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

Friday, January 29, 2010

Low Hanging Fruit - 4


Last part in a 4 part series

The "map" of spatial and temporal levels of design to manufacturing to distribution/enterprise effects we've been discussing forms the basis of identifying the information and potential actions to take over this complex space. I cannot go into all the possibilities in any detail.

The paper on which this discussion has been based ("Appropriate use of green manufacturing frameworks" authored by Corrine Reich-Weiser, Athulan Vijayaraghavan and myself) was submitted to the CIRP 2010 LCE Conference in Heifei PRC later this spring. A copy of the full paper is available - please send me an e-mail if you'd like a copy or, soon, we'll have it posted on the LMAS website. The paper goes on to review the leading life cycle analysis (LCA) methodologies (for example, process LCA, hybrid LCA and input-output LCA) and LCA frameworks and standards (for example, ISO 14040 and ISO 14044: 2006, US EPA : Life Cycle Engineering Guidelines: 2001, and NIST SLIM (SLIM stands for "Sustainable and Lifecycle Information-based Manufacturing"), and some green house gas (GHG) specific frameworks and standards (for example, PAS 2050, for publicly available standard 2050, and Corporate Reporting and Inventory Standards - Climate Registry and EPA Climate Leaders are good examples of reporting standards, while the WRI/WBCSD GHG Protocol and ISO 14064-1 are well-known inventory standards.)

We analyze how LCA and GHG methodologies apply at the various temporal and spatial levels we've been discussing. Not surprisingly, different methodologies and frameworks apply at different levels. The figures below, from the paper, show this variation. The colored dots indicate the degree of applicability - green indicates that the methodology applies well, yellow indicate a decent applicability, and red indicates poor applicability.



These figures suggest when each methodology is appropriate for each temporal and spatial level of manufacturing. Some are not applicable at all for some aspects of manufacturing.

The key differences to understand are that process LCA is most appropriate for detailed analysis of specific stages of an assessment or well-defined pieces of the manufacturing lifecycle. Hybrid assessment is best for two purposes: (1) ensuring a complete analysis across the boundaries of the analysis and (2) providing a screening to determine where process LCA is most effective.

Most existing standards are based on process LCA. In addition, existing GHG standards limit the scope of the analysis to only direct and electricity emissions, thus limiting the usefulness of the results. The exceptions to this are the PAS 2050 standards and the emerging WRI enterprise and full produce LCA guidelines. For the most part, the standards have focused on quantifying facility level emissions making extrapolation to the machine or supply chain level difficult. The figures demonstrate how these standards apply.

So, what does this all mean with respect to our discussion? Depending on the goal and scope of the assessment there is an appropriate tool available. However, all tools do not apply at all levels. Top-down hybrid LCA methodologies are effective at capturing full supply chain and enterprise level emissions; however process LCA approaches are most effective for tradeoffs at the factory or machine tool tool level of analysis. Most standards have focused on process LCA or limited enterprise LCA (just direct and electricity emissions). However, it is possible that this hole in existing standards will be filled by the emerging WRI standards on Scope 3 and product analysis.

So, back to our low hanging fruit. In addition to the methodologies identified above as applicable across the temporal and spatial scales of the manufacturing enterprise (some available for free on the internet, i.e the Carnegie-Mellon economic input-output LCA tools, see http://www.eiolca.net/) there are some more straightforward approaches.

One example addresses the concerns around sustainable packaging and the manufacture of packaging. Joe Greene, a professor of mechanical engineering at California State University - Chico has started, with industry support, the nonprofit Sustainable Green Products Inc. He starts with a "Sustainable Green Packaging Audit Checklist" which includes easily accessible information such as annual electricity and natural gas usage, car and air travel, amount of product used and recycled at the plant,waste generation and so on. Note - this is for plastic processing operations for packaging - not for everyone. But it's a start! Joe is working on a website but if you want more information contact him at jpgreene@csuchico.edu.

There are a number of similar efforts across many industries. We'll look into some of these in future postings. Whether or not they are complete or reliable depends on who's put them together. But they give you some initial data.

Next time we'll get a little less "academic" and talk about some interesting comparisons between comparable manufacturing technologies (different process paths to the same result) and how they stack up in terms of green manufacturing.

Thursday, January 21, 2010

Low Hanging Fruit - 3


Part 3 in a series

The hierarchies of manufacturing we've been discussing reflecting the different levels of "control" and "flexibility" one has from design to manufacturing have both temporal and organizational spans.  We were discussing the need to clearly identify the quality and quantity of information that passes through the interfaces between the levels because a natural result of information crossing interfaces it the potential for noise and inaccuracy. It is like the game we played as kids trying to whisper a phrase around a circle of other friends - the phrase coming out at the end was usually quite different than the one we started with!

