If cats are a good thinking aid for you, this seems very similar to the Falling Cat Problem:
From wikipedia:
"The apparent contradiction with the law of conservation of angular momentum is resolved because the cat is not a rigid body, but instead is permitted to change its shape during the fall owing to the cat's flexible backbone and non-functional collar-bone."
So a scientific paper claims to have invalidated known physics. History tells us this will go one of two ways. Either, the authors will get a Nobel prize (eventually) or it will turn into a cold fusion level debacle.
I leave you with a quote from the lead researcher with commentary from Cosmos magazine[1].
“ We let our shape-changing object move on the simplest curved space, a sphere, to systematically study the motion in curved space,” says lead researcher Zeb Rocklin, assistant professor in the School of Physics at Georgia Tech. “We learned that the predicted effect, which was so counter-intuitive it was dismissed by some physicists, indeed occurred: as the robot changed its shape, it inched forward around the sphere in a way that could not be attributed to environmental interactions.”
To make sure that the effects induced by the curvature of the robot’s space dominated, the physicists had to isolate the system as much as possible from external forces. Only then could the team ensure minimal interaction or exchange of momentum with the environment.
The curved space was produced by placing a set of motors drive on curved tracks. The tracks were then attached to a rotating shaft to produce a spherical space.
Friction was curtailed using air bearings and bushings – low heat and low mess alternatives to ball bearings. Gravity was diminished by aligning the rotating shaft with Earth’s gravity.
——
Edit: There is some critical commentary in The Register’s comments section[2].
Could this principle be used to make a reactionless drive for spacecraft? I would assume it would only be effective in an absurdly deep gravity well, like right next to a star, so it's probably impractical; I'm just curious.
Sort of, but not really. This is like reorienting yourself in midair by waving your arms in that it doesn't change your velocity, just your position. So while something like an ion engine can get away with absurdly low accelerations because they build up to large values over time, this gives you an extremely low velocity with no way to increase it. And in space with a high enough degree of curvature for this to work, you need to be moving at a very high speed in the first place to not immediately fall into the source of the curvature.
Does this also work for curved space time? Or is that type of curvature not going to help because it's relativistic? Or is the shape wrong? They mention that it doesn't work for cylinders.
The abstract says "the noncommutativity of translations permits translation without momentum exchange in either gravitationally curved spacetime or the curved surfaces encountered by locomotors in real-world environments."
The idea isn't new. This is an experimental verification of it in the case of curved surfaces, not gravitationally curved space, but it should work for both.
IANAP, but I think conservation laws are always equivalent to some symmetry (space translation, time translation, rotation, etc), so if one of those symmetries doesn't hold in a curved space, the corresponding conservation law just doesn't hold.
This seems similar to one of the kids playground toys where you free spin (think record player with handlebars kind of thing) and I am pretty sure you can get somewhere you want (angle wise) by jigging side to side with technique, but I have never though about the physics of it.
From wikipedia: "The apparent contradiction with the law of conservation of angular momentum is resolved because the cat is not a rigid body, but instead is permitted to change its shape during the fall owing to the cat's flexible backbone and non-functional collar-bone."
https://en.m.wikipedia.org/wiki/Falling_cat_problem