Kousyun Fujiwara

Scissor lift with real-time self-adjustment ability based on variable gravity compensation mechanism

Most robots involved in vertical movement against gravitation require actuators large enough to support their own weight. To improve the inherent safety of such robots against the large actuators and reduce their energy consumption, numerous gravity compensation mechanisms (GCMs) have been proposed. Our previous study proposed a variable GCM (VGCM) that uses two types of springs and can adjust the compensation force. In this paper, a VGCM-based scissor lift (pantograph lift) that uses three springs and a smaller actuator is proposed. A prototype is designed and fabricated, and the performance of the prototype is evaluated experimentally. The results demonstrate that the developed scissor lift meets the design specifications. In addition, a load estimator is established based on the dynamic model of the scissor lift. A real-time self-adjustment method that automatically changes the compensation force is proposed, and its effectiveness is verified. Graphical Abstract

COMPLIANCE CONTACT CONTROL OF MECANUM WHEELED MOBILE ROBOT FOR IMPROVING SAFETY

Recently, robots sharing spaces with humans have been spreading. Therefore, since the situations in which the robots physically contact with humans and/or environments will increase inevitably, the kinetic control in consideration of the circumference is needed. In general, the non-contact sensor is used on mobile robots. However, since such a sensor has a dead angle and it may break down, another way improving the safety is required. Although a force or tactile sensor may be another solution, a lot of sensors are needed in order to compensate a dead angle. Therefore, in this paper, we propose a method to improve the safety by estimating the external force applied on the wheeled mobile robot and by using compliance control based on virtual conveyor. The proposed method doesn’t need an additional cost of sensor, and has an advantage because of having no dead angle about driving directions of the robot.