Archive for April, 2008

One Size Never Fits All

Steve Meyer
April 19th, 2008.

Sometimes a lot is made of new technology. There is a tendency to talk about the next big thing in whatever field as “The Solution”. But there’s rarely just one solution that works for everyone.

Currently in the “car wars” (a favorite topic, since we are all effected by gasoline prices) many ideas have been advanced as “The Solution”. We heard a lot about bio-fuels reducing our dependency on oil by 30%, but now are primarily contributing to rising food prices. Hydrogen fuel cells will replace gasoline engines, but not anytime soon because we don’t have an infrastructure that can produce hydrogen as a fuel, nor an acceptable means to store it. And so it goes.  This progression of ideas, and attempts to market same, makes the point that there is rarely a single solution that suits everyone.

Historically in the industrial controls field, companies are started because of a new idea. An engineer comes up with a new solution to an old problem, and Viola!, a new business is created. And as someone who knows, no, its really not that easy. The stepping motor gets invented, and a whole new generation of technology gets rolled out into the marketplace. Too bad the original inventor didn’t file for a patent.

But what comes along with new technology is a bias that the new solution will fit every application. And new companies can be very technology centric in their culture. So it came as quite a surprise to me that Festo Corporation, known for decades as one of the leading pneumatics suppliers, entered into the electromechanical arena.

Festo’s new products include stepping motors and leadscrews and linear motors for linear actuation with micron accuracy and speed. All of the new actuators are delivered in the same format of aluminum extrusion housing so that they are compatible with the classic pneumatic products.

And the point is: give customers more choices. Pnuematics offer high speed and high power density, but limited accuracy. So Festo came to the conclusion that not every actuator requirement is best approached with pneumatics. Sometimes complex speed profiles or higher position accuracies make electromechanical solutions a better choice. In addition, the ability to program position and speed parameters make electromechanical systems more convenient, with significant savings in setup and teardown time. A real plus in production environments.

Sounds like real progress to me. Keep the choices coming, customers will take advantage of the options.

No Comments

More Mechatronic and Robotic Reflections

Steve Meyer
April 14th, 2008.

The blend of sensors and motion control become crucial in many applications. The dextrous robot hand of recent years cracks eggs like a chef with the aid of sophisticated pressure sensors at the tips of its “fingers”. Check out the Shadow Robot company for some amazing videos of their “air muscle” powered robot hand in action. Lots of interesting work has been done to mimic the human hand. A miracle of grace and efficiency that is hard to duplicate.

Which brings me to the mechatronics relationship. We spend a lot of time in the motion control part of mechatronics designing systems that are based on velocity and position. And this is the context where the “motion control” term is useful, as far as it goes.

In many of the applications we are concerned with, the positioning systems get all the attention. Probably because its more difficult. Positioning tends to be a more “real time” behavior than discrete control systems allow, and so historically, a lot of the hardware used for positioning systems has evolved on its own. (more on this subject later)

But sometimes we have to take a step back from the positioning problem and all its complexities and get perspective on the project goal. Such is the case with many crimping and press fit applications. Position alone simply won’t do. Position is necessary but not sufficient. As with the robot hand, a feeback sensor is needed to determine if proper pressure is being applied to the part.

And that isn’t always obvious, nor is the proper sensor arrangement a simple matter.

If you have to create a machine that crimps a can lid to a can body, there is a required amount of pressure. And the pressure is the most important attribute of the process. The pressure cannot be below a certain value, otherwise the crimp will fail. The pressure cannot be above a certain value, or the part could be damaged and scrapped.

You can implement a pressure strategy a couple of different ways, and the exact method makes a difference. The machine I had to deal with used a brushless dc servo and a lead screw to drive a crimping tool to close the parts to an air-bag igniter. An air bag igniter is a little like a miniature hand grenade, it contains a small explosive charge. So you want to handle it with care.

The engineer who designed the system was using the current values from the drive motor to calculate how much force was applied. In addition, an external load cell underneath the fixture gave a force value to the control system to verify each part. Data was stored in PLC memory and fed to a PC every certain number of units.

But the motion control system software did not have any programming to wait for the value of the load cell and motor current to be checked to insure that right force was applied to each part. So while everything appeared to be working correctly, parts were failing QA. It took quite a while to figure out.

So when you have a production project, make sure you know what the metrics of quality parts are. Make sure they are part of every aspect of the control system, both hardware and software.

No Comments

Robots are Mechatronics

Steve Meyer
April 06th, 2008.

Robots are a fun topic. They tease our imagination to wander into the realm of possibilities. Especially when we create film robots with personalities like R2D2, I Robot and the many strange variations of Artificial Intelligence.

And we begin to push the envelope again as to what mechatronics properly consists of. Consider the “loader” in Aliens when Sigourney Weaver starts fighting with the Queen alien. This hydraulic system, once called a “man amplifier” was designed to facilitate moving high loads easily with human dexterity and control.  A number of experimental “exoskeleton” systems are currently under test in the military to help soldiers carry higher loads of equipment in combat situations. Some of these systems are air powered with sophisticated sensors to provide feedback to the user.

Robot challenges have become popular all over the world for High School and College students to compete in “battle” or structured competitions. Students can design using any technical platform and engage experienced mentors to help them with the formidable learning curve of robot design. Sure, if you have an expert for each discipline it would be pretty easy, motors, amplifiers, motor feedback, gear reduction, mechanics and kinematics, control systems, sensors and transducers, power storage and charging, it gets pretty complicated.

Then there’s the intelligence question. What is intelligence? Can we program it? This is a little more difficult to separate out. Many systems exhibit some of the attributes of intelligence because they are able to adapt to their surroundings. As those who have participated in the recent DARPA funded Autonomous Vehicle challenge, navigation and proximity issues can be programmed, but real long distance travel is not easily rendered into programming. To what extent can we program a system to learn from experience? And what will it be able to do once it has? The limits to software are constantly being expanded.

Over the last thirty years progress have been made on many fronts. Practical robot welding systems are available at $55K price range that many manufacturers are able to afford. And they do everything expected of them in terms of productivity, consistency and speed.  Million dollar Surgical robots now exceed the performance of skilled orthopedic surgeons.

I think its an exciting time in our industry in terms of what is possible. Economies of scale, while sometimes slow, are impacting all aspects of mechatronics. I hope the students coming into the field will be able to apply their skills in solving problems and really use their creativity.  On the Big issues and the small.

No Comments