By Douglas Hixon
When it comes to flexibility, nothing beats the human arm. Need to reach for a cup of coffee behind your computer monitor? No problem. You extend your arm
up and over, carefully grasp the cup, lift it over, and
enjoy. There’s a problem here, though. You need to
grasp it carefully. One wrong move, and you’ll spill
coffee over the table.
Why do you have to be careful? Because, from a
purely mechanical perspective, your arm has a big
job to perform. To reach over the monitor, your
arm articulates on several axes, back and forth and
rotationally, and it must lift and maneuver over the
monitor with just the right amount of force, at just
the right trajectory and speed so as not to spill a
drop. And when it comes to speed, it’s the accelera-
tion and deceleration that really count. Lift the cup
or jerk en route (accelerate) too quickly, and the
coffee splashes all over. Slam the coffee cup down
on the desk (decelerate too quickly), and you still
slosh coffee everywhere.
To avoid all this, let’s say, for whatever reason,
you want to develop a machine to lift the coffee
cup for you. The simplest and most accurate way
to do this would be to build a machine that runs
on vertical and horizontal tracks—linear axes. You
then have an end effector that’s designed to lift the
coffee cup the same way every time. In this situa-
tion, the device moves precisely and consistently.
Your arm cannot reach over the monitor precisely
the same way every time, but your new contrap-
tion, with its linear axes, certainly can.
But now say you buy a new desk, and you don’t
need to reach for coffee anymore. What are you go-
ing to do with this contraption you built? Not much.
This illustrates the traditional trade-offs manu-
facturers have had to deal with in machine automa-
tion. Say you needed a machine to move from one
point to another. If you cared only about accurate
positioning at the end points—like in a pick-and-
place material handling operation—you used an
articulating-arm robot. If you needed absolute pre-
cision on the path between those end points, you
would turn to a system with linear axes. Compared
to a robot arm, such systems can’t reach into every
nook and cranny within a work envelope, nor can
they be retooled easily for a different application.
But thanks to those linear axes, they’re precise.
But what if an arm could compensate for its positioning errors? In recent years robotics technologies have narrowed the trade-off gap. This gap has
narrowed in part because of advanced techniques
used to make the robot arms themselves. At the
same time, integrated software has helped make
the robot arm more intelligent.
Evolution of Robotic Precision
Consider robotic welding (see Figure 1). For decades articulating-arm robots have wielded resistance spot welding electrodes that pinch the metal
and perform a weld. The path between each spot
weld needs to be consistent, but it doesn’t need
to be absolutely accurate. If the robot’s position,
speed, acceleration, or deceleration varies a little
bit between welds, no one really cares. It has to be
accurate enough to avoid obstructions, but that’s
about it. It’s the position of the spot weld that really matters.
Now consider wire welding along a straight or
even contoured seam. In this case, we really care
about exactly where that robot path is between
the start and end points, and we also care about
the speed, acceleration, and deceleration along the
joint. But this is wire welding, so the joint does have
some gap tolerances. It can handle a little bit of variability. For years robots have been accurate enough
to handle the process, and they’ve been more cost-effective too. Not only have prices for robot arms
dropped precipitously over the years, they are flexible (more reach within the work envelope) and can
be retooled easily for another application.
What about laser welding? This process often requires extremely high precision and accuracy, depending on the joint geometry. That’s why sometimes people choose to use a prismatic system, one
with linear drives. But today more robots are being
used even for these kinds of precision applications.
Robotic welding, such as this remote laser welding application, requires the articulating arm to make accurate
movements. Not only do the start and end points matter,
but so does every point in between.
How today’s systems handle