I’m not sure if their an art project or practical tool but these autonomous biomorphic Robot Jellyfish are interesting. From National Geographic:
Propelled by flexible, electrically driven tentacles, robotic jellyfish swim at the Hannover Fair.
Using a type of “swarm intelligence,” the Festo company’s so-called AquaJellies set their own courses and can come together or avoid each other as needed. The robots “talk” via light pulses underwater and via radio at the surface.
On Technium Kevin Kelly writes:
One one level, these autonomous robotic jelly fish illuminated the mechanism by which real jellyfish swim. … The parallels in their motions — clearly visible in the video — feel so organic that we immediately assign them life-like adjectives.
I think we are primed to find lifelikeness in machines. E.O. Wilson calls it our biophilia — our intense attraction to living things. As we design machines to approach the complexity of organisms and mimic their behavior (as these do), we will be very quick to include them in our love.
A series of video of the companies Aqua and Air Jellies are on the company’s website.
Design News describes these robots in more detail:
Their tentacle construction takes its cues from the functional anatomy of some fish fins. These bio-inspired tentacles consist of two flexible external surfaces connected by a series of internal ribs. When one of the surfaces is put under tension, the entire tentacle bends in the direction of the applied force — a phenomenon that Festo calls the “fin ray effect.”
Festo uses an electric drive, geared power transmission and linkages to actuate the tentacles. Alternating tension between the two external surfaces creates a wave-like motion that propels the robots through the water or air. Fischer describes the resulting movement as “peristaltic” since the waving tentacles seem to move by something like muscle contractions.
Whether they swim or fly, these two types of jellyfish steer themselves by carefully controlled weight shifts. As Fischer explains, their bodies contain a servo-driven swash plate connected to a four-armed pendulum that changes their center of gravity. “The pendulum shifts their weight, and they move in a new direction,” he says.
And for the AquaJelly in particular, that new direction is determined autonomously. This underwater robot guides itself with the help of a sensor array, communications systems and control software based on robotic swarm-intelligence. Fischer notes, for example, multiple AquaJelly robots can avoid each other in the water, using light sensors to pick up the presence of their tank mates. They also have pressure sensors that allow them to gauge their depth within a few mm.
AquaJelly robots also manage their own battery-charging behavior. They communicate with an in-tank charging dock wirelessly via ZigBee, for example, to make sure the dock isn’t occupied when they need to charge.
According to Fischer, giving these robots such a high degree of autonomy required a mechatronic approach in which the mechanical design, sensor engineering and control software were all developed concurrently. “Even simple autonomy is not so simple,” he says.