This may sound corny but we live in the land of plenty. As Michael Pollen, the author of “Cooked”, puts it “If you are what you eat, and especially if you eat industrial food, as 99 percent of Americans do, what you are is ‘corn.'” http://www.ecoliteracy.org/essays/we-are-what-we-eat
But I don’t want to be corn. This is where the trend of mass customization it setting in. The industrial revolution was all about creating factories that could deliver massive quantities of identical product so that everyone could have some. The logistics system delivered full skid loads of boxes of cornflakes which populated the grocery shelves. The advent of computerized ERP has changed all of that. We now want to head to the store and grab the flavor of Gatorage that suits our fancy. We just expect that the shelf will be restocked with the correct flavor without giving a second thought how the modern computerized ERP systems orchestrate that action on a JIT (Just In Time) basis triggered by scanning at the cash register. The plant that bottles Gatorade does not want to have any food remaining at the end of production each day to control bugs and pests. I am happy about that because it is a whole lot better method than using pesticides that could end up in the food we eat. Since the buying public is unpredictable the ERP system has to assemble a list for the the food warehouse of the needed flavors and quantities. The warehouse assembles (kitting) skids of the correct flavors of 80 lb gastorade concentrate packages in production sequence (in e-commerce this is called fulfillment) A delivery run is made each day so that the correct raw materials are available at the start of each production day at the bottling plant. This is just one stop on the milk run that delivers ERP kitted bulk food products from the warehouse in the working day. We all got tired of eating cornflakes every day. Computer directed mass customization systems delivery the variety of goods that we want to the store shelvesConnectivity
The first long distance communication link was actually digital. The telegraph replaced flags, smoke signals and mirrors as the first reliable long distance link. My dad’s best friend was a HAM radio enthusiast who worked for NASA. As a young child, I remember watching a teletypewriter translate radioed Morse Code and print at 20 words per minute. I thought the transmission was coming from Mars. Funny on today’s scale, in that this transmission was at the blinding baud rate of 8 bits/second. (still faster than you could send and receive Morse Code using the key.) The general public never wanted to learn Morse code so they migrated to the telephone, radio and television. My parents cottage / retirement home is on Lake Erie in a small village called Saybrook.
No it was no longer necessary to climb the pole to answer the phone. But, I can remember when a live operator came on the line to ask you what number you were calling from for billing. For us non-residents, it is mandatory to keep the cell phone plugged into the car charger so that you have enough broadcast power to last though a normal phone call. I am not surprised that Saybrook was a Finnish community. The equally sparse Finnish landscape prompted a Finnish company Telenokia in 1991 to invent the digital GSM G2 cell phone protocol. A digital link, initially at 40,000 bits/sec (the old telephone modems are at 48,000 bits/sec) replaced the G1 analog cell phone networks. This digital strategy made simultaneous voice and data transmission possible. We quickly adopted this communication link. Connected car services such as my IoT (Internet of Things) http://mileagetrakker.com automated milege report generation system became possible. Mileage Trakker even works in places like Saybrook which isn’t scheduled to upgrade beyond G2 cell coverage until next year.
Humans are funny creatures. In the near future we will all be transported by self-driving vehicles. I suspect that the resulting traffic jams will cause us to want enough cell bandwidth to synchronize the flow. This is very similar to the task of keeping track of Kiva robots scurrying around a warehouse. Very high capacity “blanket” WIFI systems at 450,000,000 bits/sec by Extricom http://www.extricom.com are the minimum needed for this task. The actual bandwidth we will want is probably higher, in that it takes 32,000,000 bits per second, to transmit TV images to entertain us during the trip. We are watching all of our communication links converge into a single interconnected system.
KISS
KISS — Keep It Simple Stupid
Early in my career I was assigned the task of commissioning an $70,000 electric paint robot. Unfortunately at that stage of robot evolution, the seal design at the robot joints prevented it from painting more than the first part. The vapors shorted it out. What I learned is that it is more effective to tackle automating the simplest tasks first. Simple tasks take less investment to automate. Even though modern 6 axis robots can now handle painting tasks, combining them with required sophisticated guard enclosures places them far from the KISS category in my mind.
I was watching a 2 axis gantry “robot” carry engine cylinder heads at a Honda assembly plant in Alabama. Reminded me of a dog playing fetch with a stick. The jaws of the gripper opened and the human at the end was presented with the retrieved cylinder head. I install gantry robot transport systems for the same benefit, their simple motion and guards allow humans to safely work along side them. These robots do not care that I am comparing them to a dog. As Ada Lovelace wrote as her work began the digital age “The Analytical Engine has no pretensions whatever to originate anything. It can do whatever we know how to order it to perform” Kiva robots are even simpler. They race to the other end of the warehouse and wait like puppy dogs for us to collect the item they have retrieved. All of this without the clutter of the rails and guards of gantry retrieval systems.
