Bench Talk

Empowering Intelligence at the Edge

(Source: wpw/stock.adobe.com; generated with AI)

In our quest to empower more intelligence at the edge, physical edge devices such as electromechanical actuators call for smarter capabilities to benefit from better real-time decision-making at the machine. These actuators provide smart, valuable, and rich sensor-like feedback. They control robots as well as manipulate and automate factory processes—transforming digital information into physical motion while offering a high level of intelligence and self-awareness.^[1] While the actuators manipulate things, sensors are used to measure and quantify real-world parameters—they transfer from physical back to digital values. Thereby, actuators and sensors are mostly still considered to be separate devices or components.

Stepper motors and solenoids represent a large portion of those electromechanical actuators and are found on every factory floor, in automotive applications, in lab automation, and more. The global multibillion USD stepper motor and solenoid market continues to grow and is driven by lab and medical applications, industrial applications, and automotive applications. These applications demand an increasing focus on higher degrees of automation and miniaturization of the actuators and driver electronics. Traditional driver solutions are not tailored to these new requirements, and they lack sensing capabilities.

The latest silicon cDriver^™ solutions, which consist of a smart controller and driver, enable intelligent actuators at the edge by merging sensor and actuator functions into single integrated components to be used inside embedded motion control solutions.^[2] System parameters and state variables that are only available directly in or at the electromechanical actuator are measured and evaluated in place (for example, temperature, solenoid reaction time, and motor load value).

This fusion of sensor capabilities with the actuator yields a paradigm change for electromechanical actuators. They advance from simple power conversion systems to self-aware sensors that perfectly control the actuator and provide in situ data back to higher control levels, the cloud, or AI productivity solutions. The electromechanical unit becomes the sensor.

Why Do We Need a New Approach to Controlling Electromechanical Actuators?

Driver IC solutions available in the market today are not really tailored for solenoid drive applications and efficient and economic implementation. They lack embedded control sequencers, application-specific functions, diagnostic functions, and protection. Whenever advanced control features (driver sequencer, dithering, fast demagnetization, current measurement) or advanced diagnostic functions (detection of plunger movement,^[3] on/off status detection, inductance measurement, open load detection) are required, the system complexity increases significantly due to the additional external workarounds and circuits needed.^[4] The designer engineers every single block (digital controller, current sensing, signal conditioning, power stages, protections) and has those interconnected. Occupation of real estate/board space, long design time, application reliability, long bill of materials (BOM), and lack of flexibility are some of the problems that designers must still face.

Enhanced diagnostics is one of several global trends resulting in additional requirements and needs for embedded control and driver solutions for electromechanical actuators.

Electromechanical actuators are prone to degradation during long-term operation and may show other failure scenarios on the electrical side (coil problems, residual coil power, overheating, insulation failures) and on the mechanical side (partial valve closure or opening, manual override, pressure differences, dirt gathering, damaged valve mechanics, grease dry-out). These challenges affect the performance, lifetime, and operational availability of those actuators and the systems they are used in. This results in a critical demand for digitization: detailed and high-quality diagnostic feedback on local system parameters for monitoring the health state of the actuator and its control electronics, for better decisions at the local machine level to react to changes and to communicate diagnostic information preprocessed or as raw data from the edge to the higher control levels. This requires feedback and diagnostics beyond simple driver error flags.

Enhanced Diagnostics: Paving the Way to Predictive Maintenance and Self Awareness

Sensor-like data is available locally in smart cDrivers. But what can be done with this abundant set of information?

The parameters available with the smart cDriver solutions include driver temperature, information on coil resistance and temperature, coil inductance estimation, supply voltage, actual coil currents, and back electromotive force (BEMF) information. Smart integrated algorithms and functions allow for deriving system and application conditions and other system parameters like the solenoid’s reaction and travel time, local current dip, open load detection, over current and short detection, detection of part closure and plunger movement, plunger displacement measurement, and real-time current monitoring. For stepper motors, the actual load information based on StallGuard^™ and the level of the CoolStep^™ current reduction can be read out as well.^[5] For many applications, the StallGuard load value is extremely valuable information as long-term drifts may indicate degradation in the mechanics and gears or defective mechanical end stops within the application. The StallGuard value is directly related to load conditions on the motor shaft and can vary in the application and over time depending on motor acceleration or external forces. The StallGuard value even allows the detection of a physical motor stall before it happens. This is then useable for sensorless end stop detection or calibration in the application.

