It's always the same. Packaging end users want higher product throughput out of smaller, lower cost machines, without sacrificing one iota of product quality. Next-gen machines must also be more flexible and scalable to meet changing market demands and simplify plant integration. From a controls perspective, the answer is quite clear. By integrating motion and logic in scalable hardware packages, designers have a fighting chance to meet these ever tougher demands.
The advent of centralized architecture for motion control and logic has provided several advantages. For one, the integration of motion control into rack-based PLCs helped reduce component counts in the control panel enclosure, making it possible to program motion and logic from a single point in a single program. This delivered an initial round of cost savings; ultimately, however, this was only true when a single processor was used with a medium axis count.
Centralized control has an inherent limitation because a fixed amount of microprocessor resources are available for all required functions: motion, logic, overhead, communications, and other tasks. In any operation, top priority is given to the motion task. Whenever an axis is added, a new burden is placed on the centralized processor.
Hitting the control ceiling
At a certain point, the processor hits its limit and starts reducing performance to accommodate the additional axis. This reduction might be in the form of a slower response to registration inputs, not being able to run complicated cams or programmable limit switches, or not being able to run the system to the machine's full potential. This in turn can set up the need to add more processors, so the machine can run at full capacity. Once this becomes necessary, there is practically no cost advantage — or operational advantage — when a design engineer is forced to install complex PLCs for simple, low-axes count applications.
The disadvantage to PLC-based motion controllers is the centralized control architecture. In a number of situations, it has proven to be the limiting factor in providing low-cost, high-performance solutions. On simple machines like fillers, augers, infeeds, wrappers, and cartoners, using a PLC for motion control can be overkill, not to mention prohibitively expensive.
In addition, centralized control can limit a designer's ability to optimize machine performance. Packaging machines are highly motion-centric, which makes the motion control critical to maximizing efficiency and throughput. For example, a vertical form fill and seal machine that can mechanically run at 200 pieces-per-minute (PPM) might only be able to do 145 PPM, due to limited controls performance. In some cases, using a centralized control architecture can double the price of the control system. As an alternative, design engineers would be wise to consider distributed intelligence.
Spreading out system smarts
Distributed versus centralized control is defined by the location of processing power for the motion control. With a centralized architecture, a fixed amount of PLC processing power is divided among all the axes. As axes are added, available processing power is reduced. Distributed intelligence (DI) solves the problem by moving the burden of controlling an individual axis out to the drive. Due to advances in microelectronics, intelligence can be distributed throughout a machine — to sensors, motors, drives, and other components.
In a DI system, each drive is capable of closing the feedback loop and can handle advanced functions such as cam tables, absolute feedback, electronic line shafting (ELS), diagnostics, and high-speed registration. It's even possible to add safety on board and predictive maintenance functionality at the drive level.
The processing power that can be built into the drive with low-cost processors and memory allows the drive to be quite intelligent. Most important, when you add a drive, you add more intelligence to the system.
Distributed intelligence not only reduces the processing load on the controller, it changes the controller's role in motion control to a supervisory status. Some of today's decentralized controls can handle up to 64 axes — with no degradation in system performance.
Distributed intelligence is a modular, responsive architecture. It supports the scalability that is an absolute requirement in many operating environments. Adding an axis is greatly simplified: just add a new servo axis. Additional expansion cards or controller functionality are unnecessary because the intelligence is in the drive itself.
Adding intelligence in a drive-by-drive distributed fashion frees design engineers to create machines that better serve end user demands for convenience and flexibility. Because processing power has ceased to be a limitation, more servo-controlled axes are practical, along with the advantages of faster setup, greater precision, and higher reliability.
Implementing a DI system requires several components engineered to work in a decentralized architecture, including intelligent drives and a DI-ready controller. Some may think an intelligent drive is one that can simply handle the position loop and receive inputs. However, this type of drive still places a heavy burden on the processor. For true distributed intelligence, a drive should be able to handle tasks such as closing the position loop, absolute positioning, high-speed registration, cam tables, ELS, and diagnostics.
As more remedial tasks are handled by the drive, the controller's workload is reduced. A good example is to specify safety and predictive maintenance tasks at the drive level. By making them drive-specific, problems can be quickly isolated, downtime reduced, and machine throughput optimized.
The motion controller is the next component in this architecture. A DI-ready controller must take full advantage of intelligent drives. Its key tasks will include running logic, overseeing drive communications, I/O peripherals, HMIs, and system networks. Involvement in the motion will exist primarily at a supervisory level.
Logic on sale
OEMs and end users can achieve dramatic hardware and software savings using distributed intelligence, while still assuring improved machine performance. First and foremost, PLC and motion control hardware can be eliminated and integrated into the servodrive. Eliminating the PLC also leads to reduced wiring requirements, installation costs, and a smaller electrical cabinet.
Industry analysts indicate that this can reduce hardware costs from $7,000 to $1,000 in a typical application. Plus, eliminating the interface card and wiring between the PLC and servodrive can provide an additional $2,000 savings. Low axes count machines such as infeeds, wrappers, cartoners, casepackers, and palletizers will be able to derive the greatest benefit.