Analysis and Presentation

I was fortunate enough to attend a one-day seminar in Philadelphia earlier this year put on by Edward Tufte, Professor Emeritus at Yale University. The $380 fee included four of Dr. Tufte’s books. I would estimate that about 1,000 people (!) attended the seminar the same day I did. I would recommend the seminar and the books to anyone who wants to communicate better or understand the dynamics of good communication.

Most of us attend meetings and presentations as part of our jobs. Many of us are asked at one time or another to present something – or to be consumers of other people’s presentations. What are the elements of an effective presentation? What follows are my interpretations of Dr. Tufte’s thoughts from the seminar and from his book Beautiful Evidence.

Your Job as Presenter

When you present, you have a story to tell. Don’t lose sight of your story. When you are finished, your audience should have no doubt about what your story was. Next comes the question, “Have I presented convincing evidence that my story should be believed?” As a presenter, you are responsible for the intellectual and moral integrity of your presentation. To be believed, you must “do your homework” and honestly communicate the findings. And as a listener, you are responsible for understanding what the story is and to evaluate its believability.

What is the best way to communicate information? Naturally, there is no easy answer. Dr. Tufte suggested: "Whatever it takes" – meaning, “Do whatever it takes to convey the story and don’t restrict yourself to one type of data or one style of presentation.” A simple approach is to simply do a ‘high-resolution data dump’ followed by a discussion of it. Said another way, “Put the information out, then ‘hold a press conference.’"  As an example, if the meeting is about improving sales, the meeting might begin with the distribution of some meaningful reports (homework!) and after a few minutes to time allotted to digest the information, a discussion is held on what the data means and what can be done. One more thing: People will be amazed if you finish early.

Notice how different this approach is from presenting a series of PowerPoint slides. The typical “PowerPoint style” of presentation has many faults, most notably that the presenter has dictatorial control over how the topic is covered, not only the content, but also the pace. As Dr. Tufte says, “(PowerPoint) presentations too often resemble a school play: very loud, very slow, and very simple.” Aggressive managers interrupt presentations like this, stop them, and almost always the meeting is better for it.

Onto some sound principles for telling your story:

Principles of Analysis and Presentation of Data

1.      Show comparisons, contrasts, differences – When we are analyzing any question, we always want to know, “Compared to what?” If we are going to use a visual tool to make our point, we should show it next to something we want to compare it to.

2.      Show causality, mechanism, explanation, systematic structure – As we are making our point, we also want to be teaching or explaining about “Why it is so.” For example, if we wanted to show sagging truss sales, we could also show them in light of housing starts, or mortgage rates, or whatever we think provides a possible connection.

3.      Show multivariate data; that is, show more than 1 or 2 variables – This is a real challenge because we have so few role models to emulate. A few examples – the baseball statistics in the Sports Section of your newspaper. There a dozens of statistics tracked for both teams and individuals and we each consume those numbers in our own way. Also, each reader will also draw their own conclusions from the same set of numbers. (That’s OK!) Another example is a road map, which shows routes, terrain, population, distance and may other forms of data. Maps, perfected over hundreds of years, show a wonderful variety of data and have almost no “fluff.” (You can tell fluff by asking, “If I left this out, would any meaning or understandability be lost?”)

4.      Completely integrate words, numbers, images, diagrams – This means that we place different types of information together, rather than segregating them. On a truss layout, we know the value of labeling everything right on the layout. Tables on the side of the layout are “second best” and should be avoided except to show summaries.

5.      Thoroughly describe the evidence. Provide a detailed title, indicate the authors and sponsors, document the data sources, show complete measurement scales, point out the relevant issues. This documents the evidence, and provides the basis for the intellectual and moral integrity of the presentation.

6.       Analytical presentations ultimately stand or fall depending on the quality, relevance, and integrity of their content. This suggests that the best way to improve a presentation is to improve the content – make it more substantial. Has the content of the presentation contributed to the questions it seeks to address?

Some examples of presenting multivariate data - completely unrelated to components: World History, How to Look at Modern Art, Napoleon’s Winter Campaign, More…

Seven Forms of Waste

In Component Talk #24 when discussing lean manufacturing, I said, “No one can afford to ignorant of its strategies and goals.” One of those strategies is the elimination of waste and the first step in that process is to identify what waste is. In putting this article together, I am quoting heavily from Lean for Dummies by Natalie Sayer and Bruce Williams, and excellent introduction to Lean.

Waste in lean manufacturing means wasted effort, wasted resources, and activities that add no value to the end user. Waste is both inevitable – it cannot be completely eliminated, and reducible – we can always improve. And so, on to the list:

Transport

The idea here is that any transportation of materials between “transformational operations” is a waste. Every time you move something it presents an opportunity for damage or injury, to say nothing of the time the transportation took. Poor layouts and disorganization are a frequent cause of transport waste. Conveyance mechanisms take up floor space, and can lead to inventory accumulation. If you’ve ever seen television program on how computer chips are made, you know that the reason that today’s chips are so much more efficient than yesterday’s is the fact that the new chips have reduced the space that the information must move through to almost nothing. When you make billions of production cycles per second, that adds up quickly. It’s no different on a component production floor.

