There's probably some engineers somewhere who thought of them, but I had a few ideas for various things:
Instead of directly motorizing joints or using hydraulic pistons for exosuits/exoskeletons, why not mount motors where it's convenient and use the same system that muscles use by attaching the wire on the target limb from the front of back? For instance, leg movement. There could be four, six motors mounted on the thigh - two or three per direction. Each motorized group drives a spool/arm to which some steel wire (possibly knotted if it's a single strand for both directions) is connected - one anchor point about two or three inches below the front of the knee, one anchor in the middle of the calf at the back. The motors move the bar/wheels, which pull at the limb. More efficient than motorizing joints, less bulky than pneumatic/hydraulic systems. It would still need a power source, but unless I'm missing something a good power pack could power it for a few hours (or maybe even a small streamlined engine, or something related to my ideas below) considering that, since it doesn't pull close to the joint, it should require less power to do work. Plus, with this system, it should be relatively easy to implement a lock system - just immobilize the bar or wheel and the thing stays as it is, without needing power. And for heavier applications? The motors, instead of driving the pulley directly, drive a flywheel - for instance, on legs, the flywheel to put the legs 'down' could rev up while the leg goes up. And I'm not entirely sure that it's not possible to use gravity or inertia to have regenerative recuperation somehow).
Another idea is general spring-power. I have no idea how efficient or energy-dense it would be, but I was thinking we could use springs (considering we already use flywheels) for energy storage. For that, I specifically thought of a few nested or stacked torsion spring designs. The first one was, in essence, solidly linking the ends (middle or outer) of oppositely-placed torsion springs (because I noticed that, in torsion springs, linking two aligned springs is much like moving the middlepoint of two stacked linear springs) and continuing to place them in series to make what is, in essence, a more compact and manageable but still huge torsion spring. Varying the girth and length of springs could power different applications - you could place a 6-inch spring along almost the whole length of a car, connected to a gearbox. Rewound by a port at the front, and it EASILY could implement regenerative braking by having a system either have another, lighter drive bar underneath it that feeds the spring back when the drive is disengaged and the car is still moving, or simply having a system put the gearbox in reverse while trying to stop. A smaller but wider two-layer spring, coupled to a small gearbox (or flywheel system) to produce power.
Third idea was for several form of wheel-motors for vehicles, and how to make them usable in areas laden with potholes, or offroad, or other suboptimal conditions. I probably have the concept drawings somewhere...
Also, somewhere I have some documents on theories on how we could make cars that handle like the ones in 'I, Robot'. Along with a nifty three-wheeler three-to-five seater sedan concept to go with it, including a few configurations (most of them electric, but I managed to klugde an engine drive in there somehow - most of the ideas on having universal steering aren't friendly to driveshafts)
ALSO, I once thought about the EMdrive and I realized that the reason why it might just work is because of radiation pressure - it literally uses photons/an electromagnetic field as a propellant (since it's open only on one end, IIRC, it pushes asymmetrically on the container which causes thrust - the particular shape might just be the ideal shape for containing and amplifying (perhaps even causing resonance) at the given frequencies. And if that's the case - pumping enough energy into it might just make it a vacuum laser).
Also, while I didn't have the mathematics down and there was a few flaws, I came to the conclusion that the reason our mathematics were all wrong is because we adhere to a vehemently cartesian point of view - while we should at LEAST use a polar, if not angular point of view - since everything is relative, we should use relative measures (and from close enough, any curve becomes straight, and because of optics any 'straight' line is actually the curve to an unfathomably large curve (and that was also the solution to the coastline issue - it's not that the coastline was infinite, it was just that our measurements weren't curved) that changes according to your point of view - and constant velocity in a straight line just does not exist in the universe - anything going at a constant cartesian speed finds itself either accelerating or part of a curve around something). Then I realized that most, if not all astronomical (including GPS) calculations were made by using angles... and converting it to cartesian metric when possible.
Another of my musings: there is life, in the galaxy, without any doubt. However, I learned somewhat recently that relativity broke simultaneity - which means that, if FTL is truly impossible, the reason why we never have and most likely never will encounter alien life is that for all we know (and I'm pretty sure it's the case, considering the whole galaxy spins like a disk instead of like orbits) the other star systems/the whole galaxy is going 'faster than c' in classical physics, which means that we're all going at relativistic speeds, which itself means that it's not impossible that by the time anyone gets anywhere (even at close-to-light speeds), everything has already either died or left... which is also why I do not believe that extra-solar colonies will, at best, become a new Earth (because by the time they get there, Earth will most likely be vacant again) and they most likely will have tremendous issues, if not be unable, to contact the other colonies. At worst? Well there IS a theory about extra-terrestrial origins for humans...
On a more lighthearted note, it just makes more sense to me that the velociraptor (and other feathered raptors) had them for aerodynamics and controls (I mean - if I had to design a vehicle that couldn't entirely depend on traction, had to deal with very rough terrain and had to make relatively sharp turns at 60 km/h, plane control surfaces DO seem like a good idea even if the thing isn't going to fly).
And on an even more 'lighthearted' note, the song "What if God was one of us" never EVER mentioned that if God really was roaming among men, he could totally be an African warlord, the President of China, a Saudi oil tycoon or the CEO/a chairman of Wal-Mart/Monsanto/Exxon depending on how fancy he felt. Or he could be a famous actor ... OR maybe that song was not just a song but a statement, and God got tired and passed away in 2003 who knows. Just saying - if God was roaming among men, it doesn't mean he's going to be an average person.