This time we'll discuss the interaction between the four temporal and spatial levels described in the last two postings. A figure in the last post showed these four levels, from product design through process design and planning (manufacturing plan) to parameter selection and process optimization to post manufacturing operations (finishing, etc.) The flexibility to make decisions decreases as we move "lower" in the levels.

This makes sense. On the factory floor we are no longer able to change the product or component design, material or other features. We may not, at level 3, be able to do much about the suite of machines we intend to use to produce the part. We most likely can adjust some of the operating parameters or, at level 4, do some finishing or alteration to overcome a problem. The difference is somewhat like experiencing building and outfitting a house - from the architect-design stage to arranging the furniture in the finished house.

In the first posting we listed a number of "spatial" levels of manufacturing (from device to enterprise) and the levels discussed above are temporal levels - relating to different times. We can represent the interaction between these four temporal and spatial levels as in the figure below (and this is another one you'll probably have to click on to see clearly). The smaller arrows represent flow of information from one decision to another.



The figure represents, at differing spatial levels, the equivalent to the four temporal levels from above, the interactions and some of the details. This figure is from a paper submitted to the 2010 CIRP Life Cycle Engineering Conference in Heifei China later this spring (see http://lce2010.hfut.edu.cn/) and co-authored with C. Reich-Weiser and A. Vijayaraghavan.

As we move up and to the right in the figure we suffer a loss of decision making capability as all earlier decisions earlier in the product design cycle, or lower in the supply chain, effect the ability to make decisions at higher levels. So, at the enterprise spatial level, level 3 (logistics adjustments here) decisions are restricted to adjustments in supplier locations or distribution strategy rather than substantial changes. Similarly, at the machine design spatial level (equivalent to enterprise design but closer to the product), level 3 (machine manufacturing adjustments here) decisions are limited to such things as adjusting consumables or tooling.

How you address what is happening at any location within this matrix depends on what information you have about the process or system represented there, what your understanding is of what this information says about what's going on, what ability you have respond to this understanding, if needed (or leave it alone if it is performing correctly), what "levers and buttons" you have at your disposal to make a response and, finally, what means you have to determine if your response had any impact and, if so, how much.

So, back to our low hanging fruit. It seems obvious that the lowest hanging fruit is found at the lowest branches of the tree. So, in this representation, the low hangers are at the lowest level of flexibility. Changing the design or material of a product is not going to be low fruit. Changing machine operation to produce that item using less consumables (or less damaging consumables) or energy (change operation) may be.  In a metal cutting operation, changing tooling to increase machining efficiency is relatively straightforward. Adjusting tool path and cutting conditions, if on a computer controlled machine tool, is a bit more complicated but also reasonable. These are also low level fruit.

And, referring to the information you have and your understanding of it discussion above, choice of the appropriate methodology for conceptualizing and measuring environmental impacts is important.

We will pick up on this discussion next time. I need to keep this edition a bit shorter as I am still traveling in Europe at my conference.

And we'll also talk a bit about "smart grids" next time and what their impact might be on manufacturing. If you are not familiar with what smart grids are, your assignment is to google it and find out!


Friday, January 15, 2010

Low Hanging Fruit - 2


The term "low hanging fruit" is employed here to address things that can be done with out a lot of staff or resources and, specially, for smaller companies. This is of particular interest with respect to measuring or characterizing your scope 1-3 impacts. And, we agreed upon a definition of cost, or what's "too much for a small company," by starting with free and moving upwards. We continue the discussion started last week.

I was discussing this with one of my graduate students, Corinne Reich-Weiser, the other day and she commented that, actually, smaller companies may be at an advantage with respect to these calculations since in general the number of details for a product, financial interactions, suppliers, etc. might be smaller. She is working with a small company herself during her PhD studies, Climate Earth in San Francisco (http://www.climateearth.com/). They work a lot with companies of all sizes and have an approach that does enterprise and supply chain carbon accounting, specially the tricker Scope 3, based on the company's financial data and utility bills, etc. So, for this data smaller is better. (I have no stake in Climate Earth's business and only use this as an example of a situation when the requirements (data, cost, time, personnel) for assessment scale with the size of the business.)

In the previous posting, January 7th, I built the comments on material recently submitted to a life cycle engineering conference in China (and Corinne was one of the co-authors). Last time we discussed facility or spatial representations of frameworks for green manufacturing. We termed this part of the process to "find the tree" so we can look for the low hanging fruit. Now we go on to the temporal aspects of the life cycle assessment.

