356
A. Thoma et al.
of what is optimal. While most ground-bound path planners define efficiency as the
shortest possible path [20, 21], this is questionable for flying platforms, where a third
dimension of movement has to be taken into account. Moving in the third dimension,
e.g., up or downward, is energetically different from moving inplane. Because of this,
2D methods cannot easily be transferred to 3D. An understanding of the flight strategies
of insects and the working principles behind their behavior might help to improve the
current flight algorithm to a high degree [22].
Currently, much knowledge is available about the behavior of bumblebees in specific
situations. Several research groups identified critical aspects of bee behavior when flying
through tunnels. For example, bees maintain equidistance to both walls [23, 24], by
maintaining equivalent optic flow on both eyes [25]. Additionally, by keeping optic flow
constant, the flight speed is adapted to the width of the tunnel [25, 26]. Even though the
behavior in an empty tunnel is understood quite well, none of the above gives information
on obstacle encounters.
If challenged with a series of vertical or horizontal obstacles within a flight tunnel,
bumblebees do not show any significant difference in the maneuver when avoiding the
obstacles horizontally or vertically [27]. However, body size has a more significant influence on the flight behavior of bumblebees than the obstacle orientation. With increasing
body size, flight performance is impaired [27]. However, this work investigates only one
possibility to evade the obstacle in a very confined space such that the bee cannot decide
between several alternatives.
When bumblebees have to decide to fly through two horizontally aligned gates, they
tend to choose the wider gap [28]. However, this situation may change when gaps are not
horizontally aligned, but the bees have to decide to avoid an object by moving upward
or sideward.
Thus, we investigated the decision of bumblebees between vertical and horizontal
avoidance in a tunnel.
2.2 Experimental Set-Up
Most research has focused on studying flight control inplane, either horizontal such as
avoiding trees or vertical as in a 110 m hurdles. However, flying insects in nature can
decide to fly above or around an obstacle. The goal of this work was to identify situations
in which bumblebees prefer to fly over an obstacle instead of around. The experiments
were conducted at the University Bielefeld. All experiments were performed with bumblebees, Bombus Terrestris, from two different colonies (Koppert, GmbH, Zeppelinstr.
32, Straelen, Germany). The hive was placed in a 0.5 × 0.5 × 0.3 m 3 acrylic box,
which was covered by a black cloth to mimic a natural situation. At arrival, the colonies
consisted of less than 30 worker bees and one queen.
The hive enclosure was connected to a 1.5 × 0.3 × 0.3 m 3 flight tunnel leading to
a 1.0 × 1.0 × 0.75 m 3 foraging chamber. A gravity feeder with a 30%/vol. Sucrose
solution was placed on a podium in the center of the foraging chamber. Additionally,
ground pollen was placed in the hive enclosure. Silicon tubes with an inner diameter
of 25 mm connected hive, flight tunnel, and chamber. Closable gates, installed on both
ends of the tunnel, were used to control in- and outflow of bumblebees. The bees had
A. Thoma et al.
of what is optimal. While most ground-bound path planners define efficiency as the
shortest possible path [20, 21], this is questionable for flying platforms, where a third
dimension of movement has to be taken into account. Moving in the third dimension,
e.g., up or downward, is energetically different from moving inplane. Because of this,
2D methods cannot easily be transferred to 3D. An understanding of the flight strategies
of insects and the working principles behind their behavior might help to improve the
current flight algorithm to a high degree [22].
Currently, much knowledge is available about the behavior of bumblebees in specific
situations. Several research groups identified critical aspects of bee behavior when flying
through tunnels. For example, bees maintain equidistance to both walls [23, 24], by
maintaining equivalent optic flow on both eyes [25]. Additionally, by keeping optic flow
constant, the flight speed is adapted to the width of the tunnel [25, 26]. Even though the
behavior in an empty tunnel is understood quite well, none of the above gives information
on obstacle encounters.
If challenged with a series of vertical or horizontal obstacles within a flight tunnel,
bumblebees do not show any significant difference in the maneuver when avoiding the
obstacles horizontally or vertically [27]. However, body size has a more significant influence on the flight behavior of bumblebees than the obstacle orientation. With increasing
body size, flight performance is impaired [27]. However, this work investigates only one
possibility to evade the obstacle in a very confined space such that the bee cannot decide
between several alternatives.
When bumblebees have to decide to fly through two horizontally aligned gates, they
tend to choose the wider gap [28]. However, this situation may change when gaps are not
horizontally aligned, but the bees have to decide to avoid an object by moving upward
or sideward.
Thus, we investigated the decision of bumblebees between vertical and horizontal
avoidance in a tunnel.
2.2 Experimental Set-Up
Most research has focused on studying flight control inplane, either horizontal such as
avoiding trees or vertical as in a 110 m hurdles. However, flying insects in nature can
decide to fly above or around an obstacle. The goal of this work was to identify situations
in which bumblebees prefer to fly over an obstacle instead of around. The experiments
were conducted at the University Bielefeld. All experiments were performed with bumblebees, Bombus Terrestris, from two different colonies (Koppert, GmbH, Zeppelinstr.
32, Straelen, Germany). The hive was placed in a 0.5 × 0.5 × 0.3 m 3 acrylic box,
which was covered by a black cloth to mimic a natural situation. At arrival, the colonies
consisted of less than 30 worker bees and one queen.
The hive enclosure was connected to a 1.5 × 0.3 × 0.3 m 3 flight tunnel leading to
a 1.0 × 1.0 × 0.75 m 3 foraging chamber. A gravity feeder with a 30%/vol. Sucrose
solution was placed on a podium in the center of the foraging chamber. Additionally,
ground pollen was placed in the hive enclosure. Silicon tubes with an inner diameter
of 25 mm connected hive, flight tunnel, and chamber. Closable gates, installed on both
ends of the tunnel, were used to control in- and outflow of bumblebees. The bees had