The local sensing capabilities and diagnostics and the availability of this in situ feedback pave the way for predictive maintenance and self-awareness on three different levels that are shown in Figure 1:

• Locally inside a cDriver component
• On the application level in the local microcontroller unit (MCU) of the embedded subsystem
• On higher levels like factory floor, plant control, or cloud

Figure 1: Diagram illustrating sensor and diagnostic data flow within smart cDriver™ electromechanical actuator systems, showing data paths from actuator to MCU to factory or cloud layers. (Source: Analog Devices)

Better real-time decisions can be made directly inside the controller and driver electronics due to the local monitoring and self-diagnosis functions. These consist of configurable thermal protection limits, configurable short detection, and driver protection in case of faults, automatic hit to hold current switchover, and immediate fault reporting; for example, when the solenoid’s plunger gets stuck.

Using the local MCU, more elaborate functions can be implemented by interpreting the sensor-like data within the application’s context. Real-time monitoring is possible via the serial interface of the cDriver. Diagnostic information and parameters are available as a stream of continuous feedback from the actuator and cDriver. This allows for more specific status monitoring, long-term failure identification, or even pattern detection. Over time, a drift of parameters such as reaction and travel time measurement, local dip search, plunger displacement, and load values is a symptom of actuator aging and indicates the need for preventive maintenance during its operating life. Sensor data can be aggregated. Application statistics beyond simple fault detection can be preprocessed and put into the right format before being communicated via the communication bus interface, such as IO-Link^®, CANopen, or even Industrial Ethernet derivatives to the higher control layers.

Conclusion

With a fusion of sensor and actuator, new cDriver components are enabling smart electromechanical actuators at the edge. cDriver components can do more than just switch a solenoid or spin a motor—they offer extensive diagnostic functions and thereby are a kind of sensor on their own. Preprocessed data make decisions locally and offer safety and monitoring functions. Such smart sensing actuators deliver additional value to the cyber-physical systems and factory floors of the future by solving mechanical challenges, hiding complexity, encapsulating sophisticated functions, providing abundant information to the higher control layers for further processing, and reducing cost and power consumption. This is a new level of digitization and a paradigm change in controlling electromechanical actuators at the edge.

Stephan Kubisch authored the blog “Electromechanical Actuators Require Smart and Integrated Driver Solutions to Empower the Intelligence at the Edge,” which was first published on www.analog.com and is repurposed here with permission.

About the Author

Stephan Kubisch is a director of Solutions at ADI Trinamic^™, INA, CMR. Stephan manages research and development activities for motor and motion control solutions in the Connected Motion and Robotics Team inside the Industrial Automation Business Unit. At Trinamic Motion Control (formerly part of Maxim Integrated, now part of Analog Devices), he held various roles including the director of product definition, head of R&D, and senior IC designer. Stephan holds a Ph.D. in information technology.

Sources

[1]  https://www.edn.com/intelligence-at-the-factory-edge-boost-productivity-and-improve-costs/
[2] https://www.analog.com/en/resources/technical-articles/embedded-motion-control.html; “Beyond Sensors and Cameras: How Embedded Motion and Motor Control Advances IoT.” Trinamic, December 2018
[3]  https://www.freepatentsonline.com/y2020/0328019.html
[4]  https://www.ti.com/lit/wp/ssiy001/ssiy001.pdf?ts=1651452039355; https://www.ti.com/tool/TIDA-00289; https://www.analog.com/en/resources/reference-designs/circuits-from-the-lab/cn0415.html; https://www.analog.com/en/resources/analog-dialogue/articles/current-measurement-in-solenoids.html
[5] https://www.analog.com/en/lp/001/building-better-stepper-motor-system.html; https://www.youtube.com/watch?v=l6r63Q7Yr58