It’s quite astonishing to see how far we have come in developing bionic arms over the last few decades. Dexterity is still not quite at the sci-fi level, of course. But the development of bionic arms is advanced enough to be universally functional, to the degree of even developing features that humans are unable to replicate with natural limbs.
This is all thanks to the different innovative ideas that helped bionic limb technologies inch their way closer and closer to more advanced sophistication. So, although it might still take a while to get your Cyperpunk-style power arm, current technologies will help develop bionic arms into better versions of our real limbs.
Dexterity and Adaptability: The Biological Marvel and Artificial Hurdle
The hand, and by extension the arm, is a very complex natural system that combines muscles, flesh, bones, and nerves to achieve very precise grip control, as well as to manipulate objects at the smallest increment of dexterity.
This is supported further by the sense of touch, allowing us to gauge our motion more accurately by feeling the objects we intend to hold and use. With enough practice and repetition, it becomes muscle memory. And with task memorization, we lose the need to actively look at our hands when doing something.
As such, not only do bionic arms need to replicate every degree of freedom of the typical human arm, but they also need to reproduce the same motion at the same speed our nervous system processes it. Keep this in mind, as we briefly review each individual design method and technology of post-2020 bionic arms.
Picking Up and Interpreting Signals as Motion
A more detailed explanation of each bionic arm via input type can be found in our previous article about how bionic limbs work in general. But to give a very short recap:
By sensing movement from residual muscles – also known as myoelectric bionic limbs. In this context “residual” means the remaining muscles left from where the limb is cut off. This is the most common method for making bionic limb usage intuitive. Advanced versions of this technology include targeted-muscle reinnervation-based systems, and implantable myoelectric sensors (IMES).
By directly grafting an inert and dense metal rod inside the bone – also known as osseointegration. With the metal rod (usually titanium) acting as base, the bionic limb can sense additional movement information via tactile feedback vibrating through the limb and into the bone (“osseoperception”).
By directly acquiring motion information through the brain – bionic limbs based on brain-machine interfaces are still significantly behind myoelectric systems, but at least research progress continues to inch on. It requires a very intimate understanding of how neurons fire in the brain in relation to the motion impetus (voluntarily or involuntarily).
Recreating Hand and Arm Positions
To be fair, despite the marvelous dexterity of the hand and fingers, practically speaking we can narrow down most of the commonly used holding positions depending on the general shape of the object being held. That is, so long as the opposable thumbs and supporting fingers provide enough grip pressure, a wide variety of tasks can still be done regardless of how slow the bionic arms work to accomplish them.
Standard Bend, Rotate, and Grip Motions – the simplest method that requires the least amount of motors and actuators. Technically this doesn’t even need the prosthetic to be “bionic” (by our definition). But, any myoelectric stimuli can help greatly in making the task process feel intuitive enough. Our best example of this would still be the very first bionic arm that Claudia Mitchell received from the Rehability Institute of Chicago in 2006.
Variable Grip Mode Systems – a step above simple close and opening of the hands, variable grip modes systems create different pre-set finger positions depending on what the user generally wants to do (or hold). Again, this does not strictly require myoelectric systems to work, but it is best to implement the system with it (using custom controls) to make it easier and snappier for the user to do. The Open Bionics Hero Arm is a simpler device fitting this category, combining both manual (mechanical) finger controls and myoelectric grip pressure levels.
Adapted Grip Mode Systems – for more advanced (and more expensive systems), the user may opt for something that could instead “learn” their movements, or the settings they commonly use, eventually letting the algorithm automatically adjust the bionic hand into the position. The learned grip modes still shift quite slowly, but they at least theoretically let the user come closest to restoring their previous habits. The most famous/trending example of this at the moment is the Esper Hand by Esper Bionics.
Experimental Individual Finger Movement Systems – even after all this time, research is still ongoing into bionic arms that can control individual fingers coordinatively (rather than just triggering them one by one). Developing this with the quick reflexes of a human being is the end goal of all upper bionic limb research. For now, we have the current research at the University of Michigan, repurposing an older LUKE Arm prototype to develop myoelectric sensor systems that can reproduce very minute hand and finger commands.
Hardware Replicating Hand Motion
Recreating the hand motion itself requires a set of microcontrollers that send specific instructions to the components that act as muscles, such as motors and actuators. The microcontrollers themselves coordinate individual movement instructions by a primary circuitry embedded with one or more processors. These integrated chips are then tasked with translating the interpreted signals in the first place.
Simply put:
- Detect sensory signals from all available sources
- Translate sensory signals into movement commands
- Relay these commands as adjustment instructions to each bit of hardware in the arm
Take note that there may be other versions of servos, actuators, and adjustment motors to help facilitate a different type of motion system. German natural motion robotics extraordinaire Festo, for example, has been experimenting with different configurations for pneumatic robot arm systems over the last decade. Similarly, Clone’s artificial muscles also aim to replicate the natural grip maintenance and adjustment of human fingers integrated valves all around the hands and arm.
Why the Arm Obsession? Why Not Specialize Using Specific Tools?
We have briefly mentioned in a previous article that creating specialized limbs for very specific situations is one way that regular prosthetics can keep up with advanced bionic arms. And so technically, we don’t even need to recreate the human arm point-per-point.
So why the almost obstinate insistence on creating perfectly human-like bionic arms?
Well, function restoration is one thing. After all, you’d want patients to get on with their lives quickly, as if the limb was never lost. But more importantly, recreating human hands and arms maintains our primary gateway of physical interaction towards a human-centric environment. This is the same philosophy used in human-like robots. The urban, populated world that we live in is inevitably created by and for humans. Using human tools, navigating man-made obstacles, and accessing places built by humans, all require the gift of human dexterity and mobility.
