren [3] modeled an obstacle avoidance behavior, in which the angular acceleration was described as a function of the goal, the obstacle angle and distance, taking into account only path adjustments, and ignoring speed changes, thereby predicting path adjustments at the same distance to an obstacle (see Fig. 1B). A last possible collision avoidance strategy, known as the time-to-collision strategy [15], proposes that the initiation of path and speed adjustments starts at larger distance to the obstacle with higher walking speed of the pedestrian, while the time point for the adjustment remains the same (see Fig. 1C).


Adjustments of speed and path when avoiding collisions with another pedestrian.

Huber M, Su YH, Krüger M, Faschian K, Glasauer S, Hermsdörfer J - PLoS ONE (2014)

Illustration of model predictions about the influence of the obstacle distance on the avoidance strategy.The dashed arrows symbolize a fast walking speed of the pedestrian, the solid arrow a slower speed. A: The obstacle is represented as a repulsive potential. The fast speed allows the pedestrian to climb up the potential higher before the repulsive force leads to a significant change of the path as compared to a slow walking speed. B: A fixed spatial horizon (dashed and solid circle) specifies when an obstacle (grey circle) is taken into account. The horizon is not dependent on the speed (dashed and solid arrows) of the pedestrian [3][7]. C: The horizon is speed dependent. Therefore, the obstacle is taken into account at a greater distance for a higher walking speed. This is comparable to a time-to collision mechanism [15].
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3935867&req=5

pone-0089589-g001: Illustration of model predictions about the influence of the obstacle distance on the avoidance strategy.The dashed arrows symbolize a fast walking speed of the pedestrian, the solid arrow a slower speed. A: The obstacle is represented as a repulsive potential. The fast speed allows the pedestrian to climb up the potential higher before the repulsive force leads to a significant change of the path as compared to a slow walking speed. B: A fixed spatial horizon (dashed and solid circle) specifies when an obstacle (grey circle) is taken into account. The horizon is not dependent on the speed (dashed and solid arrows) of the pedestrian [3][7]. C: The horizon is speed dependent. Therefore, the obstacle is taken into account at a greater distance for a higher walking speed. This is comparable to a time-to collision mechanism [15].
Mentions: Inconsistency of human collision avoidance strategies exists not only in the empirical data, but also in the models that attempt to describe this behavior. These models differ not only in their basic assumptions but also in their predictions about the deployment of path or speed adjustments in collision avoidance. To exemplify this, let us assume a scenario without considerable spatial constraints, and with only one moving obstacle (another person) walking at different speeds. Furthermore, the crossing person (i.e. the obstacle) does not react to the pedestrian by any means. A navigation model based on the heuristics [7] proposes that, as a first rule, a pedestrian chooses a walking direction that allows for the most direct path towards the goal, taking into account the obstacles in between. This model [7] uses a default maximum “horizon distance” of a pedestrian, for which obstacles are taken into account. Critically, this default “horizon distance” leads to adjustments of the path at a fixed distance to the obstacle, independently of the pedestrian’s walking speed (see Fig. 1B). A second rule of this model describes that the pedestrian tries to keep a minimal safety distance to the obstacles, which becomes relevant only within small distances to the obstacle or in the presence of spatial constraints. Given enough space to navigate, together the two rules converge to an adjustment of the path rather than of the speed. Other models inspired by Newton’s laws of motion (e.g., the “social force model” [11], or a modification of it [12]) describe the pedestrian’s motion as a combination of a driving force pointing towards the goal position, and repulsive forces originating from the obstacles. Due to the repulsive and driving forces, adjustments of the walking path are expected rather than adjustments of the walking speed. These models predict that path and speed adjustments are initiated closer to the interfering person when the pedestrian moves at a higher speed (see Fig. 1A). A third approach, based on the theories of optimal control, suggests that the smoothness of the movement is optimized [13]. In an improved version of this model, a penalty for speed changes was introduced [14], favoring path adjustments rather than speed adjustments in collision avoidance. As another approach, Fajen and Warren [3] modeled an obstacle avoidance behavior, in which the angular acceleration was described as a function of the goal, the obstacle angle and distance, taking into account only path adjustments, and ignoring speed changes, thereby predicting path adjustments at the same distance to an obstacle (see Fig. 1B). A last possible collision avoidance strategy, known as the time-to-collision strategy [15], proposes that the initiation of path and speed adjustments starts at larger distance to the obstacle with higher walking speed of the pedestrian, while the time point for the adjustment remains the same (see Fig. 1C).

Bottom Line: When walking in open space, collision avoidance with other pedestrians is a process that successfully takes place many times.Currently, the literature is not yet conclusive of how humans adjust these two parameters in the presence of an interferer.Overall, the results were incompatible with predictions from existing models of locomotor collision avoidance.

View Article: PubMed Central - PubMed

Affiliation: Center for Sensorimotor Research, Institute for Clinical Neuroscience, Ludwig-Maximilians-University, Munich, Germany.

ABSTRACT
When walking in open space, collision avoidance with other pedestrians is a process that successfully takes place many times. To pass another pedestrian (an interferer) walking direction, walking speed or both can be adjusted. Currently, the literature is not yet conclusive of how humans adjust these two parameters in the presence of an interferer. This impedes the development of models predicting general obstacle avoidance strategies in humans' walking behavior. The aim of this study was to investigate the adjustments of path and speed when a pedestrian is crossing a non-reactive human interferer at different angles and speeds, and to compare the results to general model predictions. To do so, we designed an experiment where a pedestrian walked a 12 m distance to reach a goal position. The task was designed in such a way that collision with an interferer would always occur if the pedestrian would not apply a correction of movement path or speed. Results revealed a strong dependence of path and speed adjustments on crossing angle and walking speed, suggesting local planning of the collision avoidance strategy. Crossing at acute angles (i.e. 45° and 90°) seems to require more complex collision avoidance strategies involving both path and speed adjustments than crossing at obtuse angles, where only path adjustments were observed. Overall, the results were incompatible with predictions from existing models of locomotor collision avoidance. The observed initiations of both adjustments suggest a collision avoidance strategy that is temporally controlled. The present study provides a comprehensive picture of human collision avoidance strategies in walking, which can be used to evaluate and adjust existing pedestrian dynamics models, or serve as an empirical basis to develop new models.

Show MeSH
Related in: MedlinePlus