Rat pups and random robots generate similar self-organized and intentional behavior

A guide to practical babooning: Historical, social, and cognitive contingency

“…This alteration in the environment then increases the chance that another cube, being blindly pushed around, is deposited in the vicinity because the cluster of two cubes is more detectable than is a single cube; a simple self-organizing process explains the didabots' behavior and results in the formation of ever-larger clusters as well as generating the abiding impression in a human observer that the didabots are ''trying'' their best to ''clean up''. [67][68][69] Were we to interpret the didabots' behavior as reflecting the operation of this kind of proximate ''clustering'' mechanism, reflecting their ''goal'' of cleaning the arena, the explanation would be entirely erroneous: both inaccurate and incomplete. It is also important to note that the clustering behavior is entirely dependent on the placement of the sensors on the didabots' body: move one of the sensors around to the front, and the clustering behavior disappears entirely because now objects directly in front of the robots are avoided in the same way as those off to the side, which means that no pushing behavior occurs.…”

“…This means that although they turn away and avoid cubes on either side, a cube directly in front of them is pushed along because the didabot cannot ''see'' it (i.e., its sensors received no stimulation from it). [67][68][69] If the didabot then encounters another cube off to the side, triggering its sensor, it produces avoidance behavior, moving off to the left or right, and leaving the object it has just been pushing next to the object it has just avoided. In other words, the didabot clusters the two cubes.…”

“…B1). [67][68][69] When placed into an arena in which small cubes are scattered randomly, these robots move around clustering the cubes so that eventually there are usually only two large heaps, with a few cubes left here and there against the walls (Fig. B1).…”

“…In fact, the robots have only one kind of sensor-for proximity-and they follow one simple ''rule'' that results from the manner in which they are wired: if sensory information is received on the left, then turn right, and if sensory information is received on the right, turn left. In other words, the robots are designed exclusively to avoid obstacles [67][68][69] . Why then does clustering occur?…”

Self-organising Thermoregulatory Huddling in a Model of Soft Deformable Littermates

“…Models of huddling have typically studied self-organisation from one of two directions. First, models using simulated or physical robots (even bean bags) have helped establish the importance of the animal morphology in determining how interactions between the body and the environment affect group-level aggregation patterns [7][8][9]. Second, models based on the thermodynamics of heat exchange have instead described individuals simply as gas particles bouncing around in a chamber [1,10].…”