Design and optimization of a robotic gripper for the FEM assembly process of vehicles

Highlights

• A robotic gripper based on a six-bar planar linkage mechanism is designed and optimized.

• Multi objective optimization is introduced for both path and torque requirements.

• Toggle mechanism is used and the drive torque is minimized for grip safety and drive efficiency.

• To grip the FEM safely over a wide operating range, the coupler path error should be minimized.

• Maximum path error and drive torque of the final solution correspond to 13% and 82% of the prototype design.

Abstract

The FEM (Front End Module) assembly process consists of lifting, positioning, and bolt tightening procedures in the automotive assembly line. This process requires operations for handling heavy objects precisely and repetitively. Powered wearable robots, which are regarded as human-robot cooperation systems, are expected to solve the difficulties and improve the productivity by combining the power of robots and human intelligence. Grippers are required for performing processes by a wearable robot, but conventional grippers or overhead type loaders cannot be used for this application in terms of their weight, size, and gripping force. Thus, we focused on the design of a special gripper for a wearable robot. A six-bar linkage incorporating a toggle mechanism is employed to reduce the overall weight of the gripper while maximizing the gripping force. For gripping a FEM over a wide operating range, it is necessary to follow the desired coupler path. To satisfy the requirements of both coupler path and drive torque, a multi-objective optimization approach is introduced. As a result, an optimum design is selected from the Pareto front, which satisfies the requirements of path tracking and torque capacity.

Introduction

A powered exoskeleton or a wearable robot is a robotic system worn by a human that assists or augments the movement of human limbs by controlling the actuators. It typically has the same degrees of freedom (DOF) as the human joint to generate or assist the desired motion by applying force or torque to the wearer's joint. Researchers have been designing such systems for military and medical applications since the 2000 s [1], [2], [3], [4], [5]. Due to the capacity to enhance the strength of the human operator, the application field is also extending into the industrial field [6], [7], [8]. Most modern vehicle assembly lines are composed of automated processes utilizing industrial robot arms, but some processes are still done by human operators [9]. For complex and diverse tasks where human flexibility and intelligence are required, human-robot cooperation can lend benefits in terms of productivity. The FEM (Front End Module) assembly process is one of the processes where collaboration is needed. The FEM shown in Fig. 1 is composed of a multi-piece assembly including a front carrier, radiators, cooling fans, air conditioning condensers, bumpers and headlights, etc. [10]. Workers use a FEM loader [8] as labor saving equipment to carry the FEM and assemble it in the front part of a vehicle. In the existing assembly process, the FEM loader is mounted and operated on a hoist that is installed on the ceiling in an assembly line. A gripper that can perform the same function as the FEM loader should be designed for the wearable robot.

Generally, a gripper is designed considering the shape, size, and weight of a workpiece to be manipulated. To manipulate objects of various shapes, some grippers are designed to resemble human fingers [11], [12], or flexible structures are used [13], [14]. The conventional grippers in the industrial field are not suitable for the FEM assembly process because of the heavy weight and large volume of the FEM. Meanwhile, a linkage mechanism is used in many mechanical systems to convert the simple motion of actuators into complex motion and amplify the force and torque of an actuator. These characteristics of the linkage mechanism are suitable to generate motion to hold a heavy object, and enable it to be fitted in the complicated shape of the FEM with a small actuator.

There are several studies in which linkage mechanisms are used to generate the desired trajectory or amplify the torque of an input actuator. Bai et al. developed a Meso-Gripper that can grip objects in industrial assemblies within a range from 0.5 mm to 100 mm by integrating the series of linkage mechanisms [15]. Zhao et al. proposed a unified design formula for planar four-bar linkages with n specified positions and designed a robot that can imitate a flapping wing motion of a bird [16]. Tlegenov et al. developed a 2-DOF under-actuated robotic finger with a servo motor by inserting a passive element between the links [17]. Gezgin et al. developed a hand rehabilitation robot using the Watt II six-bar structure to generate grasping motions of a healthy person [18]. Shao et al. designed a cam-linkage mechanism for gait rehabilitation to generate the complex trajectory of a human foot by using a motor with constant rotational speed [19]. However, in the above studies, the drive torque for operating a linkage mechanism has not been considered simultaneously while it is designed to generate the desired path or motion. The dynamic analysis results were just calculated for a simple confirmation after the kinematic state of the linkage was determined. Meanwhile, Park et al. used a six-bar linkage mechanism to amplify the input torque by a toggle configuration for high clamping force [20], but the path of the output link was not taken into account in their application.

In the design of a robotic gripper of a wearable robot for handling the FEM, it is necessary to consider the actuator torque to drive the linkage mechanism as well as the path of the motion. Multi-objective optimization algorithms are suitable for such problems that have more than one requirement. Many researchers have used multi-objective optimization algorithms to find optimal solutions. Chiandussi et al. analyzed four multi-objective optimization techniques with five benchmark problems, and then applied the techniques to an engineering problem where the mass of the system is minimized while its natural frequency is maximized [21]. Toghyani et al. presented a multi-objective optimization process of a Sterling engine to increase the efficiency and output power and reduce the pressure drop with four decision variables [22]. Here the optimal solutions were selected by three different decision-making methods and their results were compared. Nariman–Zadeh performed Pareto optimum synthesis of a four-bar linkage considering the conflicting objectives of tracking error and deviation of the transmission angle [23]. Datta et al. presented a robotic gripper problem considering the model of an electric motor. After finding the Pareto front of the objective functions, the optimum value of input voltage required for gripping force was analyzed [24]. Hassan and Abomoharam proposed a process of optimizing robotic systems and illustrated a case study of a robot gripper mechanism with an optimization process to minimize the gripping force variation and to maximize the force transmission ratio [25].

In this study, we propose a robotic gripper with a six-bar linkage mechanism to follow a desired path with minimum drive torque. A multi-objective optimization algorithm has been introduced to satisfy both of requirements. This paper is organized as follows. In Section 2, the design concept of the proposed gripper for the FEM assembly process is described. In Section 3, kinematic and force analyses of the linkage mechanism are performed to establish the mathematical models. In Section 4, an optimization process of the linkage mechanism with two objective functions is performed and the optimization results are analyzed. Finally, the conclusions of this study are summarized in Section 5.