Servomotors bring real-time motion control to parallel-kinematics platform

A groundbreaking demonstration recently unveiled a sophisticated parallel-kinematics platform capable of real-time motion control, precisely manipulating an otherwise unpredictable ping-pong ball. This innovative system, showcasing the advanced capabilities of a new line of integrated servomotors, combines synchronized servo actuation with visual feedback and real-time kinematic computation to achieve unparalleled precision and responsiveness. The project, spearheaded by engineers like Charan Bhamra of Monolithic Power Systems, highlights the transformative potential of integrated motion control solutions for a wide array of high-stakes industrial and medical applications.

The core of this achievement lies in its ability to address and manage complex engineering challenges, including visual-feedback latency, tight synchronization across multiple axes, and the execution of real-time inverse kinematics and proportional-integral-derivative (PID) control on embedded hardware. By demonstrating dynamic balancing, swirling, and bouncing of a lightweight, erratic ping-pong ball, the platform effectively simulates real-world scenarios where sub-millimeter accuracy and instantaneous reactions are critical.

The Intricacies of Real-time Motion Control: A Ping-Pong Ball’s Challenge

Controlling the trajectory of a ping-pong ball in real-time presents a surprisingly formidable engineering challenge. Its low mass and high elasticity mean it reacts drastically and chaotically to the slightest disturbance, making it an ideal test subject for evaluating the precision and responsiveness of motion control systems. The demonstration system, featuring a three-axis parallel platform, effectively tackles this inherent unpredictability.

Parallel-kinematics platforms, often contrasted with their serial counterparts, offer inherent advantages in stiffness, speed, and precision due to their multiple, concurrent kinematic chains connecting the base to the moving platform. This structural rigidity minimizes unwanted vibrations and deflections, which is crucial when dealing with fast-changing inputs and delicate objects. The specific servomotors employed in this setup, the MMS740100-24-C2-1 from EZmotion, are central to the platform’s success. These 40-mm smart motors are fully integrated units, incorporating a permanent-magnet synchronous motor, an embedded driver module, angle sensors, and CANopen support. This all-in-one design enables high-resolution motion, instantaneous real-time response, and tight multi-axis synchronization, significantly simplifying system architecture and wiring compared to traditional discrete component setups.

System Architecture: A Symphony of Hardware and Software

The demonstration platform’s robust control system is a testament to intelligent architectural design, separating computational roles between a powerful single-board computer and a dedicated microcontroller.

Visual Tracking and Image Processing for Feedback

At the forefront of the system’s real-time capabilities is its visual feedback mechanism. A top-mounted USB camera continuously captures video frames at a moderate yet effective rate of 40 frames per second. This raw visual data is then fed to a Raspberry Pi 5 device, which acts as the primary image processing unit. The choice of Raspberry Pi 5 reflects a growing trend in embedded systems, leveraging powerful, low-cost computing for on-device analytics.

On the Raspberry Pi 5, a lightweight Python script utilizing the OpenCV computer vision library processes these video frames. The image processing pipeline involves several critical steps:

1. Gaussian Blur: Applied to smooth the image and reduce noise, preparing it for more accurate feature detection.
2. HSV Color Space Conversion: The frames are converted from the standard RGB (Red, Green, Blue) color space to HSV (Hue, Saturation, Value). This conversion is particularly advantageous for color-based object tracking, as hue (color type) is less susceptible to variations in lighting conditions than RGB values.
3. Orange Color Masking: To specifically isolate the ping-pong ball, an orange color mask is applied. This creates a binary image where the ball appears as a distinct blob against the background.
4. Contour Filtering: OpenCV’s contour detection algorithms identify the boundaries of objects within the masked image. Filtering is then applied to identify the largest contour, which is assumed to be the ping-pong ball.
5. Center Coordinate Extraction (XY Plane): From the identified largest contour, the smallest circle that completely encloses it is determined. The center coordinates of this circle provide the precise real-world X and Y positions of the ball on the platform.
6. Z-Position Determination: The ball’s vertical (Z) position is cleverly inferred from its apparent diameter in pixel units. As the ping-pong ball moves closer to the camera, its detected pixel diameter increases, and vice versa. This proportional relationship allows for a real-time translation of pixel diameter into a physical Z-position in millimeters.

For diagnostic and monitoring purposes, the Raspberry Pi 5 also connects to a local display via HDMI, providing a live rendering of the camera feed alongside the post-processed frames, visually confirming accurate ball tracking. Crucially, the extracted position and height data of the ball are then transmitted to a Cortex M4 microcontroller unit (MCU) via a high-speed Serial Peripheral Interface (SPI) communication link, ensuring minimal latency in data transfer to the control core.

The Brains: Kinematic Models and PID Control

The Cortex M4 MCU serves as the central control unit, responsible for the platform’s core control loop and the complex kinematic model. It acts as the initiator controller for the servomotors, orchestrating their synchronized movements.

The core control algorithm is built around a sophisticated PID-based system. This algorithm continuously calculates the positional error of the ping-pong ball and translates this error into a precise platform tilt angle (and Z-translation in bouncing mode). PID controllers are ubiquitous in industrial automation for their ability to maintain a desired setpoint by adjusting system outputs based on proportional, integral, and derivative components of the error.

