Doing More In the US

The old business school motto, doing more with less, can have some interesting applications.    American manufacturing is going through a rennaisance of sorts.  Across many industries there are substantial efforts to bring more manufacturing back to the US.

A lot of it is precisely ‘doing more with less’.  How do we make the same quality of parts at prices low enough to compete with foreign competition?  It’s not easy when the typical pay scale for manufacturing labor is $8 a day in some parts of the world.

There are a couple of obvious components to price competition that don’t get a lot of attention.  Scrap rates and delivered cost.  When a US company buys parts from offshore, any defective parts are very costly.  The direct shipping cost, duties and processing fees are additional and can be 10-15%.

During my years at Rockwell Automation, we investigated the cost of selling US products in different parts of the world.  Depending on where in the world we are talking about, the shipping and logistics can accumulate between 25 and 40% additional cost to the product being sold.

So the cost of scraps and logistics are the minimum cost hurdles for companies seeking to export their products to the US.  Low cost producers have to make parts cheaply enough that the landed cost and scrap rates cost out less than the price of producing them in the US.

Doing more manufacturing in the US requires finding creative ways to lower costs.  That is the second area that is undergoing change.  American manufacturing technology is finding ways to reduce machinery and process costs.  And this area of effort may provide key strategies that will help the US gain back ground in the pursuit of more world class manufacturing.

Innovation processes like additive manufacturing allow fabrication of metal parts with no machining.  For higher levels of precision there are new machine tools that can do final machining to less than 0.001″ accuracy and the costs of machine tools are lower than ever.  These are the keys to producing high quality parts at lower costs.

There are unique mechatronic solutions that can improve machinery performance across a wide range of applications.  The Acme screw which is very inexpensive, has limited accuracy but plenty of torque handling capability.  What happens if you can add a very inexpensive linear feedback technology to the simple low cost Acme screw?  You get a really high resolution linear motion system that is very inexpensive.

The great news is that these products are currently available.  And that means that making better machines that make better parts at lower cost is practical, achievable and there are no technical challenges.  Common off the shelf parts will get it done.

The Next Industrial Revolution

Modern manufacturing is largely the result of Henry Ford’s innovation, assembly line mass production.  The goal of which was primarily to make cars available to large numbers of people due to significantly lowered costs.   No other single innovation has contributed as much to increase the quality of living conditions throughout the world.  Mass production has made more goods available to more people in more places than any other system in the history of mankind.

The electric light, for example, which was coveted 100 years ago as the great solution to night time darkness, making obsolete the candle or gas lamp.  Mass production has made the light bulb an inexpensive  commodity on the verge of extinction at about 25 cents per bulb.  The desire to reduce energy consumption is ushering in the age of the light emitting diode (LED) as the replacement technology for electric light.  Every effort is under way to reduce LED costs by any means possible so that illumination will be available that is even cheaper than incandescent lighting when the energy cost over ten years is factored into the new technology.

Even generating and delivering electricity is the result of applying the principles of mass production.  Large generating facilities are able to generate power cost effectively through economy of scale, selling the power profitably at 4.5 cents per kilowatt hour.  Wire, cable, switching systems and other infrastructure are generally costed in at an additional 2 cents per kilowatt hour to deliver the power to your door.  This is an incredible deal, trillion of dollars of resources at your disposal for pennies.

But mass production is not the answer for every aspect of modern society.  Lowering the cost of mass-produced goods implies that there is a requirement for the sufficient numbers of a product to warrant the investment in the necessary processes and tooling to accomplish the task.

Enter 3D printing technology.  Also known as “Maker bots”, this new class of tools is making fabrication a  new American pastime at incredibly low cost.  Where 3D printing equipment has recently been the domain of well-funded large corporations , selling at $10,000 to $20,000 each, 3D printer kits are available at less than $1000.  And lest you think that these are only toys for boys, the additive manufacturing paradigm has taken hold in the metals industry producing high quality parts in various steel alloys and even in titanium.

