Good ideas with bad execution, or good execution of what should be bad ideas - an analysis of inferior, off-beat or malfunctioning products, and how other people's failures can help us design better stuff.

Software Angst Has Engineers Still Reaching for Their Trusty No. 2 Pencils

In a survey last year of 2,000 techies by the software maker Maplesoft, the trusty No. 2 pencil trumped all other products as the most freqently used design and analysis tool. w25655f.jpg Given the portability of paper-and-pencil, the pervasiveness of the quick scribble on the back of an envelope may not seem all the surprising. But the question is why aren’t more engineers embracing software tools like Maple 11, a symbolic math software package that maker Maplesoft says allows engineers to perform calculations and capture technical knowledge in an intuitive way? We asked engineers to explain the enduring popularity of the pencil – and their answers might well be a software marketer’s worst nightmare.

When Maplesoft President Jim Cooper reviewed the results of a recent survey his math software company conducted to find out what calculation tools engineers use, he was floored. “We found out that our biggest competition isn’t a rival software company, it’s the old classic: pencil and paper. To a great extent engineers today are doing math calculations the same way I did them 20 years ago–looking up constants in printed reference materials and doing hand calculations. We didn’t expect that,” the former aerospace engineer told me when I first reported on this story in an editorial in Design News. The Maplesoft survey was conducted in May 2006 across 2,000 scientists, engineers, and researchers working in a wide range of end markets, from aerospace to telecommunications. Not only did the venerable “hand calculation” rank number one as the most frequently used design and analysis tool–some 52 percent of the respondents reported that they scribble equations on the back-of-an-envelope daily. It was also far-and-away the most pervasive one. Just over one third of the respondents reported that they fire up Excel at least once a day, while another third use a specialized math software program like Maplesoft’s Maple 11or Wolfram Research’s Mathematica. Even more disconcerting–to those in the software business, anyway–is that Cooper could not find a shred of evidence that suggested younger engineers are embracing these software tools in any greater numbers than their gray-haired counterparts. “We’re in the 21st Century, you’d think that engineers in the first decade of their careers would be fully digital and taking advantage of the newest tools,” he says. “But that isn’t the case.” Given the portability of the handheld calculator and paper-and-pencil, the ubiquity of these tools may not seem all that surprising. But the question is why aren’t more engineers clammoring for tools like Maple 11, a symbolic math software package that Maplesoft says allows engineers to perform calculations and capture technical knowledge more easily and in a way that they intuitively expect it to look? Engineers say that in large part it is the high cost and steep learning curve associated with many digital tools that keep them pulling out their No. 2 and a pad of paper. “We don’t own any math analysis software,” says David O’Gwynn, a principal engineer at HME, a company that designs wireless communication systems. “While there are times that it would be useful, the fundamental issue with much of the engineering software out there is that the vendors set such absolutely outrageous prices, that there is no way for me, or any engineer I know, to be able to justify the ROI.” Instead, O’Gwynn’s team does hand math, writes simple software utilities of their own, or uses calculators capable of symbolic solving. “I personally use an HP 49g+ and an old HP 28S,” says O’Gwynn. “On the other hand, if a good math program were available for under $100, I wouldn’t hesitate to use it.” Two of the leading symbolic math packages sell for considerably more: Maple 11 retails for approximately $1,895 while the price tag on Mathematica 5.2 is around $1,880. Macsyma, an open source, general purpose symbolic-numerical-grahical mathmatics software package is available for $500. The thing that really galls Engineer Barrett Clay’s is the high cost of program upgrades. “It was a struggle to cover the initial cost of our math software,” he says. “But every year, our current version is supplanted by a newer, costly upgrade with no apparent improvement in functionality.” In search of a better deal, many engineers say they troll the Internet at sites like eBay, looking for older (and presumably legitimate) versions of math software at a substantially reduced price. “It’s like paying for the upgrade myself. And at least I get to keep the software when I switch companies,” confided one engineer, who wished to remain anonymous. (Note: when I checked eBay, I came across several auctions that looked to be on the shady side: An unopened version of the latest version of Maple’s software, for example, was offered for $129.99.) The cash outlay aside, another reason so many engineers still rely on manual methods is the claim there’s simply no time to learn anything new. Unless there is a really compelling reason to do so, most busy engineers say they are too frazzled to even bother these days and just go with the status quo. Robert Aschliman, a quality engineer at Stanley Security Solutions, is a software marketer’s worst nightmare: “Most of us are just too busy to learn all the new stuff that flies in from all directions,” he says, dismissing the very idea of having to learn a new input method for inserting equations. “I believe that most of us are comfortable with our tried-and-true methods and don’t see any reason to change.” “I’ve seen reluctance even in younger engineers to learn a new