To do this, we start with the design of the product, and proceed through the design of the manufacturing process or system for the product, through to process optimization, and finally post-process control and abatement. I've been representing the "levels" of manufacturing in this way for some time. It is a convenient way to visualize the decreasing flexibility or choices (engineers might call these degrees of freedom) that occur as you move from the conceptual design to the concrete elements of a manufacturing process.

These levels are temporal in nature across the design-to-manufacturing lifecycle of the process, and can be applied in characterizing the degree of control over the environmental impact at each level. We've arbitrarily labeled the highest level as Level 1. It is the earliest in design and manufacturing - the "clean sheet of paper" stage all engineers dream about doing when they are in engineering school! At this stage all future decisions to be made in subsequent (and lower flexibility) levels 2-4 can be influenced. At Level 1 process design is integrated with part design and there is the most control over considerations of part precision, environmental impact, and manufacturing scale. Here there is scope to design the product as well as its manufacturing process to satisfy specific requirements in all the criteria.

At Level 2 fundamental process design and planning is performed for a fixed part design, and this drives the part precision, environmental impact, and manufacturing scale. Here there is extensive control over the performance of the process in all the criteria as allowed by the process design and planning.

At Level 3, process parameter selection and optimization is used to control the part-precision and the process environmental impact; control over the process scale at this level is limited by the flexibility possible with process planning and optimization.

Finally, at Level 4 post-process finishing and abatement processes are used in controlling the part-precision and the environmental impact; at this level there is no control over the process or product or system as it has already been designed.

A graphic visualization of these levels is shown below (click on the image for more resolution.) You can imagine how this basic structure can be mapped onto pretty much any product or process or system.



From these hierarchies – which span temporal and organizational spans – we get a sense of the complexity involved in information capture and transfer in manufacturing systems, especially what is required to support effective environmental analysis.

We need to clearly identify the quality and quantity of information that passes through the interfaces between the levels. As information crosses the interfaces, the potential for noise/inaccuracies dramatically increases.

Next time we'll discuss the interaction between the four temporal and spatial levels described last time and just above. We'll see how we can experience a loss of decision making flexibility as decisions earlier in the product design cycle or lower in the supply chain effect the ability to make decisions at higher levels.

Finally, I am writing this from Europe where I am participating in a production engineering academy meeting. That means I had a long plane ride with a lot of time to catch up on some reading - mostly Fortune and Economist magazines. Fortune had an article on getting a green job and pointed to community colleges as the place to go (see Getting a Green Job in Two Years, Mina Kimes in Fortune, November 23, 2009). Community colleges tend to be faster to respond to these growing markets for technical training and the article points to the example of Johnson Controls in Milwaukee partnering with the Milwaukee Area Technical College on a program to train solar installation designers and installers. Johnson Controls is installing a 2,500 panel "solar education farm" for this collaboration (and get some power out of it too!). (For details see http://www.johnsoncontrols.com/publish/us/en/products/building_efficiency/smart_environments/december-2009/urban-solar-farm.html)

In the December 7th, 2009 Fortune Marc Gunther writes about Best Buy's aggressive program to take your electronic waste back at their stores and recycle it. Besides getting customers in the stores the article quotes Best Buy's Senior Director of Corporate Responsibility as expecting this to be a break-even proposition depending on commodity prices. Since many states and some cities require electronic manufacturers to help finance recycling, the economics are tricky but breakeven, or even profit, is possible.

But there's more! And this is really interesting. Best Buy is looking at how to give products "a second life." They are partnering with a company in Irvine California (DealTree; see http://www.dealtree.com/) that helps manage trade-in and auction, processing of used items - the reverse supply chain we saw in the Ricoh comet chart. Customers can get credit in the form of gift cards for "gently used" electronics. They are also toying with the idea of the customer leasing an electronic product, for example, by guaranteeing a trade in value after some period of time.

We spoke of this concept some time ago and the real benefit being that companies design products differently if they are responsible for them (and must take them back and recover/recycle the material). As much thought then goes into taking them apart as assembling them in the first place. That means that the "lower levels" of manufacturing play a more prominent role at level 1 since they strongly affect this recovery/recycling and, importantly, any resulting profit.

And, the Fortune article states, this would remove the penalty of trading up in terms of technology every time a new electronic gadget comes along!


Friday, January 8, 2010

Low Hanging Fruit


One of the readers of this blog posed a very interesting question following the December 25th posting and the discussion there about making sure all the scope impacts (1-3) are included in an analysis of a process, system or product footprint. The question was, essentially, what can you do if you are a smaller company than Walmart to achieve any sort of measurement on the indirect stages? The follow on was that the financial and human resources necessary to accomplish this would be huge (i.e. too much) for a small company.

