“The next evolution in industrial automation requires machines to be able to independently adjust their performance parameters to accomplish tasks assigned by factory operators, or to reconfigure themselves to optimize their behavior based on input from productivity-enhancing artificial intelligence (AI) algorithms. The value of a self-aware machine is that it maximizes productivity, extends equipment operating life and reduces maintenance costs.
By Jeff DeAngelis, Vice President, Industrial Communications and Motion Control, ADI’s Industrial and Healthcare Group
The next evolution in industrial automation requires machines to be able to independently adjust their performance parameters to accomplish tasks assigned by factory operators, or to reconfigure themselves to optimize their behavior based on input from productivity-enhancing artificial intelligence (AI) algorithms. The value of a self-aware machine is that it maximizes productivity, extends equipment operating life and reduces maintenance costs.
A Journey to Self-Awareness in Motor Control
Self-awareness refers to the system’s self-awareness of itself based on its knowledge of its capabilities and system performance goals. In fact, self-aware motion control systems need to implement multiple control loops to interpret sensor inputs and expected system parameters and allow comparison of the system’s own operating behavior with expected system performance. To achieve these goals and create self-aware motion control systems, adaptive motion control agents need to be created to monitor system operation and dynamically adjust its performance based on the operating environment of the drive system. In ADI’s view, self-aware motion control systems can be built by using automated agents to detect and monitor constantly changing work environment conditions. These conditions are derived from a series of nested closed-loop real-time performance models that use the motion parameters of field-level drives. After deriving an electrical and mechanical model of the drive system, this model can be used to compare and adjust the expected system performance requested at the supervisory, planning, or management level of the automation system pyramid (see Figure 1). When a new expected system performance is requested from any level above the supervisory part of the automation system pyramid, a new set of control parameters needs to be transferred to the adaptive control part of the motion control system. The system then responds by adjusting its performance to match the new performance request.
Figure 1. Automated Systems Pyramid
The two major advantages of implementing a self-aware motion control system are the ability to self-regulate and automatically maximize the performance of the motion control system in real time. This new capability provides an opportunity for the supervisory, planning and management levels of the automation system pyramid, allowing self-aware motion control systems to be tuned by implementing productivity enhancements. In addition, AI-enabled software algorithms can be used to tune system performance to achieve better results at the factory level. To be more intuitive, ADI has designed a self-aware motion control concept map to better understand the 4 essential elements required to implement a self-aware motion control system.
Conceptual illustration of motion control with self-awareness: To achieve this level of self-aware motor control, a control system diagram needs to be developed. Figure 2 represents the 4 key elements required to successfully implement self-aware motor control.
Element I: Goal or Mission: Need to establish clear goals or tasks for the system. In the example, this means “Move the beer mug from point A to point B in an optimal way without spilling any beer”.
Element II: Expected System Behavior: After clarifying the goal, the next level of the self-aware motor control map initiates the expected motor behavior. In the beer mug example, this would be “using linear motion to move the beer mug while automatically adjusting its motion to compensate for different beer mug weights and dimensions within the required mechanical system control safety limits”.
Once the target and expected system behavior has been determined, the adaptive control engine provides a better understanding of the core drive system kinematics and its attendants by automatically adjusting the motion control drive and its integrated mechanical systems to achieve peak operating performance when operating in a unique work environment. Dynamic drive fusion between mechanical systems.
Figure 2. Conceptual diagram of motor control with self-awareness
Element III: Core Drive System: At the heart of a self-aware motor control system is its kinematics. The challenge is to observe, learn and monitor the performance levels of the motor and drive system. To create an effective drive system model, an intelligent observer needs to be implemented in order to gain a basic understanding of its motion parameters and their physical limitations. This can be achieved by using a field-oriented controller (FOC) with a dedicated position sensor or sensorless FOC approach to understand how the motor is being forced and reacted in the operating environment. The drive system response can be further optimized by monitoring and automatically adjusting the control parameter values from the motor torque-flux current loop, the velocity loop, and its positioning loop. After datagrams of this information are collected and fed into the smart observer, optimization algorithms are implemented to ensure that the motion control parameters are calculated and that the underlying motion control algorithms converge to form an optimal set of motion parameters (see Figure 3). Now that an indirect motion model is created to model and optimize the motion of the drive system, the next level of self-aware motion control solutions can be implemented by introducing an adaptive control engine. Currently, the TMCL-IDE from Trinamic (now part of Analog Devices) automatically adjusts the motion control tool to optimize motion control values.
Figure 3. Monitor and automatically adjust the torque-flux current, speed, and position loop
Element IV: Adaptive Control: Building on the system’s kinematics and FOC auto-tuning capabilities, it is now possible to focus on implementing the next level of self-aware motion control – the adaptive control engine. This level of intelligent motion focuses on communicating expected system behavior to the adaptive control engine (Figure 4). This system behavior is provided by production employees, factory supervisors, or generated from artificial intelligence productivity algorithms that collect factory data in a network of smart sensors. After the expected behavior is communicated to the adaptive control engine, the self-aware motion control system begins to dynamically reconfigure the drive system operating parameters to match the expected system behavior. Some examples of these expected behaviors include requesting increased plant throughput, or extending the operating life of a motor by operating in a safe mode. As the motion control system automatically adjusts its motion control parameters to achieve this new requested performance level, the adaptive control system continuously monitors the closed-loop system to maintain its desired performance level. This state is maintained even if the drive system encounters changes due to mechanical system wear, or even if the motor operating environment changes. The system has now reached the ultimate level of self-aware motor control.
Figure 4. Adaptive Control Model
Using a real example is perhaps the best way to demonstrate this concept (Figure 5). For example, a bartender wants to accurately deliver a full glass of beer from the side of the bar to the customer without spilling a drop of beer in the process. How to achieve this? If a self-aware motion control system is used, it becomes very simple. The goal of this task is to get beer from the bartender (point A) to the customer sitting at the bar (point B) in the fastest possible time without spilling beer. The conveyor system in this example is a cup holder with a built-in weight detector that detects the weight of beer mugs of various sizes and uses a linear motion to move the mugs across the bar. Imagine using a self-aware motion control system to deliver beer to the customer in the fastest possible time, if the customer returns an empty or half-empty beer mug to the cup holder, which is easily passed back to the sommelier for refill or discard. , and the system automatically adjusts its speed and performance. In addition, the system can also optimize efficiency if bartenders use different sized glasses to serve other types of drinks to customers.
Figure 5. Practical example of a motion control system with self-awareness (different load masses)
Although it may sound incredible, self-aware motion control technology is evolving rapidly, and it is believed that it will enter people’s lives and work in the near future. Imagine a smart factory far beyond people’s imagination when self-aware motors and smart sensors are used in equipment throughout the factory. At that time, potential failures of factory floor equipment can be repaired by themselves, the operating life of equipment can be effectively extended, production processes can be automatically adjusted, and productivity can be maximized. Welcome to an exciting new world to experience ADI’s self-aware motion control and the next industrial revolution for real.