What are the trade-offs between open- and closed-architecture controllers? Advantages of open-architecture controls (OACs) include controlling parameters other than position and speed. For example, process parameters, such as power and force control, are key in grinding operations—where the grinding wheel may become duller (or sharper) during operations. If force or power is held constant, other process parameters can be adjusted to optimize production rates.

This flexibility becomes critical for manufacturers producing large lots of components, such as bearings. For example, one major bearing manufacturer indicated that it made over 14 billion needles last year. Reducing cycle times by even a millisecond or two over 14 billion needles can yield significant cost savings. Consequently, OACs are used on advanced grinding systems.

Another very hot area for complex manufacturing systems is the use of diagnostics and prognostics for machine health monitoring, often called condition-based monitoring (CBM). OACs permit easy integration of a variety of sensors and algorithms that are on the cutting edge of CBM.

A major control manufacturer's CEO was asked recently, Why not go with OACs? He indicated that most of his company's customers did not want the added flexibility, and if it isn't requested, it's not worth the effort with the product. Furthermore, arguments are made relating to safety. OACs provide engineers with greater control access. However, inadvertent operations could create damage or, worse yet, unsafe situations.

So who is best suited to apply OACs? From what I have seen, systems integrators and equipment manufacturers have had the greatest successes with OACs. For example, machine tool builders are using open architecture controls on more advanced systems to easily integrate new features, such as CBM, or allow for control of non-traditional parameters, such as force and power. Furthermore, OACs permit quick and easy updates—sometimes even over the Internet.

While such advanced applications can be implemented by machine tool manufacturers, other crucial elements—such as E-stops and safety interlocks—also can be properly integrated into the machine. This is a critical function for all machine-tool controllers. However, access to such safety routines almost never is provided to the end-user. Thus, the end-user typically lacks the flexibility afforded by OAC. In fact, many users prefer controllers to perform a variety of advanced functions, while appearing as a standard closed-architecture controller to the operator. This demonstrates yet another advantage of OAC; it can be programmed with a wide variety of user interfaces, including mimicking just about any closed-architecture system.

Bottom line: controllers based on faster real-time PCs, digital signal processors, and field-programmable gate arrays (FPGA) are becoming less expensive, more accessible, and easier to program. Inevitably, controllers will migrate to these platforms. Even today, closed-architecture controllers are often prototyped on FPGA-based hardware.

Controller 'openness' will depend greatly on individual companies' willingness to permit entrée to their systems. In most situations, such access will not be required, and may be considered a liability. Thus, for the foreseeable future, I believe that most controllers will possess a closed or mostly closed-architecture. However, for the more demanding, and higher-end, applications, where the customer is extremely knowledgeable about controls and their process needs, OACs have made significant inroads and will continue to advance the frontiers of production capabilities.

Watch for a future 'Academic Viewpoint' on controls designed for reconfigurable machines. See the Control Engineering Online Extra, especially the link to the report: International Assessment of Research and Development in Micromanufacturing.