Moore’s law marches on. The cost of computers has dropped to the point where we can afford to automate even simple tasks. It is easy to write down date, starting odometer, location, finishing odometer and purpose for one business trip. You get pissed off enough to invent a better way when you spend a chunk of every Christmas vacation inputting a list of a thousand or so business trips so that you qualify to keep the $8000 to $9000 dollar business mileage deduction. I want to use the KISS principle to fix all the aggravations in my life. It is nice that an affordable computer http://mileagetrakker.com can talk to the car and generate the log for me
Artificial Intelligence
Back in my early engineering days I was using the largest Cray Research computers in the world to analyse the stress and mode shapes of cars and trucks. This only worked if we could simplify the finite element test to the point where the computer could complete the calculation starting at 7:00 pm and stopping at 5:00 am when its task changed to supporting CAD. When I discovered that one of my friends had a personally owned computer of this magnitude at his desk I was intrigued. He had discovered that neural networking programs that simulate the way your mind tackles problems could be used to make a 6 figure income on the stock market. I linked him into some university researchers who discovered that complex numbers were more effective within neural networking programs. He traded this improvement into a 7 figure income. This is where computerized artificial intelligence shines. Keeping track of enough of the whole economy to succeed at managing a stock portfolio is beyond our mental capacity.
The engineering community continued to advance its versions of artificial intelligence. We call it adaptive control. Early examples include electronic spark advance. By 1982, it was finally possible to purchase power transistors that were capable of handling the spark control task. Later advancements included electronic fuel injection, which continuously adapts to dirt buildup, wear. and driver behavior. In military applications, stealth bombers and Ospery aircraft are beyond the ability of pilots to fly without adaptive computer assist. Along the way, I invented an industrial process which an operator could not supervise without the help of my patented adaptive control.
There is a perfect storm approaching. It will affect the people working for the big accounting firms the most. During World War II the Allied intelligence community invented what became to be called content analysis. They guided the war effort by combining the information gathered from every publication, like company newsletters and local news sources, that they could get their hands on. Any one item was not useful by itself. The company owners (stockholders) created the concept of accounting firms to look after their financial interests. The crooks who orchestrated the major scandals such as the World Com collapse took advantage of their incestuous relationship with the accounting firms doing the stockholder oversight, to mask their actions. Firms that are actually hired by crooks are reluctant to blow the whistle. The advent of big data combined with the evolved capabilities of the neural networking side of artificial intelligence are staged to crash onto the scene. Computers using artificial intelligence and content analysis will detect an impending company failure. The resulting analysis will enable a few stock investors to sell and get out. The stockholding public will have their confidence shaken. The lack of confidence in traditional accounting oversight will cause an accounting lean revolution where new more efficient controls emerge. We are all in favor of lean improvements until our own salary is the target. Lean improvements in manufacturing have left accounting expenses as a big target for computer technology inspired cost reductions. Usually this occurs as a season change in method.
Connect the Silos
Anyone who has attempted to implement lean saving discovers that most companies operate inside little silos. The marketing department does not think that they need to link their actions to sales generated or prospects discovered. The engineering / product planning group believes that someone else should define the detail specs for the new product to be developed. The accounting group wants perfect projection of all future expenditures. The sales group wishes that the custom offerings desired by each customer can be delivered in the same time and same price as the standardized offering. We have all been around this merry-go round.
I started to wonder how we evolved to work in silos. I suspect that our upbringing is a factor. Our parents created a sheltered cocoon so that we could grow to the point that we could face all the challenges ourselves. Many of us first experienced swimming in the safety of a wading pool in our back yards.
A few of the high filers grew to revel in surfing the earthquake waves of the oceans. The work environment throws major challenges at us. I believe that silos have been created as a partial replacement for the cocoons that sheltered us in childhood.
There are high fliers in the business arena as well. In the successful start-up companies they are able to navigate in company size silos. Companies like GE have discovered that grouping high fliers into their own start-up silo is a effective formula to kick-start innovation. I am not sure whether the silo created for this purpose was an invention of the rest of the GE players. The jury is still out on whether it is possible to attract and retain a group of players who are comfortable in working in a company size silo. We will all be watching the Zappos experiment in holocracy
Rough Road Ahead
We all cringe when we see a rough road sign in our path. The jagged accelerations chatter our teeth and shake us to the very bone. Immediately we want to get the road crews out there smooth the rough edges. A die casting machine reacts much the same way when abrupt changes of speed are commanded. We have all seen the “chatterbumps” in the shot traces as the control struggles to obey our commands.