Once the target platform tilt angle and Z-translation are determined, the MCU employs an inverse kinematic model. This mathematical model translates the desired platform pose (position and orientation) into the required individual crank angles for each of the three servomotors. The process involves several steps:

• Global and Local Joint Positions: Defining the fixed-base joints and the moving platform joints’ positions.
• Rotation Matrices: Using rotation matrices around the X, Y, and Z axes (θx, θy, θz) to define how the platform moves in 3D space. The combined rotation matrix then describes the overall orientation of the platform relative to its base.
• Actual Platform Joint Positions: Calculating the actual positions of each platform joint in global coordinates based on the combined rotation matrix.
• Limb Length Calculation: Determining the necessary length of each limb (the straight-line distance between the base joint and the transformed platform joint).
• Motor Crank Angle Computation: Finally, based on the required limb lengths and the known geometry of the crank arm and connecting rod, the inverse kinematic model computes the precise crank angle for each motor.

These calculated motor crank angles are then converted into position targets and sent to the servomotors. The motors operate in CANopen Cyclic Synchronous Position (CSP) mode, a critical element for achieving synchronized and deterministic motion. This ensures that all three axes move in perfect unison, stabilizing the platform even during rapid and dynamic control inputs.

Operational Modes: Beyond Simple Balancing

The demonstration platform showcases three distinct operational modes, each demanding different levels of control complexity:

1. Balancing Mode: The fundamental challenge. The target ball location is fixed precisely at the center of the platform. The PID controller’s primary objective is to minimize any positional error by continuously adjusting the platform’s tilt angles (θx and θy) to keep the ball stationary. This mode demonstrates the system’s ability to counteract external disturbances and maintain a stable equilibrium.

2. Circular Swirling Mode: This mode introduces dynamic trajectory following. The target trajectory is defined by a two-dimensional array of predefined, time-varying reference points that form a circle with a 100mm radius in the XY plane. The PID controller, in conjunction with the inverse kinematics, continuously guides the platform to ensure the ball follows this predefined circular path smoothly and accurately for 30 seconds.

   

3. Bouncing Mode: The most complex and visually engaging mode. The target height for the ball is set in millimeters, and the target horizontal location remains the center of the platform.

   • Oscillation Phase: To initiate bouncing, the system first enters an oscillation phase. The platform undergoes rapid, synchronized vertical movement, imparting sufficient mechanical energy to lift the ball off the surface. This phase is carefully designed to gradually increase the ball’s energy until it consistently bounces.
   • Sustained Bouncing: Once a predefined energy threshold is reached (indicating stable bouncing), the controller transitions to the sustained bouncing mode. Here, the PID controller applies minimal, synchronized vertical motion to merely sustain the bounce at the target height, optimizing for energy efficiency rather than driving the ball higher. The ball’s total energy (kinetic and potential) is estimated, and its velocity is computed from height differences over time, allowing the controller to make precise adjustments.

The Integrated Servomotor: A Paradigm Shift

The EZmotion MMS740100-24-C2-1 integrated servomotor is a cornerstone of this platform’s efficacy. Its design philosophy represents a significant shift towards simplification and efficiency in motion control systems.

Traditional servo systems often comprise separate components: a motor, an external servo drive, an encoder, and a separate motion controller, all connected by extensive wiring. The EZmotion integrated servomotor consolidates all these critical components into a single, compact unit.

• Motor: A permanent-magnet synchronous motor provides the necessary torque and speed.
• Servo Driver: Built-in to efficiently provide power to the motor windings, regulating current.
• Embedded Motion Controller: Runs a nested control loop, regulating current, velocity, and position in real-time, handling the intricate calculations for precise movement.
• Angle Sensor: A high-resolution 14-bit magnetic position sensor tracks the shaft’s position with sub-degree accuracy, providing vital feedback for closed-loop control.
• CAN Transceiver: Enables direct integration into a CANopen network, facilitating robust and standardized industrial communication.
• Power Management Solution: Ensures stable internal voltages for safe and reliable operation.

This integration dramatically simplifies wiring, reduces the system’s overall footprint, and lowers assembly time, offering a true "plug-and-play" solution for engineers. It eliminates the need for external boards and complex cabling, reducing potential points of failure and streamlining design.

Synchronized Motion with CANopen CSP Mode

The ability to achieve tightly synchronized motion across multiple axes is paramount for a parallel-kinematics platform, and the CANopen Cyclic Synchronous Position (CSP) mode is instrumental in this regard. CANopen is a widely adopted, standardized communication protocol for embedded systems, particularly prevalent in automation and robotics for its robustness and real-time capabilities.

While CANopen offers other modes like Cyclic Synchronous Velocity (CSV) and Cyclic Synchronous Torque (CST) for velocity- and torque-driven applications, CSP mode is specifically tailored for deterministic, position-based multi-axis control. In CSP mode, the central controller (in this case, the Cortex M4 MCU) sends precise position setpoints to each motor at a fixed and deterministic update rate. This ensures that all motors receive their target positions simultaneously.

