EMG-Based Control of a Robot Arm Using Low-Dimensional Embeddings

Abstract

As robots come closer to humans, an efficient human-robotcontrol interface is an utmost necessity. In this paper, electromyographic (EMG) signals from muscles of the human upper limb are used as the control interface between the user and a robot arm. A mathematical model is trained to decode upper limb motion from EMG recordings, using a dimensionality-reduction technique that represents muscle synergies and motion primitives. It is shown that a 2-D embedding of muscle activations can be decoded to a continuous profile of arm motion representation in the 3-D Cartesian space, embedded in a 2-D space. The system is used for the continuous control of a robot arm, using only EMG signals from the upper limb. The accuracy of the method is assessed through real-time experiments, including random arm motions.

Building the Ninapro database: A resource for the biorobotics community

“…The success is due to the low inertia of the hand joints, thanks to which a correspondence between isotonic and isometric hand muscle configurations and positions can be established. (Attempts at mapping sEMG to whole arm position have been successfully made [14]- [17], which makes this statement even stronger.) The first example of this technique is probably [18], dating back to 45 years ago.…”

Continuous Estimation of Human Multi-Joint Angles From sEMG Using a State-Space Model

“…Refer to [9], a group of eight muscles that are mainly responsible for the studied motion is recorded: biceps brachii, triceps brachii, pectoralis major, deltoid (anterior), deltoid (middle), deltoid (posterior), teres minor, trapezius. A wireless sEMG system (Delsys, Trigno) was used to measure raw sEMG signals with a frequency of 2000 Hz, and the placement of the eight sEMG sensors is illustrated in Fig.…”

“…However, classification is only able to identify a limited number of discrete motion patterns from sEMG. But recently, instead of the discrete patterns, how to determine the human's intent of a continuous motion from sEMG has become an active issue [8], [9]. This is due to the requirements of several possible applications, such as powered prosthesis, exoskeletons and rehabilitative robots.…”

“…Normally, an EMG-driven forward dynamics model always consists of muscle activation module, Hill-type muscle contraction dynamics, musculoskeletal geometry, and joint forward dynamics [12]. However, this model is very complex and involves a large number of unknown physiological parameters, which makes the analysis and application of the model rather difficult, so only a few DOFs (1 or 2) of human joints have been successfully analyzed by the model [9].…”

Continuous robot control using surface electromyography of atrophic muscles

“…The initial ideas stemmed from prosthetic hand robotics, and various publications on the matter [10], [11], [12], [13], [14] show that EMG can serve as an excellent means to control dexterous hand prostheses. Also arm position reconstruction can be done through EMG, either using models of the arm geometry [15], [16], [17] or without [9]. The latter approach, which controls the position and orientation of the robot end-effector in 6D, leads to a nonnegligible error in absolute end-effector positioning, but by controlling endeffector velocity rather than position [18] the system can be used with considerable accuracy to grab objects.…”