I can remember making barbecue lids on a machine that effectively had a one speed shot. It was necessary to trigger the fast shot almost immediately to have any chance of accelerating to the 120 inches / second needed to fill the casting. I suspect that there are still some Lesters still working out there with a two speed shot.
Having experienced lots of “chatterbumps” it is refreshing to have the capabilities of a modern control. The Visi-Trak control
http://www.visi-trak.com/Media/Vann_Proof_withAd.pdf
running the rebuilt Prince at Cana-datum has up to 6 settable speeds and 6 settable accelerations. I was an old fashioned 3 speed shot proponent until I experienced a full speed launch of a transmission case. Applying targeted smoothing of the speed transitions made it possible to move the porosity out of the seal bores.
Having been trained to improve porosity issues by tuning gate and runner shape, I was excited to discover that tweaking of the fill profile actually was helpful and a whole lot easier and quicker to implement.
Since I published this post, my contacts at Visi-Track have given me this link because it is related.
Following is a link to where you can download the Ohio State University paper called “Experimental Determination of Slow Shot Velocity Position Profile to Minimize Air Entrapment”:
End of the Wild West
The wild west was actually a very short period of American history spanning about 40 years. We have immortalized this time because rugged individuals tamed and conquered the land. It came to a close, an organized society came in and imposed law and order and fenced in the prairies. It still remains in our hearts captured by countless books and movies.
The personal computing wild west will also span about 40 years. We are already seeing memoirs (Steve Jobs is and example) of the larger than life cowboys who roamed the plains. It will follow the pattern of every other era. When the calculators first came out spurred by the HP35 splash onto the scene, there was every imaginable version available. New and improved versions seemed to appear monthly. After a while a consensus of what a calculator should look like and what it should do emerged. Collectively we realized that it was not necessary to have a new definition of what 2+2 equals with every new release level. I remember giving my new HP45 calculator to my younger sister (also an engineer) and asking her to add 2+2. She failed because the reverse polish notation used on a HP45 makes the keystrokes 2 ENTER 2 PLUS. The modern solar cell powered calculator I use now cost $30 and has been doing the same functions as my HP (no reverse polish) for the last 15 years. The wild west in calculators came to an end.
The same evidence of the end of the personal computing wild west is showing. Open office is every bit as functional as other office suites. I suspect that we will also get tired of daily updates in operating systems that corrupt the other software that we are trying to use by separating computers dedicated to tasks from the unpredictable internet environment. Windows NT (the industrial XP version) is extremely stable as long as it is separated from the web. I suspect that hardware and software innovation which lock in program and operating systems so they cannot change will show up shortly. Yes
the wild west is being fenced in.
Lean Automation
I have to admit that many years of installing robots to reduce labor has blinded me into thinking of automation as robots to be installed. This matched the marketplace because the $200,000 robots morphed into $50,000 robots opening the door for new applications with each price drop. Shop floor manufacturing cost which was $.50 of every sales dollar in the 1950’s has steadily declined to $.20 of every sales dollar today. Since 80% of the current business costs are not on the manufacturing floor, it is not surprising that the best opportunities for reducing cost by automating are no longer on the shop floor.
It would be naive to claim that no automating activity has occurred off the shop floor. The PC, then smart phone revolutions, have brought the efficiency of software to our fingertips. We would not even consider typing and mailing a letter anymore. PC Office suite software is now available as “free” open source downloads. The more recent smartphone market is equally mature. Even as early as 2010, 90% of the smart phone apps sold less than 100,000 copies, which suggests that they failed to cover their development costs.
The 80% share of business cost off the shop floor shows that the current opportunities are still away from the shop floor. The developers agree and the latest excitement is centered around innovations that combine software and hardware. This is consistent with what I am seeing. Somewhere in the future your cars will have enough connectivity and spare computers or your cell phone will have enough battery capacity to handle extra tasks. In the short term dedicated devices such as the mileage trakker http://mileagetrakker.com that I market are good examples of the latest trend. Business people who travel are happy that a device plugged into the car handles most of the assembly of an IRS format mileage report so that they can receive their significant mileage deduction at income tax time
Demand
The introduction of new electrical users such as electric cars are keeping electrical rates on the radar screen. I tend to agree with Elon Musk that this will cause the electric utility industry to expand their infrastructure. Usually changes like this are accompanied with an increase in rates. Most businesses will be pressured to consider actions that will control the cost increase. Reducing demand charges is one of the best ways to continue to obtain the energy needed at a more affordable cost.
