Holy cow! I knew Henry when he was a postdoc and I was a PhD student at Georgia Tech. My mind was blown when I saw such a familiar face on one of my fav RU-vid channels 😄. Henry, if you see this, I hope all is well man!
I don't know what's the weirdest part about this video: the beard or the snake crawling through it or the guy inspired by this beard loving snake to build a robot version of it. Oh my head hurts!
I haven’t watched a Computerphile video in a while, but I clicked on this and stared with amazement for the whole twelwe minutes. This guy was just constantly spitting out a stream of interesting things from between his epic beards. Please give us more of him"
They should put those snakes on a plane... to do maintenance in tight quarters, just like he said! Then they could put a robot Samuel L. Jackson on there to keep an eye on the snakes.
Fantastic study on snakes body mechanics and a great way for people to learn how intricate and complex their snakes truly are something so many snake keepers spend little time learning about !!! :) I feel truly honored to have found this channel and this video having raised snakes (mostly larger constrictors) for 58 years now :)
Looks like the snake is so comfortable with him, it rather sticks to him and observes than tries to find a place that isn't spinning wildly all the time.
"Anyone can string a bunch of snake motors together to make a snake robot." I understand why he said that, but controlling and powering a bunch of servos is non-trivial. (Though I suppose it is trivial for a roboticist.) I haven't tried a snake robot yet, but my hexapods were a challenge to power and control. Thanks for the video Conputerphile.
There is another movement mode related to the concertina mode: coil and spring. We see this when a cobra strikes. I started building serpentine robots back in the mid 90s, and I didn't have access to these cool servos. Instead, I used NiTiNOL tensile springs, four per universal joint. They're pretty slow to cool and stretch, but respond very quickly and contract when you put current through them to heat them up, and can move several hundred times their own mass. They act a lot like muscles. I also didn't have Python back then in the olden days, and was limited by my budget to using 8 bit CPUs, in particular the 8085. I've spent quite a bit of time developing an AI system that operates on 8 bit CPUs or microcontrollers, which I call FUNGAL - the FUzzy logic/ Neural network /Genetic Algorithm Language. It has taken me nearly 30 years, and of course Moore's law has long passed me by, but I believe we have only scratched the surface of what an 8 bit computer can do. Sometime in the next few years I'll release the language so other people can play with it on microcontrollers or retrocomputers.
Snakes are my #1 fear. I hate them and they give me the creeps like nothing else ever has....but this video was fascinating. After the video I can understand why a snake-like robot would be beneficial. Dr. Astley was very well spoken and you can tell he has a real interest and enthusiasm for his work which is admirable.
A Lake Michigan area accent? after being in England too long? ;-) Great subject matter! very interesting, and fun, let us know when it's on batteries and escapes!
Can you make muscle cells, then link together like chain mail into any shape. If each cell starts null unnamed then ai could name each cell later and then learn to move in a simulated environment
i think to solve the problem with energy, u can make a tiny battery for each one 'cell' of snake, and each will works separated, its a lot of battery, but its a thing what u can think about.
Sure, Snakes are good at going through cramped corridors. But I think Computer Science should turn their attention back to the noblest yet humblest of reptiles, the Turtles. Computer Science has modeled basic turtle movement on surface - turns, movement back and forth, and ultimately inconsequential things like holding a pen. But millions of years of evolution prepared Turtles to be some of the most epic tunnel diggers of the Reptilekind. Why is this not modelled yet? That's the real interesting field for scientists who want to break new ground.
You're referring to the mode of snake propulsion called rectilinear locomotion. The scales don't move themselves but they are shaped to engage with the ground to resist backward motion allowing forward movement.
