https://news.stanford.edu/press/view/50023?
This looks really interesting, but is it really "a replacement to traditional motors"? It's a traditional motor attached to springs and a clutch. Sounds more like it would replace the traditional transmission moreso.
I think this provides a function closer to a motor than a transmission. It has a typical electric motor at its core providing the main power, but the springs also provide power. IMO, a better analogy would be a motor plus flywheel to replace just a motor.
The main distinction in my mind is that transmissions just convert forces from one speed/torque to another; they don't store energy.
Regardless, it's still a regular motor attached to other things. It's doesn't replace the OG motor at all. And it requires much more space that isn't useful in every application.
Again, all that stuff is cool, but the title is incredibly misleading. This is an "improvement on current power transfer designs", not a "completely new motor". It's taking the power from the motor and transferring it to other areas. That's a transmission.
If I'm being pedantic, it's because the article title makes it seem like we figured out a new electric motor besides coiled wire. That would be really fascinating. But instead we get "fancy clutch". It's literally "instead of a stepper motor, let's just use a regular motor powered on 24/7 and attach a clutch and transmission to decide when and where to deliver that power".
Energy recovery was already doable with traditional actuators through BEMF generation, so the only thing this has going for it is static torque.
The only way i'm seeing this doing things at a fraction of the cost is by enabling the use of a smaller motor and repeatedly winding each spring before doing a task requiring more effort. At that point, why not just use a linear motor and even more clutches to do that winding instead of having to move the entire arm?