Why does it matter?  Because anything that lowers the barrier to market entry for new products creates the opportunity for people to enter a market that was previously inaccessible.  The hidden relationship is financial, it is the cost of amortizing the manufacturing resources across a given number of products that makes startup of a new product impractical.  So barriers to entry in new product development are primarily the result of amortization costs.

What happens when a new technology introduces a significant reduction in the amortization cost?  You get the opportunity to experiment with things because the cost of iterating the design is low.  New products can be test marketed and improvements made because there is no major investment in tooling that would have to be modified in order to change the design.  You don’t have to get it right the first time.

And that means that anything is possible.

Next Generation Manufacturing

As a follow on to the last post, I have been investigating the cost of manufacturing equipment.  The classic machine tool is the most widely used piece of equipment for fabricating just about anything made out of metal.  The machine tool has been quietly undergoing it’s own revolution since it’s inception in the 1950′s.

The traditional metal cutting machine tool has been around since the 1800′s and was entirely manually operated.  Since the machines were manually operated, the dexterity of the operator became a major factor in accuracy and repeatability of part manufacturing.  Because of the skill required, we still have the term “master machinist” in circulation, even though most machining today is automated.

During the Second World War, the Air Force was confronting the difficulty of manufacturing airplane parts.  Through the work of John Parsons and MIT, the first “punch card” controlled machine tool was built.  Parsons’ company was using early punch card computers to generate a larger number of points along the curve of a wing brace.  The numerical information was then used directly by machinists as a look up table for manually positioning a milling tool.  Parsons realized that if they could motorized the manual process, it could greatly increase the speed of the machining process, lowering costs dramatically and increasing accuracy at the same time.

Gordon Brown’s Servomechanisms group at MIT has recently been working on early forms of closed loop dc motor control for the gun turret on B-29 bombers.  By combining these recent technologies to numerical punch card calculation approach the first Computer Numerical Controlled Machine Tool was demonstrated.

The rest, as they say, is history.  The lessons learned in computer numerical control have been instrumental in every major field of manufacturing.  Cars, electronics, robotics, would not be feasible or cost effective without the underlying control technology of CNC.

Which brings me to a 2 major points as we contemplate the next generation of manufacturing.

Additive manufacturing is maturing rapidly with a wide range of materials, steels and titanium are now available, and precision is improving at the same time.   The surface finish requirements for a large number of parts cannot be achieved with a strictly additive process.  The new wave of additive manufacturing requires a complementary subtractive technology at complementary prices.

Secondly, while there are an increasing number of machine tools at low cost, they are not CNC.  This will likely be the next “breakout” technology.  There are a number of technical hurdles that have to be addressed in terms of reducing the cost to a level comparable with the Makerbot.  With the current generation of dedicated motion controller chips, lower cost step motors and low cost feedback technology, this should be a slam dunk.

Get your pencils out and get after it!  There’s some serious money to be made here.

2012, Year of Opportunity

2011 was a difficult year for many sectors of the US economy.  World markets were about the same as at home.  The general weakness was felt world wide with currencies, especially the dollar and the Euro, declining due to bank failures in EU, financial crises in Greece and Italy, unrest in the Middle east creating concerns about stable supply of oil, runaway spending in the US and low sales numbers in new car sales and new home sales, two of the “bellweather” indicators of economic strength.

Alternative energy in the US has failed to produce the return on investment or to create new jobs in any significant numbers.  Car sales have picked up significantly over 2011, but not nearly at the rate of 16 million cars/year as in the heyday of that industry.  The real estate bubble has burst after a decade of speculation and bad lending practices that continue to depress new construction.

The second industrial revolution, Henry Ford’s industrial revolution, was about mass production and cost reduction.  For almost 100 years we have been perfecting the centralized manufacture of almost everything around us.  Economies of scale that enable volume manufacturing of consumer electronics at lower cost year after year are the result of the Ford approach to manufacturing.

So where do we go from here?  We start re-inventing the industrial revolution.

