We are saddened to report that Professor Dornfeld passed away in March, 2016. If you enjoyed his blog, please consider making a contribution to either of two funds at UC-Berkeley that have been established in his memory.

David A. Dornfeld Graduate Fellowship
David A. Dornfeld Scholarship

Thursday, July 9, 2015

The 3rd Machine Age

Where are all the people?

I recently bought and read a book “The Second Machine Age” by two visionary authors at MIT predicting (touting?) the rise of the next industrial revolution and the age of the robots … like the phrase in Lord of the Rings movie predicting the demise of men (“The Age of Men is over. The Time of the Orc has come." — Gothmog to his army in The Lord of the Rings: The Return of the King). We see here the rise of the machine (again). 

Other recent publications, in a series of books and articles, tout the role of automation and robots in enhancing productivity. In “Plant of the future” (a publication of IndustryWeek) the benefits of computers, automation, robotics and “digitalization” are reviewed with the summary comments: “Automation allows people to complete tasks faster with fewer errors at cheaper costs. This increases productivity, which means people don't have to spend as much time or money to accomplish tasks, which generates new wealth for society. True, if your job was eliminated through automation, you are personally less wealthy. That's the painful side of economic disruptions.”

The benefits are clear for the economy on the front side … the challenges created by the “if your job was eliminated” part are less clear.

First of all, I am not a luddite. I love technology and have spent a lot of time researching and developing aids to automation that make factories and machines hum and perform efficient, quality, cost effective operations. And, I also am aware that many times in the past (just check Amazon for this Second Machine Age book and they’ll  give a list of a bunch of earlier books predicting the machine age and its benefits others have bought) there have been similar predictions but calibrated to the buzz word of technology of the time. Recall a post in this blog back in October 2014 on the digital revolution where referenced was made to a Fortune article from November 1994. That article predicted that the digital factory could have the effect to “stabilize or even increase the number of production-worker jobs in the U.S.” Well … not so much in the U.S. Maybe elsewhere.

But, there is a growing concern about this enthusiasm for the age of the machine - at least among some of us - or at least by me. Part of the concern stems from the simple question “Who takes things apart?” The last posting on the circular economy summarized the results of a study on material flow in the economy as follows: “… the degree of circularity of the global economy measured as the share of actually recycled materials in total processed materials is quite low - only 6%. Most of the  processed materials (66%) left the global economy as wastes and emissions.” This is a problem and it is not necessarily helped by automation or increases in efficiencies on the “input side” of the economy. Let’s look more at this.

You may recall early on in postings in this blog a reference to the Ricoh Comet Circle (way back in 2009 in fact. Reproduced below for reference (from http://www.ricoh.com/environment/management/concept.html) , the comet circle was described as an excellent 


way to represent the “supply chain” feeding the consumer. The forward (counterclockwise loop) at the top is from materials through production to delivery to the consumer and use. The reverse (clockwise loop at the bottom) is after the consumer is done with the product back through recycling, recovery, and return to material supply chain. Usually when a green supply chain is mentioned it is in the context of the return loop - resource recovery. The posting noted that the forward loop is only half the battle and, if that loop is done correctly, the reverse loop is much easier to implement and overall better.

“Better” here refers to design for everything necessary to make the product successful and efficiently producible, energy efficient in use, etc., but, also, adaptable to the reverse supply chain. This is the circular economy! The reverse loops in the comet circle are the return loops in the butterfly diagram on the technical side of the circular economy.

So what’s the problem?

Automation with its robots, assembly systems, mechanized manipulators and tooling put things together. And usually in a way that they stay together. They they don’t take things apart. If you Google “robotic disassembly” one of the top hits is a 2013 posting in Popular Science on “How it Works: The Robotic Chicken Butcher”. Elon Musk refers to the recent explosion of one of his booster rockets as “rapid unscheduled disassembly.” That’s not what we are talking about! OK - I know there are other examples of robotic disassembly - some hardware is available but the field of technology applied to disassembly - for the purpose of repair, remanufacturing, reuse, recovery or recycling - is not at all developed. So how do we move up from 6% circularity?

How about people? What if the design and production of products was done in consideration of the return loop in the comet circle? What if we could take apart for the purpose of remanufacturing (essentially extending the life), repair, reuse, etc. the products that course through the veins of the industrialized economy? And do this for everything - from cell phones to dishwashers to automobiles to airplanes and buildings. That would be a big step forward.