Waiting

Any time a production person is waiting, it is a waste of that resource. Waiting most frequently comes from shortages, unbalanced work loads, need for instruction, or by design (such as watching a machine complete its cycle.) If you see a production person waiting around, ask yourself, “Which one of the reasons listed above is causing them to wait?” and address it. In some cases workers wait because they expect (or know) that whatever they are waiting for will come to them – even though they could simply go get it much faster. They ‘wait for it’ out of habit, and with full knowledge that eventually it will come to them.

Overproduction

Producing more than the customer needs is a waste and causes other wastes – inventory costs, manpower and conveyance cost to deal with the excess product, and consumption of raw materials.

Defect

In keeping with the all-too-obvious idea that we want to ‘do it right the first time,’ any process, product or service that fails to meet specifications is waste. I’ve seen the time taken to fix one missing plate on a truss shut down not just one line, but two (if they share a conveyor) equal to the time that it normally take to build three trusses. That’s a lot of waste, and missing plates are only one kind of defect. How many different types of defects do your products have? Which costs the company the most wasted time and resources? What are your mitigation strategies?

Inventory

Inventory anywhere is non-value-added to the consumer. It not only ties up capital, but it runs the risk of damage, obsolescence, spoilage, and reduced quality. Inventory takes up floor space and other resources to manage and track it. Keeping large inventories may be covering up other issues, such as equipment reliability or poor work practices.

Motion

Any movement of a person’s body that does not add value to the process is a waste. Examples are walking, bending, lifting twisting and reaching. Tacking a plate in place on a truss adds value, but turning around and having to reach to get the plate does not. Motion of the product can also be waste, such as building the truss flat, lifting it into the vertical position to stack or band, and lowering it back to flat for shipping.

Extra Processing

Any process that does not add value to the product is waste. In component manufacturing, this is frequently caused by inadequate technology, such as when we have to re-do an operation (like a finish press.) And given that we do so many things to add value to the products we sell, why not take the time to verify that those activities are indeed valuable? By talking to customers we can make sure that what we are doing is really adding value and/or making sure that the customer understands how to get the maximum value from the things we are doing.

And see? TPI-2007

Two new building codes are now in the pipeline

Within the next year many local and state jurisdictions will start to adopt and enforce the 2009 International Building Code (IBC 2009) and the 2009 International Residential Code (IRC 2009). These new codes reference, and thus require, the use of a new standard called ANSI/TPI 1-2007 (TPI-07). The most significant changes required by the new standard are in plating, but there are changes that affect lumber design as well.

Because of the new and modified plating checks, some plate sizes will increase, although this will be greatly influenced by the type of trusses you design. One estimate is that plating square inches will increase an average of 3% to 5%, and lumber costs will increase perhaps 1% over current design standards. Changes to individual truss designs may vary greatly.

Why should plates get bigger at all?

If you are an old truss designer like I am, you might assume that “everything there is to know about the analysis of trusses is known.” But that isn’t the case. Knowledge about plating and lumber design grows over time. As this happens, and the standards are reviewed, there are areas where certain assumptions of the past are replaced by the new knowledge gained. Is that ‘new knowledge’ “right,” and the previous knowledge (standards) wrong? Not necessarily – but both are based on ‘the best available knowledge at the time,’ and governed by the fact that each standard represents a consensus of those that participated in the creation of the standards. Each new standard therefore “tightens” areas that were based on assumptions, and in doing so frequently results on more stringent rules. I'm not sure it is widely understood that the development of new standards is not a “closed” process. In fact, plate companies, component manufacturers, academia, building officials, model code officials, consulting engineers and other groups in the US with an interest have participated.

What Type of Joints Will be Most Affected?

Because of more stringent checks in TPI-07, joints that plated previously (under the TPI-02 standards) may require a larger plate and in some cases may no longer plate. Those most affected are corner joints, splices, and peaks. Here are some suggestions to help:

·         A corner joint, such as a raised heel with vertical, may benefit from running the vertical through to the bearing or having a 2x6 (as opposed to a 2x4) end vertical.

·         A design option that analyzes the pitch breaks using a semi-rigid model will help a lot with some trusses. (In MiTek Engineering software, this is located in Design Info, Fixities.) This feature reduces the amount of moment at the perimeter break joints for which the program is designing. Using Semi-Rigid Joints will generally increase the stresses on the lumber while decreasing the moment on the plates. This will typically decrease the plating requirements, but it may increase (we think this will be minimal) the lumber size or grade.

·         With splices, an option to “Use solid bar tension values” (MiTek = Plating Options) will help. The downside to using this feature is that it requires stringent quality control. You must make sure the solid bar section of the plate (as opposed to the section through the plate with slots) is centered over the splice line. There is only an eighth of an in. (1/8-in.) positioning tolerance along the plate's length and the rotational tolerance is very limited (about 2 degrees). Although not recommend to use this feature as a rule, it may allow you to plate some joints that you could not otherwise.

·         Reduce the length of a panel that is next to the peak or hip joint. Doing so reduces the moment at the joint so that a smaller plate may be used.

·         Expanding your plate inventory to include more 18g high strength plates.

The use of these techniques may eliminate a large amount of the plating increases that we would see from TPI-07 using “out of the box” defaults.