The Power of Embedded Drive Solutions
POSTED 07/15/2010 | By: Kristin Lewotsky, Contributing Editor
Embedded drive solutions simplify integration, enhanced performance, and speed time-to-market.
Standard, discrete drives are off-the-shelf solutions designed to address a broad range of applications. What happens, though, when the application has special requirements for size, performance, or function? There was a time OEM machine builders had only two options: accept performance that didn't meet their needs or design a custom solution in-house. No longer. Today, companies throughout the industry provide embedded drive solutions with varying degrees of customization. The approach simplifies integration and potentially speeds time-to-market while allowing OEMs to achieve the performance and packaging required to address their customer base.
At its most fundamental, an embedded drive solution consists of an intelligent co-processor or an integrated drive/motion controller. It's a package designed to include the minimum hardware required to deliver the software capabilities to the end-user to simplify their development of their motion control subsystem. Embedded drives can be implemented as plug-in modules designed to snap into the motherboard of the system, or as open-frame versions that can be attached to a circuit board mounted to some other part of the machine. That allows them to be used with both centralized and discrete architectures.
An embedded drive typically performs a key function that the customer cannot obtain from standard, discrete products. Perhaps they need a specific communications protocol or communications bus or have specialized packaging requirements caused by environmental factors such as heat, cold, vibration, space constraints. Perhaps it's a matter of performance, such as a system that needs velocity regulation to 0.0025%, when standard components can only reach 0.25%.
The beauty of the approach is that through software, the basic motion functionalities can be augmented by a range of capabilities including filtering, I/O features, communications, and more. “You can bring in some control electronics in the way of particular relays or different forms of isolation, or transforming low-level signals to high-level signals or vice versa," says Karl Meier, Director of Marketing at Advanced Motion Controls. “You can, in this same form factor, bring in functionality that you would have had to implement in other cards or in other PLC or control systems and you can do it at a much lower cost. It gives the OEM the ability to think outside the box—the boundaries of the box are not even defined anymore.”
When To Choose An Embedded Solution
Embedded drives can provide enormous benefits but only when applied in the right circumstances. Industrial automation markets such as packaging get their best performance from off-the-shelf discrete drives that are manually connected to traditional panel mounts. More specialized applications like lab automation or camera orientation may not only present more unusual requirements, their production numbers can to run to hundreds or even thousands of units per year. An OEM machine builder producing eight or ten cap, fill, seal machines per year can afford to use standard, discrete solutions wired up by electricians. OEMs building product in the triple digits need a faster method. “If you're building 1000 units a year or something, you typically don't want to be hand connecting wires," says John Chandler, Vice President of Sales, North America at Technosoft Motion Technology (Canton, Michigan). "You just want to plug modules into a circuit board where all the cabling can be connectorized to the end-user requirements. It reduces a lot of cost in volume manufacturing by eliminating the error prone processes associated with manual wiring.”
A key benefit to the embedded drive approach is that it allows the OEM design team to concentrate on their core value proposition rather than on a support technology like motion control. Consider an ultrasound imager. The primary value that the manufacturer brings to the product lies in their ability to generate, capture, and process the signal into an image. Motion control consists of typically two axes of motion in the hand-held scanner, and while the task is critical it is of less concern to the development team. An embedded drive solution takes that part of the task off their hands in a plug in module that can be snapped into their motherboard.
The capabilities can also dramatically speed time-to-market. He points to lab automation specialist Blockwise Engineering LLC, which needed a solution that would give them a fast time to market for their Rotating Beam Fatigue Tester Model FTX, a device that fatigue-tests stents and small wires (see figure 1). Blockwise’s key value proposition is a patented mechanical chuck system that holds these small components. In the motion system for the FTX, two motors, one at each end of the test piece, needed to continuously rotate the piece and bend it once per revolution, all without applying torsion. It was a nontrivial problem. “You have to synchronize two drives, and have them remain phase locked to each other while rotating counter to each other so that they're continuously bending the stent but they're not twisting it,” says Chandler.
The FTX contains a motherboard and pack housing that handles the chucking system, machine level operations, and the user interface. For the motion, the Blockwise design team chose an embedded drive approach, integrating the plug-in modules and then connecting motors directly onto the motherboard. It allowed them to get their prototype operating within weeks. "They used off-the-shelf drives and motors mounted less than half an inch apart from each other, linking them together with CAN communications," says Chandler. "That allowed them to synchronize [the motors] positionally so they can rotate from 0 to 40,000 RPM, run for days on end and then stop and they are forever phase locked together. The whole motion control challenge would have been a time-consuming technical hurdle for Blockwise, considering they were trying to get the product to market quickly. The embedded drive approach streamlined their product development cycle.”
