Solving a Problem

We’ve got a particularly interesting problem on our desks here at Atlas Geotechnical. There’s a lot at risk, various stakeholders are frustrated with and suspicious of each other, and there’s not enough time. While working this problem through to a pretty tidy conclusion this afternoon, it occurred to me to share the process that we use to achieve a safe, efficient design.

It goes without saying that rigorous project framing is critical to any problem. Define the boundary limits and success factors. Write, refine, and document the basis of design. There’s no point in working really hard late into the night when you haven’t defined the problem you’re trying to solve.

Even when the project is framed and bounded correctly, the juiciest problems always offer sticking points; places where the natural tension between resources, budget, and performance simply don’t allow a path forward. When I get stuck at one of those obstacles, these are the techniques (in order) that I use to crack it:

  • Collect More Data: Usually when moving quickly through a conceptual design you adopt conservative and simplifying assumptions about important parameters. The best way to solve a problem is to collect real data and refine the assumed parameters. This is the most self-contained and linear problem solving technique.
  • Challenge Your Assumptions: Sometimes you’re limiting yourself. A classic is that soils are normally consolidated, when really there’s a desiccated crust and settlement will be less. The always-dependable Mohr-Coulomb constitutive model is another bountiful source of limiting assumptions embedded in our most useful analytical tools. Engineers in my office call this “doing it the hard way” but if it solves the problem, and nothing else would, how hard was it, really?
  • Push Back on External Constraints: This one is particularly effective here at Atlas, but you need to understand the discipline that you’re challenging along with the hopes and dreams (and fears) of the team member who imposed the limit. Someone tells you that you can’t drill through a pilecap? Can’t tolerate more than an inch of differential settlement? Can’t pump more than 150 gpm? Discover the simplifying assumptions embedded in that limit; perform Steps 1 and 2 on someone else’s work, and find a way to preserve project performance without complying with a simplistic limit.
  • Call a Friend: I can’t tell you the number of times that this one has saved my bacon. If I weren’t so proud it would be higher on my list. Clever engineers have been solving problems for millennia; one of my friends has, almost certainly, previously solved the problem that has puzzled me for an afternoon. This one can be humbling; try to be gracious. The corollary to this technique is “try to have clever friends.” I’m good friends with several old guys who’ve been everywhere, done everything, and shoots do they ever help me crack troublesome problems
  • Hold a Meeting: Just kidding. Meetings never solve problems.
  • Get Away from the Problem: Irv Olsen used to go see a movie; one of my best friends, an astonishingly effective engineer, hikes like a maniac; I thought up this post while swimming laps. You serve your clients best when you’re thinking creatively and clearly. Don’t stay at your desk putting on a show of hard work when really you should stretch your legs, clear your mind, and actually perform engineering. Sure, you’ll need to start again with Step 1 once you’ve blown the cobwebs out, but you already got down to this last step once without solving the problem, so what other choice do you have?

I’m considering distributing laminated cards to the younger engineers here at Atlas outlining these four steps. That or hardhat stickers.

While not a panacea, I’ve found that there are very few intractable problems when clever engineers, given a clear mandate through good project framing, apply themselves vigorously and enthusiastically.