In Herr’s paper, a knee prosthesis is presented which automatically adapts knee damping to the gait of the amputee using only local sensing of knee force, torque and position. The local sensing refers to all the sensors positioned close to the knee axis allowing the amputee to employ vertical shock pylon technologies critical to overall prosthesis shock absorption. The controller cycled through the state machine as the user moved through each gait cycle. At each stage of the gait cycle, certain conditions must be satisfied in order to move from state to state.
Through this iterative process, the user adaptive controller determines how swing phase damping could change with foot contact time or walking speed. Stored within the memory of the knee’s processor is the full biological range of foot contact time. One key piece in this research is that the external knee prosthesis must provide stance control to limit buckling when weight is first applied to the prosthesis. In addition, a knee prosthesis must provide swing phase control so that the biologically realistic leg dynamics emerge during swing.
One issue that I noticed with this device is the same issue that plagues all modern mobile devices: batteries. The situation is certainly a grim one where a user with one of these devices needs to go back home to recharge or hook their leg up to a car charger. In evaluating the future work for this device the development of improved power supplies and more efficient knee actuator designs where both joint dissipation and mechanical power generation can be effectively controlled in the context of low-mass, high fatigue-life, commercially viable knee prosthesis. The improvements need to come in power supplies, knee actuation strategies and distributed sensory architectures.
This research demonstrated that the user-adaptive knee successfully controls early-stance damping, enabling amputees to undergo biologically-realistic, early-stance knee flexion. This gives demonstrable proof that the user-adaptive control scheme and local mechanical sensing area all that is required for amputees to walk with an increased level of biological realism compared to mechanically passive prosthetic systems. This is an important point to note about this research because mechanical/computer based implants are going to become more common place within the coming decades. The user needs to feel that the device is a part of them and not trying to control their movements. Even with implants, there is a social implication that if the person does not move like a human with the implant, than they are not normal. For any of these prosthetics, it is important that it acts as a direct replacement to their original biological counterpart and not as a impedance. Further research will most likely exploit this and make further refinements upon the devices so that the implants provide even more performance advantages over their original body parts.
Reference:
- Herr, Hugh. “User-adaptive Control of Magnetorheological Prosthetic Knee.” <http://biomech.media.mit.edu/wp-content/uploads/sites/3/2013/04/User-adaptive-control-of-a-magnetorhelogical-prosthetic-knee.pdf>