tool or change the tools they already know. After all, there is a limit to how many programs any one person can learn, I can barely fit another thing into my brain,” agrees Engineer Karl Hanson, whose tone on the phone suggested his cranium was indeed overstuffed for the day. “For my needs, I prefer writing programs in Visual Basic.” Even getting the software for free and wanting to learn it might not be powerful enough motivation for some engineers to tackle the learning curve. Jon Titus, a former engineer, gadget enthusiast, and technical writer, says he recently received a review copy of the newest version of a math analysis tool. Still in its shrink wrap, he says it’s been sitting on his shelf for months: “I know it would be a great piece of software to learn, but I just haven’t found the time,” he explained sheepishly. But perhaps the biggest hurdle of all for software developers is the degree of separation from the actual process that many engineers say they feel. “I have found that the ‘digital tool revolution’ has caused an unintended problem – brain disconnect,” says Tom Ehresman, a project manager at Firewall Forward Technologies, LLC. “Years ago I was using spreadsheets, math software, and other digital tools to perform and keep records of my calculations, but I then discovered a disturbing trend. In some instances where it was more difficult to estimate “ball park” figures, I found that after awhile I would blindly take whatever the computer spit out without taking the time to verify the results.” The more he used the tools, the further Ehresman felt his brain was getting from the process. I noticed mistakes in calculations, some small, some potential disasters. I even once nearly launched a project based upon a bad formula. I fortunately discovered the calculation was flawed before disaster struck.” Since that scare, Ehresman has gone back to manual calculations to start and verify computations and analysis. “I’ve even gone so far as (gasp) pulling out my slide rules and practicing and doing initial calculations in that manner and also going through my old college math books.” Now, he says that he is much quicker in problem solving and more confident and accurate in his analysis. “With the manual method, my mind is engaged before the process begins.” Chase Engvall, an RF & digital engineer for the Army, has used math programs like Matlab, Mathcad, and Mathematica profusely for many years but is convinced that they leave out an important side effect–intuitive understanding. “After putting in the data and seeing the answer that pops out at the end, I am always left with a sort of empty feeling that I didn’t really do it myself and am missing the details of how the answer was derived.” To illustrate his point, Engvall describes a tragic discovery he made about a canned RF calculator program used by his engineering department. “One of our technicians discovered that no matter what value he entered into a field labeled “Observer distance,” the output results didn’t change. So I requested a copy of the software, and guess what I found? That particular data item was not even utilized in the calculations! It really wasn’t an error in the calculations, which I suppose we might have discovered, but there were many wasted man-hours spent over the years deriving and inputting this data.” Some engineers simply describe the greater sense of accomplishment they get out of taking pencil to paper. “I actually enjoy doing the brainwork required for most of the simpler math,” says John Wyatt, an engineer based in Raleigh, NC. “It keeps my ‘co-processor’ sharpened, plus the act of holding the pencil just feels more grounded than hitting keys and clicking buttons.” “For me, there is a sense of satisfaction that comes from knowing that I figured out a problem the old fashioned way, without the “help” of a software package. It allows me to think logically and accept the notion that I may be worth my pay,” says Quality Engineer Robert Aschliman. Though it may take time, it’s likely that math analysis tools will evolve and become more accepted in the general engineering population. Even the venerable No. 2 isn’t exempt from progress. “I frequently pull out my trusty Pentel 0.5-mm automatic pencil and Machinery Handbook to do a few quick calculations,” says Aschliman. He explains that he has given up the wooden pencil mainly because most are made in Asia and don’t possess the quality of their predecessors. Plus it’s getting tougher all the time to find a sharpener! View the full results of Maplesoft’s 2006 Engineering and Scientific Calculation Tool Survey.

Tags: analysis tool, pervasiveness, rank number, reference materials, technical knowledge

Related posts


  1. Karen Field
    September 06, 2007 22:39

    Thanks for the update on Macsyma — for those who dare!!

  2. September 06, 2007 19:14

    I haven’t seen the commercial version of Macsyma for a while, I don’t think it’s still being maintained.
    But I do use the free, open sourced version occasionally:
    It IS a mess compared with the commercial packages, since it was their progenitor. Not well organized at all, but it’s useful for some things.

  3. Richard
    August 02, 2007 08:27

    I design software and often use pencil and paper, or whiteboard if in a group and I need to share. Similarly for us the design software costs thousands of pounds and because it can only convey a fixed set of concepts it can seriously hinder expressiveness or force you to say something in a very round-about way.
    The plus side of the software is the management features, but few companies use this. We are using our own addons to open source design software later in the design phase to auto-generate large parts of the system, but I think that places us as a minority. Microsoft promise basic auto-generation from design in some future version of their software.