Crucially, the individual motor controllers do not generate their own trajectories. Instead, they rely entirely on the initiator (the MCU) to continuously feed updated position targets. To ensure continuity and smoothness between these discrete updates, each motor internally performs linear interpolation between consecutive setpoints. This allows the motor shaft to follow a seamless trajectory, even though the position targets are sampled at discrete intervals.

From a communication standpoint, CSP mode involves the initiator broadcasting Process Data Object (PDO) messages containing the absolute target positions. These PDOs can be mapped, configured, and stored to the motor’s non-volatile memory, enhancing system flexibility. The calculated crank angle values from the inverse kinematics are thus converted into CSP position targets and synchronously broadcasted across the CAN bus. This meticulously orchestrated communication ensures highly coordinated and jitter-free motion across all three limbs, essential for the platform’s dynamic capabilities.

The Demonstration Sequence: A Controlled Performance

The operation of the parallel platform follows a well-defined sequence, designed to showcase its capabilities from initialization to advanced motion control:

1. Homing Routine: The system begins by executing a torque-based homing routine for each servomotor. During this process, all three limbs retract inward at a controlled speed until they reach their mechanical hard stops. This establishes a precise and repeatable "zero position" or reference point for all subsequent movements.
2. Initialization: Once homing is complete, the platform automatically rises to a predefined mid-height along the Z-axis, preparing for operation. A green status LED illuminates, signaling successful initialization and indicating that a user can safely place a ping-pong ball onto the platform.
3. Active Sequence Loop: With the ball in place, the manipulator enters its active sequence, which continuously loops through the three primary modes:
   • Balancing Mode: The system first demonstrates its ability to stabilize the ball at the platform’s center.
   • Circular Swirl Mode: After a period of balancing, the platform transitions to a circular swirling motion for 30 seconds, guiding the ball along a predetermined trajectory.
   • Bouncing Mode: Following the swirl, the system initiates the bouncing sequence for another 30 seconds, actively maintaining the ball’s height.
     This sequence continuously repeats as long as the system remains active, providing a comprehensive display of the servomotors’ precision and the control system’s adaptability. Throughout the entire demonstration, a 7-inch screen provides live visual feedback of the ball’s motion, allowing observers to directly witness the system’s performance.

Broader Impact: Revolutionizing High-Precision Industries

The implications of such a highly integrated, real-time motion control system extend far beyond the impressive feat of manipulating a ping-pong ball. The technologies showcased in this demonstration are poised to drive significant advancements in various high-precision industries.

Medical Robotics

In the realm of medical robotics, architectures similar to this parallel platform are crucial for enhancing surgical precision and patient safety. Integrated servomotors and real-time control can:

• Stabilize Endoscopic Instruments: Provide steady, tremor-free manipulation of tiny instruments during minimally invasive surgeries, reducing surgeon fatigue and improving outcomes.
• Guide Surgical Tools: Enable highly accurate positioning and movement of surgical tools, vital for delicate procedures like neurosurgery or ocular surgery.
• Assist in Minimally Invasive Procedures: The fine resolution and jitter-free motion are essential for tasks requiring extreme dexterity and control within confined spaces. This can lead to smaller incisions, faster recovery times, and reduced risk of complications.

Industrial Automation

The manufacturing and automation sectors stand to gain immensely from such advancements. Integrated servomotors and parallel kinematic structures are ideal for:

• Optical Alignment Stages: Achieving nanometer-level precision in aligning optical components, crucial for manufacturing sensors, lenses, and semiconductor equipment.
• Pick-and-Place Machines: Significantly increasing the speed and accuracy of component placement in electronics assembly, reducing defects and boosting throughput. Delta pick-and-place machines, known for their high speed, are particularly suited for these integrated motors.
• Micro-Assembly Manipulators: Handling microscopic parts with extreme delicacy and precision in micro-manufacturing processes.
• Delta 3D Printers: Many consumer-grade and industrial delta 3D printers already utilize parallel kinematic structures, and servo-driven systems can offer superior print quality, speed, and reliability compared to traditional stepper motor setups.

According to Charan Bhamra, a Mechatronics Engineer with Monolithic Power Systems, the company behind EZmotion, "This demonstration effectively validates the real-time capabilities and precision offered by our integrated servomotors. The ability to manage visual-feedback latency and achieve tight multi-axis synchronization on embedded hardware opens new avenues for innovation in medical robotics, high-speed automation, and advanced manufacturing. We believe these integrated solutions will dramatically simplify system design for engineers while simultaneously elevating performance standards across the board."

The continuous drive towards greater automation, miniaturization, and precision in critical applications underscores the importance of such integrated motion control technologies. As industries demand ever-higher levels of accuracy, speed, and reliability, solutions like the EZmotion integrated servomotors, coupled with sophisticated real-time control architectures, will undoubtedly play a pivotal role in shaping the next generation of intelligent machines and robotic systems. The demonstration platform serves not just as a showcase of current capabilities, but as a compelling preview of the future of precision engineering.