Colony Collapse Disorder Virtual Reality
the problem is that the poison (Imidacloprid, the most widely used
insecticide in the world, which is mostly made by Bayer) destroys
the bees' ability to navigate. so the bees are able to get to the
poison but they can't get back.
so what happens in Colony Collapse Disorder (CCD) is that the bees
"get lost" and "disappear" and die from the poison, while the bees
that stayed at the hive die from starvation.
but even if they could come back we don't want the poisoned bees to
come back inside the hive, because then they will feed poison to the
young ones and the queen.
bees Virtual Reality (VR) simulator
first of all putting bees into VR has been done already by the Free
University of Berlin although in another context:
"Honeybees Enter Virtual Reality So Scientists Can Study Their
Learning and Its Neural Correlates in a Virtual Environment for
my solution has 2 components: on the one hand - helping the bees not
reach the poison in the first place; and on the other hand - if a
bee somehow got to the poison, then to prevent this bee from getting
into the hive.
currently a bee hive has a single entrance which is also the single
i suggest we attach to this entrance a virtual reality system for
bees, which will basically be something the size of a shoe box only
cubed. this will be the chamber where the bee experiences the
virtual reality "ride".
if we wanted only visual effects for the bee we would make each of
the inner walls of the system a monitor screen.
but we also want the bee to be able to freely move fly in 3-D inside
the simulator, just like we want people to freely walk in 2-D
(forward, backward, left, right) inside a human VR simulator, so
bees have all these plus the third dimension (up, down).
so the equivalent of this for bees will be walls with electric fans,
something like the fan in a desktop computer at home. so each wall
will be able to push the bee as if the bee is flying into a wall of
wind. the bee will feel itself flying through the air and
"progressing" in that direction, when in fact the bee is really
flying in the same place. just like a person walking on the
if we wanted only haptic effects for the bee we would make each of
the inner walls of the system a wall of fans.
but we want BOTH: we want the bee to see (visual) and feel (haptic)
that it is moving. how do we do that?
for this we use the fact that bees are not sharp-eyed. see for
example the bee orchid flower.
so we use a trick from the 18th century of horror theater
phantasmagoria, where they used magic lantern projection on smoke of
we also use rear projection effect, but instead of on smoke or sheet
we will use a mesh like a mosquito net. this will allow the wind
current from the fans outside the chamber to reach the bee inside
the chamber. this will also allow the images from a small projector
outside of the chamber to reach the bee inside the chamber. behind
each wall of the chamber there are one projector and let's say four
we want 4 of the walls of the chamber to be on hinges so that they
can open like a doors:
the 1st wall that needs to act like a door is the one closest to the
hive's entrance/exit. through this door we let an incoming bee into
the hive after it passed the simulator. the simulator (with the fans
rotating in reverse).
the 2nd wall that needs to act like a door is the one farthest from
the hive's entrance/exit. through this door we let an outbound bee
out to the field after it passed the simulator.
the 3rd wall that needs to act like a door is the floor which will
have a jar with honey or something that the bees really like, this
is the reward. the jar needs to close right after the bee finishes
and flies home.
the 4th wall that needs to act like a door is the ceiling which will
be like a death door for a bee that failed in the simulator and is
thus presumed to be poisoned. the fans blow it into a trap where it
another component that we need in the simulator is an ultraviolet
(UV) flashlight, because bees can see ultra violet light and they
navigate by the sun, which looks like a UV flashlight to them.
this needs to move along the walls so we need either 1 UV flashlight
on 1 robotic arm, or a cheaper solution will be 1 UV flashlight on
each wall of the cube. so we need 6 UV flashlights. each UV
flashlight will be mounted on an XY rails like in a CNC machine or a
3D printer. so the computer tracks where the insect is moving
through a camera and activates in any given time only one of the
flashlights. or i think a cheaper option will be one static
flashlight with optical fibers that leads it to the XY rails. we
need also a mechanism to choose which of the 6 fibers is the one UV
flashlight connected to. we can take a small electric motor, like
Lego motor, that can rotate a Lego axle. then we place the
flashlight in front of the motor so that when it lights, the light
goes along the axle. now we position a tiny mirror in 45 degrees,
half towards the flashlight and half sideways towards a direction
perpendicular to the axle. and then we position the 6 fibers around
the mirror so that when the light from the flashlight can reach a
fiber, if the motor rotated the mirror to that fiber.
another option to control them which i'm sure is a lot simpler is
with electronics, unfortunately i don't know anything about
electronics. it's like we choose which of the 7 segment display to
activate, but here it's 6 "segments" and we only need to activate
one segment at any time.
when the doors open there is a short time of stress for the bee
because it might be that that Sun disappeared, but after a few
seconds, something very exciting is happening: entrance or exit or
reward or death, so i hope the bee is like humans and it will
concentrate on the exciting thing. of course in the death door case
the stress will not happily resolve but this option is a rare and
necessary evil to save the rest of the colony.