The new wave of manufacturing has begun with the advent of the Maker Bot and a family of low-cost fabrication tools that can manufacture based on 3D printing techniques.  While solid model prototyping has been around for some time, the magic ingredient is a new family of machines that cost less than $2000. and some recent new entries around the $1000 mark.  At these price points, it doesn’t take a lot of volume to justify the purchase of one of these machines.

There are advanced processes that are becoming available to generate sintered metal parts, even titanium parts, using processes resembling the additive manufacturing process.

Amortization cost is the secret.  Lowering amortization costs and minimum order quantity at the same time results in a fantastic breakthrough in productivity.  It also lowers the barrier to entry into new markets.  So if you have an idea for something that’s never been done before, the cost for development may be a lot lower than you think.  And thousands of people have begun to jump into the mainstream economy because of the availability of these new tools.

While the “maker” tools are limited to plastics, there is progress in the metals arena as well.  The computer numerically controlled (CNC) machine tools have traditionally been the domain of 6 figure costs, HAAS has been making $50,000 machines and last year Tormach entered the market with the “Personal CNC”, a high quality machine that is priced at $10,000.

My prediction?  There is going to be a lot of activity at the $10,000 and below price point to come up with low cost machine tools as a complement to the “maker” bot 3D printer technology.  Additive manufacturing will require a complementary subtractive manufacturing infrastructure at a comparable price point.

And creative American engineers and tinkerers will be leading the way.

Friction, Friend or Foe?

Friction is rarely talked about in motion control circles (pun intended for those paying attention).  It is the “waste” energy in mechanical systems.  We spend a lot of time and sometimes cost, trying to eliminate it.  Many times we just ignore it.

This was the case when a friend of mine was designing a material handling system for newspaper bundles.  A very exotic conveyor system with 8 servo driven belts and a design that involved 10 pages of hand calculations of inertia.  We shipped the servos and sent out a field engineer to start up the project only to find out that the motors and drives were too small.  The designer had forgotten to account for friction.  In this case the frictional load was 3 times the mechanical load due to the unique belt and roller configuration.

So the first lesson is; don’t forget to look at friction as 1 of 3 components of the torque load.  The three being; steady state torque, torque of acceleration and friction.

Then there is the fanciful wishing that there wasn’t any friction to worry about.  Kind of like doing experiments in the space station and having no gravity.  It’s fun to think about, but there are few real world situations where this is likely to work.  The only exception is air bearings.  Of which there are a few.

If you have ever played air hockey, air bearings are like that.  Parts in motion tend to stay in motion when there is no friction to worry about.  And that would be good in a lot of applications.  No friction will generally result in smaller servos, so there are savings in the hardware requirement.  No friction means no mechanical wear, nothing to service as the machine runs up cycles.  No friction also means high speed motion is a lot easier to achieve.

Cars coast to a stop because of friction.  That’s a good thing.  Without friction, parts would end up flying off the conveyor instead of going where you want them to go.  In conveyor belt applications there is usually a lot of friction and that helps the system slow down and stop.

So the second lesson is; friction can be your friend.

In between systems with friction and systems with no friction, there are rolling bearings.  Systems like the Bishop Wisecarver “Vee Guide” are among many products on the market are examples of this.  Rolling element bearings have very low coefficients of friction, so losses are low and therefore the energy needed to overcome them is very low.  This also results in very low wear, so maintenance on this type of mechanism is also low.

The are dozens of linear actuators on the market and each vendor has developed unique bearing solutions, whether sliding or rolling, that perform well at varying price points.  There are no universal rules for selection.  The typical criteria are move speed, positioning accuracy, life expectancy and cost.

Programming Software and Control

Writing software for control applications isn’t the easiest thing.  It’s probably been 15 years since I had to write any actual code for a control application.  I have scripted and taught training classes in PLC programming and am very familiar with ladder logic programming and Boolean instructions, although I am less familiar with the latest editing software which has become very sophisticated.

