Integration Tool Needed for Embedded Programming

Oct. 1, 2005
One exciting thing happening now with embedded systems is that a lot of programmable logic is being focused to get tighter and faster controls, says Shelley Greitlein, manager of National Instrument’s (NI, www.ni.com) LabView Realtime and Embedded Group, in Austin, Texas.

An example she mentions is the use of NI’s CompactFieldpoint out in the oil patch. The technology is a programmable automation controller (PAC)—a hybrid device that combines personal computer (PC) flexibility and power with programmable logic controller (PLC) ruggedness—that is coupled with wireless communications. The combined package allows that oil company to transmit real-time data from production wells to a central data repository. “There is an embedded controller and PC in the device. It has a Web server and can actually serve it (data) up on a Web browser or PDA (personal digital assistant),” she explains.

But the real challenge with embedded systems now is getting disparate types of systems, such as vision and motion control, to communicate, she says. “How do we deal with these complicated distributed intelligent systems? How do we help them communicate as a distributed system?” Gretlein says NI hears from its clients that the communications between the systems and the management of them are the most difficult challenges they have.

In mixed systems, such as vision and motion control, a combination common to machine builders, users deal with different input/output, thus different types of data. At one time, they had to bring in specialists if they needed to integrate such differing data, Gretlein says. But things are changing and need to change, she says. “The engineers and scientists—the control experts, the automation experts—want to be able to do that. They’re the local experts. Let them have the power, the ability to take advantage of the processors.”

There are different integration tools, Gretlein notes, but what is needed is a single-interface environment in which various embedded devices—such as PACs, vision systems and motion control systems—can share variables and allow a single communication interface. The new technology must be OPC-DA (an open communication standard) compliant, she adds, and work with standard Ethernet applications such as ModBus/TCP (transmission control protocol). It must not only allow users to view disparate systems from a single view, but it must also allow them to control those systems, she adds. NI, for one, has announced such a new technology.

Embedded growth

The importance of programming embedded systems grows because these systems are more prevalent than in the past, she adds. Due to this growth, Gretlein says more people will have to understand and program them, if they are to fully utilize embedded systems.

One technology they’ll use is PACs. “These PACs are becoming popular when a bottleneck needs to be fixed,” she says. A PAC’s utility is local optimization “at one place in the process where you need a little more control, a little more something,” she explains, noting that PACs are more intelligent devices that can also be programmed.

Yet another trend that Gretlein sees is growth in the use of Field-Programmable Gate Arrays (FPGAs), a type of logic chip that can be programmed. Gretlein believes FPGAs are taking off in machine control, when users need high speed and reliability and when they want to embed Internet protocol. This is particularly evident in machine control in the integration of vision and motion technologies, she says. “Users are getting more out of FPGAs because they are programmable and [enable] higher-speed applications because the application lies in the hardware.”

C. Kenna Amos, [email protected], is an Automation World Contributing Editor.