Control in Bioengineering:
Applications to the Nervous System
EXOSKELETON
A powered exoskeleton or exosuit is an external suit designed to assist its users perform limb movements with much less energy. The machine is powered by a combination of engine motors, hydraulics and pneumatics.
Some of the main purposes of powered exoskeletons are to enhance strength and endurance in healthy users and replicate motor capabilities in disabled users. Powered exoskeletons have a wide range of potential uses and research are being done to probe potential applications for civilian and military purposes. [1]
The human nervous system is inherently a control system, therefore it is vital for the exoskeleton to have an in-built control method that corresponds to the nervous control system in order for the exoskeleton to operate as intended. A proper exoskeleton control system must be able to optimize the power generation by the machine components and ensure it follows the nervous signals sent through the user's muscles.
History
The idea of using an exoskeleton to augment the human body is far from novel. The earliest incarnation was a device designed in 1890 by a Russian inventor named Nicholas Yagn. While able to assist with movements, the device required the user's self-generated power to start the initial movement. [2]
It took until the 1960s for the first machine-powered exoskeleton to be developed. General Electric and the United Stated military developed a suit called the Hardiman, designed to enable its user to carry heavy weights. The suit never went beyond prototyping phase due to several crucial limitations, namely its 680-kg weight, very slow movement speed, power supply issues and lack of stability. [3]
The earlier failures did not deter military and civilian researchers in working towards the goal of creating a functional exoskeleton. Only after the turn of the millennium, the whole concept started to seem workable. Major breakthroughs in electronics and material science in the past several decades increased the feasibility of constructing powered exoskeletons. This fuelled a boom of exoskeleton developments, which continues until now. [4]
Control Mechanism
Figure 1: Comparison between the 2 control algorithms used in a prosthetic leg. The black arrows represent the footswitch control, while the gray arrows show the the proportional myolectric control path. [4]
Figure 2: The main difference between footswitch and proportional myoelectric control is higlighted in the simplified block diagrams above. In proportional myoelectric control, the control signal is directly related to the motor signal sent from the CNS to the physiological muscle. Consequently, the efferent copy can allow the CNS to estimate the behaviour of the artifical muscle more accurately. [5]
Having a proper control algorithm in a robotic exoskeleton is crucial as the wearer will need to be able to take control of the various components of the suit. Controlling the limbs is especially vital as mobility depends on them. The most widely used control method employed in contemporary exoskeleton systems is proportional myoelectric control.
Direct proportional myoelectric control is used to move the lower limbs of an exoskeleton. There are several steps of processing done by the controller:
1. The controller, either an in-built microcontroller or in an external computer, receives electromyography (EMG) signals detected from the user's leg muscles.
2. Noise (i.e. movement artifacts) in the signal is filtered using a high-pass filter. A low-pass filter is then applied to smooth the signal, giving a normalized reading that can be used for actuation. As the control is proportional, it is important to ensure the gain of the controller is appropriate.
3. The processed EMG signal is delivered to the actuator. The power received by the actuator will be proportional to the amplitude of the processed signal
4. Torque is applied by the actuator on the pneumatic muscle(s), creating movement.[4]
This control algorithm follows a similar pathway as human's neural signals. Users are able to walk in a more natural pattern compared to users of exoskeleton with a footswitch control, which relies heavily on motion instead of nervous signals.5 A dorsiflexor inhibition rule can be added into the proportional control to limit the effects of co-activation of antagonistic pneumatic muscles, allowing the user to move more easily and naturally.[5]
References
1. Eveleth, R. 2015. The Exoskeleton's Hidden Burden. The Atlantic, [online] 7 August.
2. Yagn, N. 1890. Apparatus for Facilitating Walking, Running and Jumping. United States. Pat. 420,179.
3. Kellner, T. 2010. The Story behind The Real "Iron Man" Suit. General Electric Company, [online] 23 November.
4. Cain, S.M., Gordon, K.E. and Ferris, D.P. 2007. Locomotor adaptation to a powered ankle-foot orthosis depends on control method. [online] Journal of Neuroengineering and Rehabilitation, 4:48.
5. Ferris, D.P., Gordon, K.E., Sawicki, G.S. and Peethambaran, A. 2006. An improved powered ankle–foot orthosis using proportional myoelectric control. [online] Gait & Posture, 43(4), 425-428.