A few weeks ago, I had the opportunity to write a couple of small “C” programs in a training class and re-discovered why I don’t like to write control software.  I don’t have much background in C programming.  It’s not that C programming is inherently good or bad, it’s just another language.  What is difficult to deal with is each controller having it’s own library of C language instructions.

It’s not that any particular language difficult, it’s that every language is iterated on different controllers and the instruction set and programming quirks have to be learned on each platform.  Ladder Logic instructions have become largely standardized and the difference from one platform to another are becoming less and less significant.  Turning discrete inputs and outputs on and off is pretty straightforward.  Reading analog signals, doing some mathematical operations and setting analog outputs is also fairly straightforward.   Even when there a lot of I/O to deal with, the knowledge base required to understand the applications of the technology are ultimately very repeatable.

The variations of how to do motion control on different platforms are very significant.  Each controller company has to come up with a complete programming environment that defines how to command the controller to execute motion tasks.  The creation of commands and processor executables requires coding and testing the code over man years of development.  This is a complex form of knowledge capture and there are a lot of nuances as programmers come up the learning curve before good effective programming environments can be created.

This is part of the reason why the motion control field hasn’t progressed as much as other control disciplines.  There is no agreement on any standard programming methods past trapezoidal move profiles.  The situation becomes more complex because each motion control vendor develops its own programming environment based on the selection of processor platforms and what its programmers come up with for the programming suite.  This creates a barrier to entry for new companies, and makes improved code solutions problematic.

Many of the PLC programming suites include dialog boxes that provide scripting for the motion commands in the ladder logic program.  The technology is readily available to make a high level motion programming suite that is processor independent and capable of addressing 80-90% of all motion control applications.  This will make motion more accessible to a wider audience and simplify the programming aspects of motion and machine control.

We need to bring the industry into the 21st century and make everyone’s lives a little easier.

Tranportation, Economics and the New Solutions

In the US, and most industrialized economies, the second largest expense of individuals and households is personal transportation.  Given the hundreds of man-years in development of the technology, the extremely low cost and high energy density of gasoline and diesel as fuels, it is not surprising that the dominant means of transportation is combustion powered.  Cars, buses, motorcycles, even bicycles are powered using the same basic combustion approach.

Considering the possibilities of improved personal transportation, the consequences of a major change in transportation technology are significant and should be carefully considered as we move forward.

The major impact of all the technology being promoted these days is increased efficiency and reduced fuel consumption.  Whether your motivation is reducing emissions and cleaning the air, or you are interested in reducing your cost of transportation, the requirement is the same; get more miles out of a gallon of gasoline or eliminate gasoline usage altogether, as is the case for a pure electric vehicle.

Across the entire population of cars in the US, the average fuel efficiency is around 20 miles per gallon. Despite the demand for higher fuel mileage from consumers, this situation hasn’t improved much in the last few decades.  A dismal thought in contrast to the claims being made these days for the new solutions.

The US consumes 383.25 million gallons of gasoline and diesel fuel per day.  This all goes into transportation.  The only fuel going into electrical generation is in diesel gen-sets for backup and remote power, just in case anyone is thinking about the barrel- of-oil-to-electricity energy equivalency.

Imagining a future in which gasoline usage declines is not difficult.  I drive a Ford Fusion for work which is averaging 30 mpg combined city and highway.    If the US fleet average is 20 mpg, increasing that average to 30 mpg implies decreasing the amount of gasoline sold by 1/3.  Currently, gasoline retails for $3.25/gallon, or $453 Billion annually at the pump.

So a sharp change in usage due to efficiency or an increase in the number of electric vehicles, is cause for concern from oil & gas exploration companies, gasoline refiners, distributors and dealers.  Unless gasoline prices continue to go up.  In which case there would be less gasoline solid at roughly the same total revenue, which suggests that higher profits might be the side effect if the true cost doesn’t go up.

