Some of the latest developments in software ecosystems aim to make motion systems easier to design and commission. Artificial intelligence (AI) has made rapid inroads here. For motion systems specifically, the trend is to use AI for:
- Servotuning — before a system is put into service.
- Predictive maintenance — when a system is running in service.
First consider servotuning. Most vendors of servo controls offer motion-control autotuning software. Usually, the servodrive and controller move the system at multiple frequencies and then tuning parameters are autoset for optimal response.
Even now, autotuning can’t always produce the most optimized motion — especially in dynamic systems. In these situations, manual gains and filters tuning is typically required — and done by highly experienced integrators, consulting engineers, or in-house engineering with lots of institutional knowledge. Such manual tuning is common.
But in a few high-end (demanding and dynamic) applications, AI-equipped servotuning can now run prewritten algorithms that proceduralize, formalize, and thereby replicate the so-called “black magic” of tuning experts.
For example, one AI-driven software called EasyTune can autotune servosystems with approaches honed by machine learning. The software is targeted to semiconductor, photonics, aerospace, and medical-device makers and based on massive amounts of data gleaned over years to shape recommend adjustments. From our webinar sponsor Aerotech, this software lets engineers fine-tune their systems without needing in-depth knowledge of servomechanic tuning.
Panasonic Industry is another supplier. They sell a servodrive that runs an AI-equipped servotuner (called precAIse Tuning) to cut position settling times to better than those of possible with manually tuned systems. That lets end users take full advantage of a Minas A7 servo drive-motor-control system having 27-bit encoder resolution and response frequencies to 4 kHz.
Other suppliers that offer AI-equipped servotuning software include Fanuc and Mitsubishi Electric with their Compact AI software.
In fact, Compact AI also executes certain predictive-maintenance functions — the second motion-related use we mentioned. Here, the trend is towards real-world uses of AI to leverage all the data that’s now easy to collect from automated installations. For example, the programmed versus actual move profiles of machine axes can be noted during commissioning and then tracked over time for deviations or performance drifts as well as more dramatic changes such as payload. Some suppliers’ tools include AI that can also inform parameter servotuning to adjust for a machine reassigned to new tasks or showing wear over time.
AI aside, many of today’s software tools can also detect when servomotor temperature (proportional to current draw) rises over time — typically because of wear of the mechanical assembly. Preventative maintenance with AI can track these trends to best time relubrication or replacement.
Vibration detection with accelerometers (in some cases on encoders) can also be paired with controllers … those that run Fast Fourier Transforms can allow the visual display of frequencies normally present when a machine runs … and when those frequencies change.
When informed with sensor feedback, software can also monitor multi-axis synchronization, energy consumption, advanced tuning settings, and the reliability of system communication and controls.
To be clear, the most advanced versions of these functions are typically associated with dedicated motion controllers … and hardware does have a bearing on what software is employed. Of course, software associated with PLCs and industrial PCs handle motion … and so-called software ecosystems have really blurred the differentiation between motion and the rest of the automation space.
PLC and industrial PC systems offer tight I/O integration, automation synchronization, vision, RFID, and SCADA and MES applications for digital-transformation (DX) functions.
Trends in design-engineering software consolidation
Regarding software for electrohydraulic and servo-pneumatic systems, leading options offer realtime tuning and advanced condition-monitoring functions once systems are in operation. For example, some offerings from Danfoss allow programming and commissioning as well as valve configuration.
For electromechanical systems destined for discrete automation, it’s now pretty standard for motion to seamlessly meld with machine functions — confusingly called process control in a lot of contexts. For example, Aerotech’s Automation1 — billed as a software-based controller — can run on PC or drive hardware … it lets engineers program single axis moves or build complete machines. The suite’s Motion Development Kit is configuration, control-device setup, and programming … including that EasyTune (which we covered earlier), and live build checking.
For automation employing PLCs, Siemens and Rockwell dominate; these include TIA Portal and FactoryTalk along with the SIMOTION, Studio 5000, and other pieces for motion and drive functions.
Standardizing on a given ecosystem obviously unifies design-engineering workflows.
