National Instruments’ LabVIEW has been used with robots for decades. It mostly has been employed by developers looking to take advantage of graphical programming tools (see “LabVIEW: Graphical Programming”).
LabVIEW also was the underlying platform for the popular Lego Mindstorms robots (see “The Mind Of Mindstorms”). Lego Mindstorms (Fig. 1) spawned a generation of kids playing with robots from an early age. It remains the basis for the FIRST Lego League robotics competition (see “Future Engineers Brace For Battle Of The Robots”).
The ARM-based Lego NXT control block was a significant step up for most new robotic developers with a 48-MHz Atmel ARM7TDMI processor. Children can start out using the graphical programming environment that comes with Lego Mindstorms, but there are lots of alternatives ranging from free platforms like RobotC and the Java-based LeJOS to commercial products like Gostai’s URBI and Next Byte Codes NBC. Even LabVIEW and IAR’s Embedded Workbench are available options.
But more advanced systems like those often used in the other FIRST competitions needed support for peripherals that NXT could not handle. Likewise, the applications tended to be larger as well, exceeding the storage capacity of the ARM-based NXT brick.
National Instruments’ Compact-RIO (Fig. 2) and its sibling, the Single Board RIO (sbRIO), were popular platforms (see “RIO Boards Target Control And Data-Acquisition Apps”). With a processor and FPGA at their heart, they’re designed for industrial applications, not just robotics. These platforms also accept a range of standard and custom plug-in modules, which allows the RIO platforms to handle data acquisition as well as process control.
Handy as an intelligent sensor in robotic applications, the Smart Camera can communicate with other devices via Gigabit Ethernet. However, it also has enough processing power to perform a wide range of analysis and compression algorithms, reducing the communication load.
Given this wide range of hardware and software support for robotics, it wasn’t surprising when National Instruments released its LabVIEW Robotics Starter Kit (Fig. 4). It’s based on the sbRIO 9631 platform, which includes a Freescale MPC5200 processor, 110 digital lines, up to 32 analog outputs, four analog inputs, and 32 industrial 24-V digital I/O. The kit also adds a frame, motors, wheels, and sensors in a compact robotic development platform.
The system runs off a 12-V nickel-metal-hydride (NiMH) battery. Gear-driven 4-in. wheels provide good traction and ground clearance. The robot design is rather wide to accommodate the sbRIO system mounted on top. It has Ethernet and serial port connectivity. The kit is not ambitious enough to have a Smart Camera included, but it does have a Parallax Ping))) ultrasonic sensor mounted on a servo.
BUNDLING ROBOT SOFTWARE
The hardware is impressive, but the software makes the starter kit stand out. LabVIEW is at its center, of course, but robots require so much more. The system includes LabVIEW RealTime, LabVIEW FPGA, NI Vision, LabVIEW Control Design and Simulation, NI SoftMotion, LabVIEW Statechart, LabVIEW Mathscript, and LabVIEW PID Toolkit. The Wind River VxWorks real-time operating system (RTOS) provides the RealTime support on the sbRIO.
National Instruments includes a variety of virtual instruments (VIs) targeted at robotic sensors and controls such as the Smart Camera and simpler systems such as the robot’s motor control. This is the start of a framework for a common robotic design and control system, although it isn’t as general or encompassing as Microsoft’s Robotics Developer Studio (see “MS Robotics Studio”). For example, the Robotics Developer Studio includes a simulator where applications can be tested in a virtual environment. Microsoft does provide a graphical programming language called VPL (Visual Programming Language), but it lacks the massive support of LabVIEW. It does meld well with Microsoft’s other tools that do have comparable support, though.
Still, the Robot Starter Kit is a major step forward for robotic development, especially since linking together tools such as planning and image recognition is relatively easy with an environment such as LabVIEW. The CompactRIO platform is already in the hands of FIRST teams, so designers looking for a compatible but less expensive solution will definitely like this latest kit.