原创 Electric motors plus pneumatics equals awesome robotics

Robotics is hot these days, and it's certainly deserved. Today's robotic arms and systems are faster, cheaper, more precise, and lighter by large factors compared to those of even just a decade ago. The result is the deployment of these units in stationary and mobile applications that were impractical but are now very feasible.

If you think this is just the usual post-CES type of public-relations hype, think again. Check out this article in a recent issue of Design News, "Mechatronics Solution Is Adept Enough to Produce a Bow-Tying Robot," which details a robotic system which ties decorative ribbon bows on boxes of chocolate. (The YouTube video here is both mesmerizing and hypnotizing, IMO.) The article is more than a puff piece on how smart the design team was; it explains some of the problems they encountered, the constraints they faced, the tradeoffs they juggled, how they used simulation tools to test out ideas in advance, and more. Also note that this is not a "science project" robot demonstration–though there is nothing wrong with that–this one is in daily commercial use on a production line.
 

If you have ever worked with earlier robotic arms powered by hydraulics, you know how powerful and terrifying they can be, and at the same time, they can deliver lots of torque and power (different but closely related quantities). But they are nasty, with pumps, values, seals, leakage, and more. When things went bad, the result was often both a mess and dangerous, such as hydraulic fluid leaking, or worse, spraying out. Some fluid-powered systems eliminate the hazards of hydraulic fluid by using compressed air (pneumatics) in its place, but the air pressure has to be fairly high to get useful power, and high-pressure systems bring their own risks. Still, it's a very good alternative for low-load situations, such as the bow-tying one.

Many of today's robotic systems have gone all-electronic, with fast, low-mass, precisely-controlled motors (BLDC or stepper). Even if hydraulics did not have the fluid pressure and leakage issues, they still have a major drawback compared to an electric-motor system: due to the inherent mass and compressibility of the fluid, there is an unavoidable time lag between changes directed to the servovalves and the end-effecters. These lags complicate the closed-loop control algorithms when you need fast, precise operation. While it is possible to add various compensation factors to the control algorithms to take this mass-based reality into account, this adds complexity and risk to the software along with additional errors.

In contrast, with electric motors there is only a negligible lag of electronic switches and current flow that occurs between directing power to the motor and having it initiate motor action. This means the algorithms are dealing with a much crisper system, in control terms. Further, the lighter weight of the today's motors with their powerful magnetics compares favorably to hydraulic actuators, and this high power/weight ratio simplifies the remainder of the control design. By combining the best features of electric drives and pneumatic ones, a good system engineer can develop a system which meets previously unattainable objectives. Think about that fast-food production line; that's a likely application area.

Coupled with improved algorithms for both sensored and sensorless control (such as vector field control, also called field-oriented control or FOC), the electric motor has become a powerful element – literally and figuratively – in the advances of high performance, easy to use, reliable robotics. Hydraulics still has many places where it is the best choice in terms of massive power delivery, torque, non-sparking, or other factors, of course. If you doubt that watch this 50-minute documentary "Sunken Ship Rescue" on the righting and salvage efforts of the cruise liner Costa Concordia. Still, for many small-to-moderate applications, electric is the way to go.

What has been your experience with electric, pneumatic, and/or hydraulic robotic drives?