A novel approach to compensate
delay in communication by predicting teleoperator behaviour using deep
learning and reinforcement learning to control telepresence robot
Fawad Naseer1,✉, Muhammad Nasir
Khan1, Akhtar Rasool2, and Nafees
Ayub3
1 Electrical Engineering Department, The
University of Lahore, Lahore, Pakistan
2 Mechanical Engineering Department, Beijing
Institute of Technology, Beijing, China
3 Computer Science Department, Government
College University Faisalabad, Faisalabad, Pakistan
Email: fawadn.84@gmail.com
Robots with telepresence capabilities are typically employed for tasks
where human presence is not feasible due to geography, safety risks like
fire or radiation exposure, or other factors like any epidemic disease.
Time delay is a significant consideration in controlling a telepresence
robot. This study proposes a deep learning-based approach to compensate
for the delay by predicting the behaviour of the teleoperator. We
integrate a recurrent neural network (RNN) based on the Long Short-Term
Memory (LSTM) architecture with the reinforcement learning-based Deep
Deterministic Policy Gradient (DDPG) algorithm. The proposed method
predicts the teleoperator’s angular and linear controlling commands by
using data gathered by embedded sensors on the specially designed and
built telepresence robot. Simulations and experiments assess the
operation of the proposed technique in Gazebo simulation and MATLAB with
ROS integration, which shows 2.3% better response in the presence of
static and dynamic obstacles.
Introduction: Healthcare paramedical staff spends significant
time and effort in the sensitive and critically vital field of patient
interaction. Robotics has been employed extensively in the healthcare
environment for at last few years. The Puma 560 robot performed
neurosurgical biopsies in 1985, making it the first robotic surgery
[1-3]. The market for surgical robots has grown remarkably in recent
years and is expected to reach $15 billion by 2030, with a compound
annual growth rate of 17.1%.
ICT and robotic solutions have been developed over the past decade and
are legitimate support that could enable persons to survive freely.
Robots like Double robot from Double robotics and Nao from Softbank are
a few examples. Increasing attention towards telepresence robots has
been seen recently in creating for treating patients. Before testing the
robot in a real-life environment, several researchers conduct brief
robot assessments in a lab setting. However, other works use focus
groups or laboratory experiments to analyze the system [4-5]. The
telepresence robot generalized system is shown in Figure 1.