What about tax revenues?  The direct state and federal tax on gasoline is about 40 cents per gallon.  This does not include large excise taxes collected by the states, taxes paid by refiners and distributors, etc.  In fact, it would be hard to calculate how much of gasoline pricing is taxes and how much is the cost of the product.  Regardless, at 40 cents/gallon, the daily revenues are $153 million and the annual is above $55.8 billion.

Given the current economic picture, is there any level of government that is willing to give up the tax revenue from gasoline?  Probably not.  Is this any different than “Dollars for Oil” at the UN a couple of years ago?  Probably not.  But we thought that was a scandal.

EV’s Everywhere, and More!

Alternative energy fans are getting  good news this year end, 2012 will be the year of the electric car.  No matter what flavor of technology, dual drive train hybrid, true hybrid, plug in electric, there will be something for everybody.

American buyers will be able to buy American hybrid cars.  The Chevy Volt will be flanked by the Ford Fusion Electric scheduled to be released for sale in 19 US markets in March of 2012.  The Nissan Leaf might be the first production electric, so most commentators will make comparisons regarding driving range, speed and recharge time based on the performance of the Leaf.  At present, the claimed performance of the vehicles is very comparable.

It’s all speculation until there are a few units out there and the actual life cycle of the batteries can be measured.  100′s to 1000′s of vehicles will have to be built and consumer experiences cataloged in order to get a handle on how the batteries really work.  With all due respect to the development and testing efforts, it’s educated guesswork until there is real world experience.

Will the batteries be able to cycle enough times to make them cost effective?  When will they require replacement?  What will the price tag be for the battery pack?  Hopefully less than the $13,000 Tesla battery pack.

EV’s are coming.  But they are, like all the alternative energy technologies, still not cost competitive with Internal Combustion engines.  Most vehicles carry a $39,995 starting price tag with a $7,500 Federal rebate.  The basic purchase price puts EV’s out of the price range for many people, which fundamentally defeats the purpose.  The point of alternative energy technology is that it must become widespread in order for any impact on the environment to take place.  High prices are a major barrier to broad adoption.

Meanwhile the internal combustion engine is seeing some revival.  New approaches are being built and tested that offer dramatic improvements in efficiency and engine weight.  The EcoMotors opposing piston engine has been under DARPA development since 2007.  EcoMotors technology has been demonstrated to 40% efficiency, more than double that of traditional ICE.  In addition, it weighs less, takes up less space and gives of dramatically less heat.

Recently, the University of Michigan announced a new breakthrough called the wave engine that is expected to increase combustion efficiency to 60%.  And the rotor only turns in one direction like a scroll compressor instead of a piston, so there are no reciprocating motions to deal with.  This will also lower vehicle weight substantially, so the engine efficiency improvement leads to further overall efficiency in fuel required per transportation mile.

If these ICE improvements translate directly into miles-per gallon, then based on average 20 mpg cars today, we are talking about 53+ mile per gallon in town and possibly 70 mpg highway for EcoMotors solution.  At these levels, the equivalent energy cost per transportation mile is at parity with electricity.  If the wave engine proves successful, in town ratings of 80 mpg and 100 mpg highway become feasible, making electric options more expensive.

The future is what we make it.  Let’s make it the best we can with choices that make sense economically and environmentally.

Control System Theory and Feedback

Control system performance is based on feedback.  Control of electric motors, however, continues to be a bit mysterious because the common conventions associated with motor control are often driven by cost considerations.  The feedback component is often target for elimination in cost constrained systems.

Control systems can be described as “open loop” or “closed loop” depending on the whether or not the system being controlled is well characterized.  Many forms of motor control seek to be “open loop”, that is, without the use of a feedback device.   However, this notion should be modified to open loop meaning without an explicit feedback device. This is because great effort is expended to “infer” what is going on in the motor through various means. The most common of which is current.

In the world of electric motors, the alternating current motor of Nicola Tesla is well understood, and rarely requires a feedback device.  Motor speed is derived from the frequency of the power being supplied minus losses depending on the details of rotor construction and how a specific load affects the motor.  The standard ac motor has a small amount of rotor “slip” from 1800 rpm to 1750 rpm which reflects the magnetizing current losses in the motor and magnetic features in the rotor that would be needed to maintain perfect synchronism with the line frequency.

