Introduction to Parallel Robotic Arms
Parallel robotic arms are a sophisticated branch of robotics that serve a variety of industrial, medical, and research applications due to their exceptional precision and load capacity. These mechanisms are constructed based on the parallel kinematic principle, where multiple arms (or legs) support and manipulate a common end-effector or platform.
Parallel architectures, initially designed for tire-testing machines and flight simulators, have expanded into various applications involving heavy load manipulation with high acceleration, including simulators used in vehicle driving. These structures are increasingly popular in manufacturing due to advantages like high structural rigidity and dynamic performance over serial robots, influencing the design of new high-speed machine tools. Compared to serial robots, these robots are generally characterized by higher speed, strength, and precision. However, their workspaces are more constrained and complex due to the presence of both serial singularities and parallel singularities. Serial singularities represent configurations where the end-effector loses one or more degrees of freedom (DOF), defining the boundaries of the working modes, each corresponding to one of the four solutions of the inverse kinematics problem. On the other hand, parallel singularities occur in configurations where the actuators are unable to compensate for external forces or moments applied to the end-effector, subsequently reducing the size of the workspace.
The example provided illustrates a fully parallel robot with six degrees of freedom, where the movement of the mobile platform is controlled by six legs that connect it to a stationary base. Adjustments to the lengths of the legs, achieved through actuated prismatic joints, allow for the precise positioning of the mobile platform. The Stewart platform consists of a base and a moving platform connected by six variable-length struts arranged in parallel. This configuration allows the platform to move with six degrees of freedom: translating along the x, y, and z axes, and rotating about the roll, pitch, and yaw axes. The precise control of the strut lengths enables the high accuracy and stiffness that make Stewart platforms suitable for tasks requiring intricate motion control, such as flight simulators, telescopes, and surgical robots. This design concept has historical significance, having been utilized by Gough in 1947 for tire-testing machines, and later influencing Stewart to create a flight simulator in 1965. This particular design framework is recognized as the Gough-Stewart platform or parallel manipulator robot.
Focusing on the 5-bar parallel robotic arm design, this simplifies the concept by utilizing five bars in parallel fashion to control the position and orientation of the end-effector. Essentially, it’s a specialized version of parallel kinematics that reduces complexity while still offering many of the advantages, like high-speed operation and increased rigidity compared to traditional serial robotic arms where joints are connected in series. The five-bar design’s limitations in degrees of freedom are often offset by its simpler control algorithms and reduced cost, making it a desirable choice for applications that do not require the full mobility of a Stewart platform, such as pick-and-place tasks, 3D printing, and lightweight machining operations.
While less versatile than the six degrees of freedom offered by the Stewart platform, the five-bar parallel arrangement (also known as five-bar linkage) is optimized for planar tasks, providing two translational movements and one rotational movement. This efficient design results in a robotic arm that is both swift and accurate, making them particularly well-suited for high-speed, repetitive tasks on a single plane.
Introduction to 5-bar linkage and Literature
Five-bar robotic arms (aka five-bar linkage) belong to the family of parallel manipulators, distinguished by their unique kinematic chains that allow for precise and rapid movement within a specified plane. Unlike serial manipulators, where each joint connects in a “series” leading to the end-effector, the 5-bar design features two parallel chains that connect the base to the end-effector.
The five-bar robotic arm consists of five rigid linkages or members: two arms that connect the base to two interconnected arms, which in turn connect to the end-effector. The relative movement of these bars is typically controlled by two servo motors located at the base. This configuration allows for two rotational degrees of freedom and enables the end-effector to cover an area within the arm’s reach with high precision and speed.
Designed with an emphasis on simplicity and efficiency, five-bar parallel robotic arms offer several advantages. The inherent mechanical stiffness of the design, due to the parallel bars working in concert, contributes to the machine’s high precision. Moreover, the setup allows for excellent payload capabilities relative to the arm’s weight and size. Moreover, because the motors are stationary at the base—unlike in serial configurations where they might be distributed along the arms—the five-bar robots can achieve faster acceleration and deceleration with less energy consumption.
Applications of 5-bar parallel robots
Applications for five-bar robotic arms are diverse, ranging from industrial automation where they perform tasks such as sorting, assembly, and packaging, to fields like animation where they facilitate motion capture. There are many academic research examples as well, such as geometrical hierarchy trials, dynamic performance analysis and trajectory optimizations.
