Project Manager: Dr. Rafaella Carloni
People with moderate or severe upper extremity disabilities, caused by various diseases or injuries, find themselves in need of continuous personal care, since even the simplest task cannot be performed independently. This complete dependence on others has a tremendous impact on the impaired person’s quality of life and on societal healthcare expense. Therefore, robotic aids to assist impaired people is becoming increasingly popular. Unlike task-restricted devices like feeding-aids and pageturners, multi-purpose robotic arms can assist the impaired in a variety of tasks and are often mounted on a wheelchair for increased mobility. However, since these arms are traditionally powered by stiff actuators, interaction safety is not easily guaranteed, because interaction control is not robust due to control bandwidth limitations. The novel variable impedance actuation technology enables the adaptation of physical actuator output impedance (e.g. stiffness) independently from the actuator output position. This means that the physical impedance can be tuned to instantaneous safety requirements, thereby providing intrinsic safety by means of physical elasticity.
The goal of this project is to research and develop an intelligent variable impedance arm, aimed at assisting physically impaired people. Intelligence that can be embedded is, for instance, the autonomous tuning of impedance based on specific task intentions and exploiting the internal actuator elasticity for more energy efficient motion control.
Results in 2013
modeling of a variable stiffness actuator in graph theory has been studied and
optimal control methods for these models have been researched. A compliant
robotic system model composed of multiple variable stiffness actuators has
been made to analyze the optimal stiffness distribution along the complete
chain. This 1D Euler-Lagrange model was extended to a 6D Port-Hamiltonian model
to be able to analyze a genuine manipulator. Moreover, a novel actuator, as an
evolutionary design on the vsaUT-II, has been researched and developed.
Results in 2012
beginning, a literature survey on current assistive robotic arms has been
performed. Moreover, possible control methods inspired by human extremity
control were examined and the possible route for optimizing a robotic
manipulator with regard to the actuation type (stiff, series elastic and
variable stiffness actuation) and arrangement of actuators (uni- and
bi-articulation) was researched. The mathematical background needed in the
optimization route is investigated to gain a better understanding of the
subject. This entailed research on port-Hamiltonian systems in general and
specifically the port-Hamiltonian modeling of rigid bodies, kinematic
constraints (joints), physical springs, stiff actuators and variable stiffness
Project duration: 1-1-2012 / 1-1-2016
Project budget: 270 k-€ funding