Connected Intelligence Closes the Motion Loop

Outside of manufacturing, entities such as medical treatment centers using robotics also rely on this formula to operate smoothly and speedily.

With distributed intelligence systems on the factory floor, end-users may face unfamiliar problems. But these can be overcome, says Systems Engineer Lee Stephens, with Danaher Motion’s (www.danahermotion.com) Applications Center, in Radford, Va. With analog voltage-controlled servo motors, signal loss typically stopped motion, he says. In today’s digital world, even with electromagnetic interference or cable breakage, a servo with embedded distributed intelligence could continue operating. What’s the secret? Have an error-handling program, called a watchdog, within the system to appropriately terminate the servo’s motion, Stephens states. Redundant communications protocols may be needed.

The good news is that control-loop algorithms exist in such protocols, Stephen adds. Two such protocols are Profibus and SERCOS (Serial Real-time Communications System). These deterministic protocols permit loop closure in time and position, he says. But while Ethernet or FireWire—the Institute of Electrical & Electronics Engineers-1394 High Performance Serial Bus—could close loop position, these non-deterministic protocols wouldn’t “close time,” Stephens says. That means it’s unclear when loop closure will occur.

But quicker and better real-time machine behavior comes through distributed intelligence, asserts Rick Rey, product manager with Bosch Rexroth Corp.’s (www.boschrexroth.com) Electric Drives and Control Division, in Hoffman Estates, Ill. And to work with multiple-vendor solutions, machine builders can now create motion profiles from programmable logic controllers (PLCs) and then send those profiles, via common bus connectivity, to the distributed-intelligence drive, Rey notes.

PLCopen (www.plcopen.org) promotes this trend, he says. The Netherlands-based group created standard motion programming function blocks that conform to the International Electrotechnical Commission’s (www.iec.ch) Standard 61131-3, “Programmable controllers—Part 3: Programming languages.” The standard defines four interoperable programming languages: Instruction List, Ladder Diagram, Function Block Diagram and Structured Text, plus the organizing principle of Sequential Function Charts.

Beyond traditional manufacturing applications, robotic-based medical procedures demonstrate how the closed-loop-feedback relationship combines with distributed intelligence to enhance procedure precision.

The core of a radio-surgery technique using robotics fuses millisecond feedback and response. Sunnyvale, Calif.-based Accuray (www.accuray.com) created CyberKnife, a non-invasive, automated, surgical X-ray technology. It uses Weingarten, Germany-based Kuka Controls GmbH’s (www.kuka-controls.com) VxWin software to position the X-ray gun to the visual imaging system that tracks a tumor or lesion’s exact location, says John Patchin, an applications specialist in Kuka Robotics Corp.’s (www.kukarobotics.com) headquarters, in Clinton Township, Mich. The software allows, on the same processor, Redmond, Wash.-based Microsoft’s (www.microsoft.com) Windows XP to be added to,

for example, Windows CE or Alameda, Calif.-based Wind River VxWorks (www.windriver.com) real-time software, he says.

The control loop, embedded in the robot, handles feedback, which in this application means any movement by the patient. Infrared (IR) cameras monitor the positions of IR-light-emitting diodes that are known distances from a lead marker placed on the patient’s body. Every 12 milliseconds, the cameras report the positions to the robot’s controller, Patchin says. The robot then dynamically repositions the X-ray gun. It’s seamless, instantaneous-feedback motion control, enhanced by distributed intelligence. It’s life altering.

C. Kenna Amos,ckamosjr@earthlink.net, is an Automation World Contributing Editor.