Their reliability and speed also make them suitable for precise tasks in harsh environments where human presence might be impractical or unsafe. Given their operational efficiency and cost-effectiveness, five-bar robotic arms continue to be an area of interest in both academic research and industry innovation.
Some Literature for 5-bar linkage robots
A range of studies have explored different aspects of the five-bar parallel robot.
Campos (2010) focused on increasing the robot's usable workspace by proposing a design with equal-length links. Five-bar planar parallel robots for pick and place operations are always designed so that their singularity loci are significantly reduced. In these robots, the length of the proximal links is different from the length of the distal links. As a consequence, the workspace of the robot is significantly limited, since there are holes in it. In contrast, they propose a design in which all four links have equal lengths. Since such a design leads to more parallel singularities, a strategy for avoiding them by switching working modes is proposed. As a result, the usable workspace of the robot is significantly increased. The idea has been implemented on an industrial grade prototype and the latter is described in detail.
Wang (2016), developed a modularized practical training table for the robot, integrating a control system and virtual operation machine. The invention relates to a modularized five-bar parallel robot practical training table. A five-bar parallel robot and a virtual operation machine are centralized on one platform and are suitable for customization with different bases; the modularized five-bar parallel robot practical training table is convenient to assemble and disassemble, reasonable in space utilization and multipurpose.
Yi (2010), addressed the interpolation algorithm for the robot's trajectory, proposing an algorithm based on the Jacobi matrix. The interpolation problems of five-bar parallel robot's trajectory with 2-DOF is studied in this paper.An algorithm which can get incremental step through small line is proposed.The interpolation algorithm of the trajectory is achieved according to the given error requirement by using the Jacoby matrix. A method to realize the movement of the five-bar parallel robot with the usual control system is put forward.The implement of prototype experiment is realized on the Mach3 system, which is based on the above proposed algorithm with the simulation experiment.The expected results is obtained.
Hong-bing (2009), presented the dynamics equations of the robot, analyzing the influence of configuration on equivalent inertia, coupling inertia, and driving torques. This paper presented the dynamics equations of planar five-bar parallel robot based on the Lagrange equation, and obtained expression of driving torques.The equivalent inertia, coupling inertia and driving torques of the active joints of mechanism were analyzed through the existing model.The results show that, for a given motion, the configuration has a significant influence on the equivalent inertia, coupling inertia and driving torques.
Bourbonnais (2015), focused on trajectory planning and control to optimize performance. This paper presents trajectory planning optimization and real-time control of a special five-bar parallel robot. Planning is based on a cubic spline stochastic approach that minimizes trajectory time and selects the best combination of working mode regions to circumvent all parallel singularities, allowing the size of the workspace to be increased. Identification of the dynamic model of the robot and its actuators allows a precise implementation of the trajectory planning and real-time control approach. The optimization algorithm achieves a fast trajectory and a controller that operates with an error of less than 0.7° for both actuated joints of the robot.
Jian (2014), used bond graph modeling to simulate the robot's dynamic performance, providing a valuable tool for analysis and design. This work focuses on the bond graph modeling method and its application on multi-body systems, especially on the five-bar parallel robot. Experiment results show that bond graphs can simulate robot dynamics performance without having to make a large number of equations. It is able to simulate and solve five-bar kinematics problems in the process.
These studies collectively contribute to the understanding and application of the five-bar parallel robot.
Controlling the Five-Bar Robotic Arm with Acrome SMD
Smart Motion Devices (SMDs) are essential in the world of autonomous robotics, particularly when it comes to the nuanced operations of 5-bar parallel robots. These sophisticated machines require an intricate level of control that SMDs are uniquely equipped to provide, thanks to their capability for real-time monitoring and dynamic adaptation. They are the maestros conducting an orchestra of motors, ensuring each note is played at the perfect pitch and tempo.
The real game-changer with SMDs lies in their ability to synchronize the control signals necessary for the precise actuation of motors, be they linear or otherwise. This is no small feat when considering the complex coordination required to emulate the fluid motion of a five-bar parallel robot. Those interested in the technical details of connecting and utilizing multiple SMDs in a single network are encouraged to delve into our comprehensive article, "Synchronizing Linear Motors and DC Motors".
In robotics, every millimeter and millisecond count, especially when five-bar parallel robots execute tasks involving highly precise trajectory movements, such as delicate placement in confined spaces. It's here that synchronous motor control showcases its value, meticulously guiding each actuator to ensure the robot follows its pre-programmed path with the grace of a ballet dancer and the precision of a surgeon.
