Page 304 - Handbook of Biomechatronics

296 Andres F. Ruiz-Olaya et al.

estimated based on EMG signals. The hip-knee-ankle-foot (HKAF) lower-
limb exoskeleton presented by He and Kiguchi (2007) has been designed to
assist the movements of physically weak people. It consists of one passive
DOF for the ankle d/p, and two active DOFs for the hip and knee f/e joints.
The desired assistances for hip and knee movements are estimated
through an EMG-based neurofuzzy controller, from eight muscles on the
thigh. LOPES is the first application of adaptive oscillators on a lower-limb
assistive exoskeleton (Ronsse et al., 2011). It is based on a trunk-hip-knee
frame-based treadmill-mounted exoskeletons with actuated hip f/e, a/a, and
knee f/e. The RoboKnee is a knee exoskeleton designed to assist the wearer
during stairs climbing and squatting with heavy loads (Pratt et al., 2004).
RoboKnee, consists of a thigh and a shank brace, jointed on the knee
and connected by a linear SEA joint.
The exoskeletons presented above, have been designed to assist different
kinds of human lower-limb movements, such as supporting heavy loads,
ground-level walking, sit/stand transitions, squatting, ascending and des-
cending stairs, and even running. The subjects used in the studies are also
diversified, including elderly people, healthy people, people with muscular
weakness, people with lower-limb disability, or totally lost lower-limb
functions (Tingfang et al., 2015).
Even though the kinematics and kinetics characteristics of lower-limb
joints greatly differ in each kind of locomotion, in the control process,
the exoskeletons are usually divided into a series of phases: detection and
prediction of these phases are based on the exoskeleton sensory systems,
which are fundamental for the control strategy. Due to the complexity in
evaluating user’s psychological effort, in the reported examples, there are
only few works involving these indexes. In addition, due to the current pro-
totypical nature of human augmentation/assistance devices, safety and
dependability factors have been poorly dealt with. In the examples above,
very few validations including two or more continuous tasks, and at the cur-
rent level they rely on state machines and vocal commands which do not
facilitate switching between tasks, thus interrupting the user’s movements
(Tingfang et al., 2015).
Generally, there are two main issues associated with the strategy for
developing assistive technology for upper and lower extremities, with
respect to the mutual interactions: the physical interaction, that is, the
mechanical power transfer, and a cognitive interaction, for information
exchange (Pons, 2010). These two issues affect each other: a consistent
and effective mechanical power transfer is fundamental for the comfort of