I commented that this was a great question and really gets to the heart of the issue ... what can be done that doesn't require a lot of resources but is effective? Then I referred to this the "low hanging fruit" strategy. I did not mention that, according to Fortune magazine and their Fortune 500 listing, any company other than Exxon is "smaller than Walmart" (see http://money.cnn.com/magazines/fortune/fortune500/2009/full_list/ if you want to check this out). But, I believe the question was intended to cover companies in the small to medium size category! So, let's go with that. And, we could have a long discussion about what, in terms of resources is "too much for a small company" but let's start out with free and move upwards.

In my response I mentioned that some "low hanging fruit" ideas include the resources on the Carnegie Mellon LCA website (http://www.eiolca.net/) that allow a quick look at aspects of your business and calculates a footprint (rough but helpful),  a simple questionnaire to major suppliers, resource providers to see if they are aware of their impacts (embedded energy, resource use, etc.), charting  "where things come from and where they go" (and this could be a useful group activity) to get a sense of the complexity of your production, and, finally (but not free), some of the resources listed in earlier blogs (like the lean green work or economic bottom up LCA tools) can be used effectively. There are also many groups forming in various regions that are struggling with this same problem and they try to network and share approaches to this, and other, issues.

I promised to work on this and use it as the content of a future blog so and go into more detail. It's 2010 and the future is now so I'd like to start the conversation off today and continue it over the next few postings.

By way of the discussion I will be extracting some material from a paper that two of my graduate students and I just prepared for submission to a life cycle engineering conference in China later this year. The paper was on "Appropriate use of green manufacturing frameworks" and was co-authored with Corrine Reich-Weiser and Athulan Vijayaraghavan (Corrine is a current PhD student and Athulan just completed his PhD - both are the kind of students that make being a professor fun and rewarding!).

In this paper, we started out with the comment that the question usually put forward is ‘where to begin?’ This is similar to the set up for this posting and is part of our strategy to address the low hanging fruit. But, first, we need to find the tree!

One of the challenges in assessing the environmental sustainability of a manufacturing process (or system) is the need to parse the process or system in a way that makes it appropriate for application of some kind of analysis - that is "find the tree". When we find the tree (or actually trees in this case) we can then determine our reach and see what is, in that context, "low hanging." And, I promise, this is as far as I will push this analogy!

This is challenging because it’s complicated. Sorting out when and how to use various analysis tools makes it easier to begin. This usually involves determining reasonable size elements of the problem (bite sized chunks so to speak) based on process or system complexity and the level at which the process resides in the design to manufacturing space.

A number of questions arise that must be answered, such as:

- How do we find the optimal balance between productivity, cost, quality and sustainability?
- What performance characteristics do you track and how do they relate to each other?
- In your analysis, what metrics, LCA, decision-making tools can scale over multiple levels?
- What decisions made at one level are not tracked at other levels?

Some of this came up in our discussion about lean and green. We will begin this discussion here and continue it over the next few blogs. First we'll take a shot a defining what our field of view is (sort of the google earth view of manufacturing we had some time ago - see Sept. 15, 2009, http://green-manufacturing.blogspot.com/2009/09/green-manufacturing-technology-wedges.html). I also appreciate that this discussion will go well beyond the original focus of the question raised but, hey, I'm an academic and there are no short answers!

The complexity and sophistication in the organization of manufacturing systems and processes, large or small,  requires a keen understanding of the organization for accurate environmental analysis. To assist in this effort, manufacturing can be broken into “levels of study” across two orthogonal frameworks, spanning organizational and temporal levels. From the perspective of the organization of the system, we can consider manufacturing processes as being composed of four levels, from the level of the individual devices where unit processes take place, through to that of the enterprise, incorporating all the activities in the manufacturing system, including supply chain externalities.

These four levels are as follows:
1- Device – Individual device in the manufacturing system, which is performing a unit process. Support equipment for the unit process are included here such as gage systems, device level oil-circulating systems etc.,

2- Line/Cell – Logical organization of devices in the system that is acting in series or parallel to execute a specific activity (such as manufacturing a part or assembly). Support equipment for the collection of devices are included here, such as chip conveyers, tool cribs, etc.,

3- Facility – Distinct physical entity housing multiple devices, which may or may not be logically organized into lines, cells, etc., Support equipment required at the facility level are also included here, such as power generators, water purifiers, HVAC systems, etc.,

4- Enterprise – The entire manufacturing enterprise, consisting of all the individual facilities, the infrastructure required to support the facilities, as well as the transportation and supply chain externalities.

Given these levels, which are essentially, facility or spatial representations, we'll go on to the temporal aspects of the life cycle - next time.