Load variations can be measured by sensing the current in the line going to the motor.  So there is a feedback element available from which a great deal of information can be derived.  This is where the ambiguity about feedback comes in.  The current needed to run the motor with no load is fixed value, so more current read on the motor leads is load, until the motor reaches locked rotor current or stall.

In brushless dc systems a similar approach is used.  Detecting the zero crossing point of the phase current establishes precise timing of the rotor speed and is used to regulate timing of current pulses to all three phases of the motor.  In this way even the brushless dc motor can be operated without an explicit feedback sensor.    However the tradeoff here is very poor low speed regulation of the motor which makes this approach unsuitable for many applications.

From a control system standpoint, feedbacks are the last, slowest loop in the control scheme of the motor.  This makes sense in the context of position control as it is normally executed in a PLC or motion controller.  However, this makes load regulation more of a challenge since the actual error detection of the control system is being done a level removed from the actual load.

A host of mathematical tools from the signal processing domain have traditionally been employed to characterize the lag created by the control system and the interaction of the controls at varying speeds. All of which works well, but has also lead to “rules of thumb” that are not very clearly understood and which are sometimes misleading.

Unique Solutions in Material Handling

Moving products around is mechanical work.  When the work is done by a control system and actuators its mechatronics.  Mechanical work, whether by humans, by horses, by hydraulics, electrics or whatever, is still work.  Figuring out what technology approach will be the most cost effective way to get the work done is the challenge.

Many of the constraints of the work are environmental.  If work is being done outdoors, then temperature and humidity are a factor.  Felling trees and in the forest requires extremely high forces due to the work needed to cut through a tree and drag it to a truck to be hauled off for processing.  Processing trees, even in a plant environment, requires some serious hardware, 125 horsepower band saws are not unusual.

Doing work on a ship or oil rig has additional constraints because of the presence of explosive fumes and fuels.  Often the need to avoid any possibility of igniting a combustible atmosphere causes engineers to apply pneumatic control systems.  Yes, there is still a compressor somewhere to generate the compressed air supply, but that is usually remote or contained to avoid exposure to the volatile atmosphere.

Environmental constraints come in many forms.  Extremely high temperatures push the limits of what is possible.  Making glass, semiconductors, and primary metal processing are all high temperature environments where engineers have developed whole technologies in order to bring us the materials we use in everyday life.

The simplest action of rolling or sliding becomes a real challenge when environmental constraints are added to the work statement.  Sawdust becomes a potential abrasive in woodworking environments that can introduce severe wear in moving parts.  Corrosive and explosion proof atmospheres as well as food industry applications introduce all sorts of chemical compatibility problems that require special materials and processes in order to meet strict guidelines for safety.

As always, resourceful engineers have worked out solutions for all of these difficult applications.  One family of solutions to rolling applications is the use of all ceramic bearings.  No steel, no lubrication.  None is needed because the ceramics are extremely high purity to start with and have extremely high precision surfaces eliminating the need for lubrication.  No outgassing or contamination to worry about.

Other solutions take the form of air bearings and non-contact material handling devices.  Air bearings have become more readily available for conventional applications, but are particularly compelling in large machinery applications where precision is required.  Large flat screen display glass  presents unique challenges that successfully addressed using a combination of air bearing regions and vacuum regions to move the glass without actual contact and with overall flatness measured in millionths of an inch.

A unique solution in pneumatic material handling takes compressed air driven into a funnel shaped recess and creates a vacuum in the center and an air cushion at the edges where the air is exiting.  This creates a vacuum pickup that never quite comes in contact with the part, leaving no marks.  Perfect for solar cell and some food and beverage applications.

Engineers continue to meet the unique challenges of industry and create commerce at the same time.  And that’s what it should be about.

Next Page »