Some are already working on this. A company named CoreCentric Solutions in Chicago focuses on what is called “reverse logistics supply chains” - part lifecycle management and remanufacturing solutions. They work with large and small companies to take excess inventory back, take back for part recovery and reuse out of date or replaced products, remanufacturing, etc. Their website says “Often times parts removed from an appliance after a service call are thrown away. This creates unnecessary waste in our landfills. Recycling appliance parts in a proper manner helps to "go green" by keeping harmful chemicals and metals out of land, water and air.” Then another section of their website is a shopping area where they offer “quality remanufactured” parts for the major appliances among other products.

I have not been to their facilities but I would be surprised if they were highly automated. That’s the challenge! How could you scale such an operation to handle the reverse logistic loops of the circular economy? Any attempts at “disassembly” that have been published have been for a narrow set of products with very specific characteristics (modular, take apart/disassemble from one axis motion, removable fasteners/not welded/glued/riveted together, and so on). These are not usually the characteristics of most products produced today - think cell phones, consumer electronics, wash machines, or automobiles. But these are manufactured with high degree of automation.

So, what’s the solution? What if one used less automation to assemble and produce products (but still could use tools and methods to make production efficient) so that one could more easily  take these apart for the purpose of remanufacturing, repair, reuse, recovery, recycling? But, we’d need to see the numbers! We’d need to do a reasonable comparison of any “losses” due to reduced efficiency on the input supply side of the Comet circle (assuming there are some … not necessarily true if done right!) and how those would be offset or balanced by gains on the output return supply side.

Design for assembly guidelines focus on ways to smooth the flow and assembly of components. Ulrich and Eppinger (Product Design and Development, 5th Edition, Irwin McGraw-Hill, 2012), a leading design textbook, lists the elements of design for assembly as:

     1.    Minimize parts count.
     2.    Encourage modular assembly.
     3.    Stack assemblies.
     4.    Eliminate adjustments.
     5.    Eliminate cables.
     6.    Use self-fastening parts.
     7.    Use self-locating parts.
     8.    Eliminate reorientation.
     9.    Facilitate parts handling.
   10.    Specify standard parts.

And, for automation, one might add to this design for assembly from “one direction” … the “stack assembly” element listed above.

If one reviews the list, several guidelines pop up that, while making assembly efficient (and hence the supply side chain productive), in fact may frustrate the return loop in the comet circle for disassembly. For example, minimizing part counts encourages the use of parts with many materials or elements integrated as one … making replacement of pieces or recovery of materials challenging. Using self-fastening parts means things need to be mechanically or thermally separated rather than just the removal of a fastener. This is even more challenging if adhesives or welding/bonding (non removable fastening) are used.

The good news is that most guidelines for robotic assembly should lend the product to easier disassembly. Parts need to be easy to grip (the facilitate parts handling element above), directions of assembly (stacking) as above accommodates the less dexterous movements of robots compared to humans. So, there is a basis for accommodating both assembly and disassembly.

The real challenges come from “how things are held together” - what engineers refer to as fastening technology. This ranges from screws, to nuts and bolts, to rivets to crimping, adhesives, epoxies, welding, soldering. These are meant (at the weld, solder, adhesive end) to be one time no return joining processes. The image below of one small set of the fastener world illustrates the challenges of accommodating only one type of removable fastener.


 


Rivets are close but can often be removed by a simple drilling process. If one follows all the rules of design for assembly and then creates one giant monolithic block of product with integrated materials and essentially unassemblable components we have a problem. Take a look at most of the consumer products around your home or work place - you’ll see what is referred to- designed to be put together and work … and then disposed of, not disassembled. Finally, all parts flowing together in a symphony of assembly are known as to type, location, etc. Robots require exact location, orientation, delivery, etc. to support automatic assembly. When things get taken apart it is easy to lose the orientation, part knowledge, etc. So, to facilitate reuse or remanufacturing it would be necessary to provide labels or marking on parts to keep track of them for future use or recycling.

But this is, like most of the topics discussed in this blog, an engineering problem - even a manufacturing engineering problem.  And, the role of humans in the reverse logistics loop must be considered. The flexibility of humans which was so important to “work around” in the forward loop to insure maximum automation might be just the ticket for the reverse logistics loop. But this would have to be teamed with designed tooling and systems to insure the reverse loop is just as productive as the forward loop.

So, if you get your pink slip from the assembly side because a robot stole your job take your application to the reverse logistics side … they should be hiring!