Applying COTS principles to motion control
When building a motion drive or controller, manufacturers traditionally implement advanced control techniques and package them in a black box with a few gain settings, a slider, or an autotuning feature, and the machine builder or end user may never know or care how the control is implemented. Machine builders send high level commands to the system and rely on the expertise of the manufacturers to provide the best performance possible. They shouldn’t have to care what goes on inside that box, as long they get the performance they need. But what happens if they don’t? What if their application doesn’t match the use case originally defined by the manufacturer? Or what if the application is more complex than a “standard “application and requires better performance than the out-of-the-box firmware provides? In today’s world, where technological innovation is driving an increasing need for innovation in manufacturing, it is unlikely that all of the requirements for a new machine or process will be met with off-the-shelf controllers and drives.
So what happens then? Engineers could try a different supplier that has implemented a different black box that could meet their needs, but chances are these customers already selected an industry leader for controller and drive technology, and it’s doubtful that a competitor could significantly outperform the existing solution. Users could try going back to the original manufacturer to ask for new features, but it’s uncertain whether their needs will be prioritized high enough to be implemented, or and at what NRE cost. Because of these challenges, today’s machine builders are often turning to full custom design solutions. Though custom design can meet their individual needs, the downsides of choosing this option are significant increases in cost, development time or even team size. Additionally, the engineers will have to maintain and take ownership of any changes to use this custom design for another project in the future.
An ideal solution in this common situation is the ability to pair the convenience, cost structure, and usability of off-the-shelf black box components with the flexibility and performance of a custom design. This type of system is sometimes termed a COTS (customizable off-the-shelf) system, and this design philosophy is starting to be applied to motion control applications through the use of “SoftMotion” tools.
SoftMotion is the concept of modularizing motion control tasks and making them accessible to the user for modification. A core principle of designing an effective softmotion system is the expectation that the end user will need to modify components to meet their specific needs. Of course, many users may be fine with the tools off-the-shelf, and the goal of any good manufacturer is trying to make off-the-shelf functionality applicable in the overwhelming majority of cases. But it is unrealistic for a vendor to meet the demands of every customer use case, especially with the current pace of innovation and the variety of applications in machine building. SoftMotion is all about providing an extensible platform that will allow end users to take the off-the-shelf architecture and extend it to meet their needs. In other words, end users avoid starting from scratch when it comes to their motion applications.
These extensions to the platform also present an opportunity for system integrators and machine builders to create significant value for their customers without reinventing the wheel. These engineers can add their own knowledge and expertise to a system in a specialized field while leveraging the surrounding framework to optimize effectiveness. For instance, an engineer could add a specialized flying shear algorithm on top of a standard gearing implementation, or even rip out an existing observer model in a control loop and implement a specialized version that can account for disturbances characteristic of a particular application.
This approach to motion control can revolutionize the design process for machine builders. A softmotion system is generally realized by integrating motion functionality into a general purpose controller where it can be modified and redeployed by the end user using the same standard programming environment they would use for the rest of their control system. This brings up another valuable aspect of a softmotion approach. Because the motion functionality is now implemented on the general purpose controller and with the same programming environment that would be used to program the rest of the control system, integration of motion with other subsystems and synchronization with I/O should be seamless. Think about the amount of time spent today on integration of subsystems, of I/O, with vision, with the HMI, with the SCADA system, for diagnostics and prognostics. This methodology can provide huge gains in productivity for the engineer building advanced machinery. Especially if their application does not match the designed use case the component manufacturers used in defining their feature set.
As an example, NI provides a SoftMotion module for LabVIEW that runs on CompactRIO or PXI controllers; devices that combine a real-time processor, user programmable FPGA, and modular I/O. The module provides motion functionality in the same environment used for overall application control, machine vision, and a variety of measurements. Motion is inherently well integrated into the overall machine design. Applicatinos for this module include high performance machinery for semiconductor processing and electron beam welding.
SoftMotion does not just mean open access to a flexible set of motion functions in a common programming language to create value with user extensions. It also means seamless integration with other machine subsystems. This approach provides a tool set that can be used off-the-shelf for standard well defined applications, but also provides the capabilities needed by engineers working on the frontiers of their industries -capabilities that were previously only possible through full custom design. Also, if an application fails somewhere in the middle, , an engineer can crack open the specific part of the toolset that needs to be modified, make changes accordingly, and seal it back up to be used within the context of the larger application framework. The combination of a modular open software approach with customizable off-the-shelf hardware has the potential to revolutionize the design process for machine builders.