A major drawback to using such suites is vendor lock-in. It’s also true if an engineer picks B&R Automation Studio or Beckhoff TwinCAT or Bosch Rexroth ctrlX or Omron Sysmac or Schneider EcoStruxure because these also define entire machine architectures. Of course, machine builders can’t always pick the platform … once a company is locked into a vendor, it may be impossible to switch. Software compatibilities can also evolve over time — and that can also necessitate the need to manage legacy versions.
New trends in sizing, selecting, and commissioning tools
In 2025 we take for granted all the online sizing and selecting portals (and downloadable software) — especially when those tools are for configuring and procuring mechanical components. Tools abound from all kinds of linear-motion, gear, encoder, and coupling suppliers. But even electric motors are configurable now.
Examples abound, but one industry-typical example is the Leroy-Somer Configurator from Nidec to let machine builders select motors, gearmotors, electronic drives, and brakes (if applicable) to obtain CAD files and technical specifications.
With some motors, it’s also become possible to import component parameters into control software.
For example, ECM PCB Stator Tech software lets engineers specify motors with stators that are printed according to the engineer’s requirements. At the core of this technology is software that takes engineers from simulation to prototyping to fully functional motors in a go.
Generated by the design tool is the aforementioned hardware as well as tightly integrated firmware for realtime control.
A new company called Alva Industries now offers a similar software tool for procuring custom motors. Called TorqStudio, this online platform was first shown at the Smart Production Solutions or SPS 2025. The software combines motor simulation and design — from the exploration of concepts to detailed performance verification before machine builders have to commit to hardware.
For motor creation, the engineer just defines their required motor size. The platform then automatically generates an optimized motor variants with different tradeoffs in motor mass, motor constants (Km), torque capabilities, and efficiency.
It’s entirely analytical so results come in milliseconds instead of minutes or hours. Every generated motor is mathematically accurate, physically realistic, and manufacturable. [The company’s] Users can also analyze any motor created within the software or from a stock catalog of torque motors and frameless motors.
We’ve already touched on AI for servotuning … and expect to see more in coming years. Software with these tools and other commissioning tools often come from companies traditionally known for their drives. For example, Yaskawa’s free SigmaWin+ is the de-facto setup and tuning suite for Sigma SERVOPACK amplifiers for mechanical analysis, simulation, and autotuning. Kollmorgen, Elmo Motion Control, ACS Motion Control, Galil … these companies also lead with drives and the software follows.
Edge computing and cloud computing
Ever-more tools are built into today’s software and yet many need vendor add-ons. Base software packages are often free. However, motion add-ons, safety editors, and visualization and simulation modules usually require purchase … so there can cumulative licensing costs tiered into base software packages. Or the opposite can also be true (as with Siemens products, for example) … so purchase of the base package allows access to a library of modules. Either way, hardware comes with free customer experience or CX software just to get the user going out of the box.
Cloud computing for system setup and configuration aren’t especially common yet — unless there’s reuse of some configuration for projects being scaled perhaps. But what is common are operations that leverage offsite computational capabilities to aggregate, manage, filter, and analyze data. That usually offers more computational power than local platforms.
For edge computing in discrete automation, typical edge devices include actuators, sensors, and connectivity components. These can also include gateways (for cross-system communications) … and in fact, smart motors can also be considered edge devices if they include onboard drive and control electronics. These eliminate data bandwidth and latency issues.
Essentially, these filter and analyze data before it’s sent onward to central controls or the design’s cloud presence. Here, AI (in the form of machine learning) is already leveraging edge-generated data to monitor and improve individual component and overall machine performance and life.
Code generation in engineering software is nothing new … it generates models and executable PLC or motion-controller code so machine builders don’t have to write it by hand. Commonly supplied motion blocks (prewritten) are axis control, camming, gearing, and diagnostics. Oftentimes, this is Structured Text or Function Blocks that can run deterministically on the target hardware — and in IEC 61131-3 programming formats. IEC 62443-4-2 is increasingly important as well for cybersecurity of discrete automation.
A nice-to-have is an ecosystem function that gives engineer access to the entire stack including lower-level code — even that generated by autotuning, sampling, and interpolation modes. Where there’s a bit of code transparency, there’s the possibility of debugging control logic, drive parameters, fieldbus issues, and various machine-safety states.
Leave a Reply
You must be logged in to post a comment.