another element that we need in the simulator are "breadcrumbs"
which will hint the bee in the right direction of the good flowers
field. again we need this behind every wall of the cube. this will
be the nozzle of a spray can of a smell a similar to the flowers'
nectar as possible. it can even simply be fresh flowers that we
grounded into a powder.
like with the fiber optics, we can use a system of valves and one
pump. this is also similar to Lego in this case the pneumatic air
pressure Lego. my dad bought me this one when i was a child
First generation Pneumatics excavator from 1984
and it's super cool! so in the same way that a combination of valves
can use the same one pump to either extend this section or that
section or contract this or that section, we can switch between the
tubes to reach the desired end nozzle.
you can see how one valve works in this video, but when you connect
a few valves you can control more outputs with one input pump:
LEGO Pneumatic experiments by The Technic Gear LEGO Reviews
or instead of all this we could use a rotating nozzle like the
mirror in my first suggestion was rotating, and instead of a
flashlight we have the spray of flowers.
recording the correct "ride" and a little around that path using
now we need to use the ability of bees to share knowledge with each
other in the form of waggle dance.
so ideally we can train one bee and teach this bee the directions
from the beehive to the good field of organic flowers which have no
so first we use drones to make a video of this path, in all the
directions that a bee can look (right left forward backward up
down). it's like what Google Maps do in Street View that they film
the streets, but also on the axis of up and down.
for the up down, i think we can use one high flyby and another low
flyby on the same coordinates and interpolate between these with the
but in case this complicates things then we can forget the up down
axis, it's negligible.
ok so we feed all the frames of the video into the computer, so if
the bee inside the simulator "is" in a specific coordinate, the
computer can project the right image on the 6 walls around the bee.
of course for showing the bee the way in the other direction we
don't need to record the film again, it will just play backwards,
because the bee will be in the coordinates in the opposite order.
training the first bee to use the right path
now we take the first bee when it wants to get out of the beehive.
it doesn't get out to reality. instead it gets into our simulator.
so we tempt it with scents to go to the virtual good path and when
it reaches the virtual good field location we reward it with the
honey which is the ultimate prize.
now after the bee memorized the way to the virtual field and drank
its full, we let it return to the beehive which is just the same
virtual way in reverse. when it reaches the virtual hive, we open
the door to the real beehive, and this first bee teaches its friends
that there is tasty treats and teaches them the virtual path which
is also the right path in reality.
of course we can map with drones a few good fields in the area and
teach the bees a few alternatives.
so ideally without our simulator, the first bee exits the hive,
finds a good field and drink nectar, then the bee comes back to the
hive, and teaches its friends through dance how to get to the good
but the problem is that the good field is not the closest to the
hive, so the bees get opportunistic on the way or get tempted on the
way to try out the poisonous field.
so the first thing we do is set a password to get back into the
beehive. the password is not verbal it's navigational. let's say for
example when the bee exits the hive it must go in the simulator: 100
meters to the north, then 50 meters to the east. when the bee comes
back to the hive this is of course flipped: 50 meters to the west
and 100 meters to the south.
we teach this to the first bee before it gets out of the hive. so
from the bees view point this is just part of the path. but for us
this is the secret part to know when the bee is back whether the bee
was poisoned on the way, because then it will not know how to do
this password part, and then we take that poisoned bee to the trap
behind the death door, and save all the other bees at the hive, the
queen and the young ones, from poisoning.
summary of the process
(1) record good path with drone
(2) connect simulator to beehive (possibly at night).
(3) teach first outgoing bee the password and the virtual good path
(4) return the first incoming bee (that same bee from step (3) )
through the virtual good path and the password (still reinforcing
memory, not testing password yet)
(5) now switch the simulator from learning mode to "only password"
(6) meanwhile the first bee taught its friends the good path and
(7) the next outgoing bee does the password and gets to the
real-world good field.
(8) this bee is rewarded with real good nectar.
(9) this bee returns to the beehive which is the entry to the
simulator. if it passed the password successfully it teaches more
bees and we go to step (7).
(10) in the following day we no longer need the simulator, because
all the bees know how to get to the real-world good field.
since this machine is needed only in the beginning flights, it
doesn't make sense for each beekeeper to buy a machine of his own.
after one day of training the bees are happy, until that good field
is depleted. so one machine can be shared by multiple beekeepers,
each with multiple beehives.
so the body which buys the machine and the drone should be the
central organization that the beekeepers are affiliated to, such as
the American Beekeeping Federation or British Beekeepers Association
and so on. this body will rent the machine to the farmers in low
prices or even for free, because it's also in this body's interest
that the honey bees will flourish.
and it goes without saying that since it's a human survival issue -
the world governments and international organizations must subsidize
and help as well!