“Saving Money by Understanding Demand Charges on Your Electric Bill” — This Tech Tip explains demand charges, how they are calculated, why we need to monitor them, and what we can do to lower them. Demand–related charges usually represent 30 to 70 percent of most commercial customers’ electric bills. Electrical demand in some parts of the Nation stretched the capabilities of electrical generating and transmission systems during the summer of 1999″ Dave Dieziger, Project Leader US Department of Agriculture
Many process cells have a spike load which is much higher than the average load. This fact opens the door for reducing demand charges
Leveling the peaks- The size of the wires (and generating equipment) is directly related to the maximum demand that your equipment requires. Historically we had few affordable options for reducing the peak. With the advance of modern computerized electronics, better options exist. In a recent installation a soft start on a 200HP motor proved very cost effective. Even discounting the future energy saving, it was possible to cover the extra cost simply by reducing the size of the wires feeding the machine. It is interesting that more than the power company is concerned with wire size.
Automation is becoming a popular business survival technique. Virtually all automation requires electric power. The wires feeding an existing cell (or transformer feeding the plant) may not be sized to support added automation or robots and handle the spike loads. In many cases it is cost effective to reprogram the PLC computer running the cell to level out the spikes. This frees amperage to run the added automation. In the die cast rebuild I accomplished last year this turned out to be nothing more than adding program steps to insure that all users did not grab power simultaneously. 100 amps out of 450 amps became available for auxiliaries. Yes a slight increase in cycle time for this change occurred. After I was done, this cell in total had a quicker total cycle time because I overlapped other actions.
Filling in the valleys – We are seeing process controllers, software, tooling and equipment evolve to the next technology level. In mold making this can be disruptive. Historically we we pleased when the cutter path software could simply create the shapes. We used large margins to insure that power demand did not exceed the supply. Most of the time the mill motors were barely off of their minimum power.
Recent advances in milling cutters and driving software have made it possible to stay at 70% spindle power instead of 10%. The reduction in time from 3 days to 1 for removing the 6000 lbs of chips to create a mold holder is disruptive to the CNC build marketplace. The CNC machine builders will have to adjust to selling less machines.
Storing energy – Many hydraulic powered machines use accumulators to store energy and balance demand. The original equipment builders in many cases did not work with the end user to optimize this storage.
My recent rebuild at Canadatum showed that it is possible to level the power usage of the machine by paying careful attention to the timing of the recharge.
Combined together engineered power management has can knock 10% or more off of the power bill.
Mechanic Shortage
We probably will see an end to the Mechanic shortage when the little Greysons of the world enter the workforce. In the interim the retiring baby boomer multi-craft maintenance technicians are leaving a big hole in the workforce. It would be nice to think that the workers entering the market could instantly fill the hole. The reality is that the retiring workers gathered their broad spectrum skill over a lifetime in the field. As a double whammy the adoption of automation as a solution to other competitive issues has expanded the need for maintenance mechanics.
Most people still in the manufacturing game are well aware of the problem. The question is what to do about it. I have elected to list a few avenues to consider.
1) Factory rebuilds. Ralph Nader in his ignorance made planned obsolescence a generally known term. As engineers we are proud of our ability to design products with a known lifetime. This term can be good or bad depending on whether you are trying to use mechanical items beyond their design life. Smart business people recognize the benefit of rebuilding back to better than new on a planned schedule. For example the life expectancy of the first rebuild of a diesel 18 wheel truck engine exceeds the life of the original engine. Trying to staff enough maintenance talent on the third shift or weekend to keep equipment past its planned life running is a losing battle
2) Offsite maintenance. The thing I like best about Kiva Robots is their interchangeability. If you are automating a warehouse it is not necessary to hire a large staff of robot mechanics. These items can be sent offsite and quickly replaced with identical replacements. Many industrial trucks and lift trucks fit into the same category. Contrast this with central conveyor systems that shut the plant down if any section is not operational.
3) Autonomous Maintenance. In today’s lean environment we have to get past the notion that only maintenance craftsmen are capable of keeping the plant running. Operators know their work area best. Systems can be put in place to guide each operator in performing actions which prolong equipment life and improve uptime.
4) 5S. Shine is a key factor in keeping equipment operating reliably. Many times a repair is nothing more that cleaning within the mechanisms. The time for making a repair is diminished if the mechanic does not also have to clean the outside.
5) Team maintenance. Scarce multicraft technicians are a lot more productive if they have an assigned helper. This effective strategy also is helpful is broadening the skills of apprentices.
6) Scheduled Maintenance Free. Smart business people are purchasing items that require less maintenance. I was talking to a Tesla owner who only changed tires after three years of heavy driving. This is possibly why they do not need a dealer network. Wesley an industrial truck builder changed to flat-free aperature tires and maintenance free batteries to offer a vehicle without scheduled maintenance.
7)HMI. (Human Machine Interface) Much the same way that it is next to impossible to fix your car without obtaining the error codes, the computers attached to modern production equipment need to identify what is wrong.