Of course, for companies to take full advantage of the potential of embedded solutions, they need an engineering staff with the skill set to at the very least be guided through the process by the manufacturer. Rather than electricians and controls engineers, their engineering staff needs to feature electrical engineers and software engineers familiar with programming microprocessors. In fact, the staffs might very well be capable of developing their own embedded drives, but that process is not nearly as simple as it might appear on first blush, especially when time is of the essence. "Designing one of those little plug-in modules, while they look simple, might keep a staff of engineers busy for two years," says Chandler. "The hardware is not trivial. These things are surface mounted top and bottom, they're multi-layer circuit boards.” Of course, the hardware is just the tip of the iceberg. Because so much of the functionality is implemented in software, level of support required there is intense. "Years ago, they might have two or three hardware engineers and maybe one software engineer. Now that most of the functionality is pulled into the embedded software side of it, one hardware engineer can keep half a dozen software engineers busy." Choosing one of the standard or customized embedded drive solutions available from a range of manufacturers allows OEMs to spend their engineering time enhancing their products core value proposition rather than focusing on motion.
It seems counterintuitive that a highly integrated, value-added component could be cost effective, but the reality is that embedded drive solutions and embedded motion solutions can offer significant economies of scale. Standard, discrete drives are designed to address a wide range of customers and applications. They feature connectors, standard line filters, packaging, and so on. They've typically been through various approval cycles for UL, CE, etc. These are all important features for the target customer base but they also raise the price.
“If you make a standard product, you have to put the features into it to meet a wide range of the market or you are not going to sell very many," says Ed Wilhelm, Engineering Manager at Electrocraft (Ann Arbor, Michigan). “You have to have the extra inputs and outputs, you have to have the safety agency ratings for the different markets you expect to sell it into.” For some markets and applications, that can be a big benefit. For embedded drive customers, it may not be necessary. “In some cases with an embedded drive, a customer can just take their machine through safety regulation testing and certification and not worry about the individual components inside, so they're not paying twice for that.”
A certain amount of non-recurring engineering (NRE) hours go into the implementation of an embedded drive solution, particularly if it requires customization. Given the trade-offs that discrete solutions require, however, the price can ultimately settle in favor of the embedded drive. Particularly when amortized over large volumes, the cost of the non-recurring-engineering (NRE) becomes quite reasonable, especially when combined with a potentially smaller component count and faster, more efficient assembly than the discrete approach. "Say your piece price for a standard drive is X," says Wilhelm. "At the same quantity your piece price for a custom drive could be X minus 20%. Over the life of a product line, that 20% savings would more than pay for the engineering to develop this custom embedded drive." Also worth noting is that the volumes don't have to be high to create this value proposition. "Many times you can be down in the double digits to triple digits to really make it economically feasible.”
Of course, no solution is without some trade-offs. Instead of simply being able to swap out the failed drive, repairing a fault can require pulling the motherboard and shipping it back to the manufacturer. The drives also require careful attention to thermal management. Instead of being in separate housings or even mounted in a cooled cabinet, they’re attached to a motherboard that is already probably generating significant amounts of heat to begin with. In the case of open-frame designs that allow the board to be mounted anywhere on the machine, the machine itself can serve as a heat sink. For cases in which the chips are mounted on the motherboard of the machine, standard electrical techniques such as passive cooling the heat sink for active cooling the fans may be required.
Cautions aside, for the right application and the right user, embedded drive solutions offer significant benefits. It's an approach that has grown up organically around the market, as part of a general trend toward intelligence in the system. A few years ago, there was a lot of excitement about the idea of integrating drives with motors, for example. Embedded drives took the whole concept one step further. “It doesn't matter that the drive is with the motor anymore because we've gone beyond that,” says Meier. "Now you take that embedded drive and motor and make it indiscernible from the rest of the machine. We're making a whole control solution that's part of the electronics of the machine, so the costs go way down compared to buying even the standard product.”
"Many OEMs are not aware of where the technology has evolved to,” says Chandler. “[The industry is] adding more and more capability, performance and software, and we keep driving the cost down and minimizing the components that it takes to pull together a solution and minimizing the packaging. If I was a CEO, I would really encourage my engineers to go look around before going to design our next generation, because the embedded drive products just get better and better.”