Yet motion is not the only factor at play, as safety and adaptability are equally paramount. Continuous monitoring of motor current levels is crucial; it serves as the robot's sense of touch, alerting it to any unforeseen obstacles or resistance. SMDs come with integrated current limiting and active current monitoring capabilities, which are vital for collaborative robot (cobot) applications where human-robot interactions are common. These features ensure that cobots operate within safe parameters, preventing accidents and damage to both the robot and its human counterparts.
Lastly, the sensors: the eyes and ears of the robot. Without them, a five-bar parallel robot is blind to its environment, unable to react or interact with any degree of intelligence. The sensor add-ons provided by SMDs are not just an afterthought; they are a well-integrated solution that satisfies the robot’s need for environmental feedback while also simplifying the mess of wires that robotics can often be associated with. The daisy-chain capability provided for both motor drivers and sensors means that data flows back to the controller in a streamlined fashion, along a single consolidated line of communication.
This article serves as an example to the robustness and ingenuity of Smart Motion Devices and their usage in robotics technology. Whether you're a hobbyist looking to understand the basics or an engineer seeking to refine a complex robotic system, the intersection of SMDs and 5-bar parallel robots represents an area rich with potential and exploration.
Connection Diagram
Here is the connection diagram of the Acrome Five-Bar Robot. It contains two SMD RED boards, two DC Motors, one USB Gateway Module and a Power unit. Optional feedback with QTR, Potentiometer and IMU add-on modules can be added to the SMD network, but they are not depicted in the figure. Built-in current shunt sensors are not drawn.
In general, the USB gateway can be connected to a PC or to any single-board computer such as Raspberry Pi, Jetson Nano etc. Following, you may find an example code in Python that can run on the PC or SBCs.
Five-Bar Robotic Arm Python Code Using SMD
from smd.smd_types import Index
from smd.smd import Master
import time
import math
from 5bar_InvKin import *
if __name__ == "__main__":
# End effector coordinates
# x, y = float(sys.argv[1]), float(sys.argv[2])
# s1, s2 = calc_angles(x, y)
m = Master(4096, 'COM16') # Com port number should be altered
print(m.AutoScan())
m.send(m.SyncWrite(Index.OperationMode, id_val_pairs=[[0, 0], [1, 0]]))
time.sleep(0.3)
m.send(m.SyncWrite(Index.TorqueEnable, id_val_pairs=[[0, 1], [1, 1]]))
time.sleep(0.3)
m.send(m.SyncWrite(Index.MotorCPR, id_val_pairs=[[0, 6533], [1, 6533]]))
time.sleep(0.3)
m.send(m.SyncWrite(Index.MotorRPM, id_val_pairs=[[0, 100], [1, 100]]))
time.sleep(0.3)
m.send(m.SyncWrite(Index.PosPGain, id_val_pairs=[[0, 3.4056], [1, 10.3439]])) # can be optimized
time.sleep(0.3)
m.send(m.SyncWrite(Index.PosDGain, id_val_pairs=[[0, 6.4178], [1, 23.2778]])) # can be optimized
time.sleep(0.3)
m.send(m.SyncWrite(Index.MinimumPosition, id_val_pairs=[[0, -5000], [1, -5000]]))
time.sleep(0.3)
i=0
while True:
i=i+1
x=math.cos(math.radians(i))*4
y=math.sin(math.radians(i))*4
#plot_plot(x, y+27,m) #enable/disable for plotting
References
[1] Bourbonnais, Francis. “Minimum-Time Trajectory Planning and Control of a Pick-and-Place Five-Bar Parallel Robot.” 2015.
[2] Campos, Lucas. “Development of a Five-Bar Parallel Robot With Large Workspace.” 2010.
[3] Hong-Bing, X. “DYNAMICS ANALYSIS OF PLANAR FIVE-BAR PARALLEL ROBOT.” 2009.
[4] Jian, Shengqi. “Five Bar Planar Manipulator Simulation and Analysis by Bond Graph.” 2014.
[5] Wang. “Modularized five-bar parallel robot practical training table.” 2016.
[6] Yi, Fang. “The Interpolation Algorithm of Five-bar Parallel Robot Trajectory.” 2010.
[7] Duperret, Jeffrey & Koditschek, Daniel. (2015). An Empirical Investigation of Legged Transitional Maneuvers Leveraging Raibert’s Scissor Algorithm. 2531-2538. 10.1109/ROBIO.2015.7419720.
Author
