Photo credit: Flickr/ martinak15

As much as I hate to admit it, engineers are fallible (Gasp!). I know, I know, we spent years in school learning second order derivatives, every function possible in Excel, and have the Greek alphabet memorized. But still, it happens from time to time that we screw up as engineers.

Sometimes, our designs just plain miss the mark.

Today, I want to talk about five of the most common reasons that engineering designs fail. I’ve seen these mistakes time and time again in my career. If you can make an effort to avoid these mistakes, I promise that your abilities as an engineer will have improved dramatically. I can’t promise you’ll never mess up again, but you’ll certainly be moving in the right direction.

Why Engineering Designs Fail

Here are my five top reasons why engineering designs fail.

#1 Failure to understand the specifications

Engineers need to remind themselves that their work isn’t for themselves. Their work is for some client, be they an internal marketing department, as is the case in the auto industry, or someone external to the organization. That means that you need to sort out what the client actually wants before setting to work. All too often, I see engineers make assumptions (or wild-assed guesses, really) as to what the client wants for their product or service. This is engineering suicide. If it’s not crystal clear what the client expects of you or your design, stop what you’re doing, put down the CAD software, and step slowly away from the computer. That’s it, niiiiiice and easy. It’s nearly impossible to “guess” correctly. Do yourself a favour and get clarifications from the client.

If you’re client doesn’t know what they want, that’s OK too. Educate them. Get their buy-in for the ideas you come up with. Don’t assume that they’ll love your designs because they can’t think of anything themselves. You may be the expert, but clients tend not to sign cheques unless they get what they want/ expect.

#2  Scope creep

This issue is related to some extent to item #1. Scope creep is the inclusion of additional project scope over time. Normally, this is the result of either the client asking for more (or better, or faster, etc.) than what was originally specified. Sometimes, it’s the result of engineers trying to “perfect” their designs, and over deliver. Both can be problematic.

You see, there’s nothing inherently wrong with giving more than what you signed up for. In fact, that’s a great way to keep clients happy and build a name for yourself. The problem is that often these little bonuses you provide get out of hand. You lose track of all the extras you promised. And with each additional little feature or function, your design gets more complex, more difficult to integrate, and harder to design. It’s better to stick to the plan, and only over-deliver in cases where you are confident you won’t be sabotaging  the whole design or blowing the schedule and budget.

Engineers need to remember: Perfection is the enemy of getting things done. Just because you could make the design better than what has been requested doesn’t mean you should. Delivering exactly what you promised on time and on budget is a beautiful thing.

#3 Lack of knowledge

Sometimes, engineers take on projects that are beyond their capabilities. This can be because they want to prove themselves, they’re being pushed by their organization, or for any number of reasons. Regardless, the results are the same – poor design. Even worse, lack of knowledge can lead to dangerous designs. This is an absolute no-go. Where I practice in Ontario, Canada, the Engineering Act legislates that engineers are only allowed to practice in areas in which they are competent. We have a responsibility to the public to keep them safe over all else. My assumption is that most other jurisdictions in the world have similar legislation.

The take-home message here is simple: if you don’t know what you’re doing, don’t do it. You can’t possibly do a good job, and you could be putting people in harm’s way.

#4 No prototyping

This one baffles me a little bit. Engineers today tend to be unwilling/ unable to experiment. To me, this is the best part of the job. I learn more from playing and experimenting on the shop floor than anything else in my job. Engineers need to be comfortable stepping away from their desks, into their steel-toed boots, and trying things out.

I had a professor once in undergrad – my favourite – who used to hammer this home. He used to reject design submissions that didn’t indicate that the design team had actually built at least a rudimentary prototype as part of their design activities. He didn’t care how we had built the prototype, he just wanted to see that we had done it. The reason for this is that you can learn unexpected things about your design that wouldn’t necessarily come to you through calculations and modelling alone.

The company I work for recently commissioned a very fancy woodworking shop in one of our offices in Quebec, Canada. Seems a funny thing for a company that builds trains, doesn’t it? Not when you consider how much cheaper it is to build mock train parts with wood before actually going to production. Building wooden prototypes has been incredibly valuable for helping point out physical clearance issues and problems with cable and pipe routing – problems that would have been very expensive to fix after “real” train parts were being used in production.

#5 Poor intermediate reviews

In his excellent book, the Lean Startup, Eric Ries discusses the importance of building “minimum viable products”, or MVPs. His point is that if you build a product that is at least viable, you can take it to your client to see what they think of it. Then, you take that feedback, alter your design, build it up, and take it back to your client for more review. The idea is that if you get your client’s buy-in along the way, then you won’t waste your time building something just to find out they don’t like it at the end.

Many poor engineering designs come about because their intermediate reviews are poorly planned, poorly run, poorly attended, and/ or poorly documented. Intermediate reviews need to be strong in order for engineering designs to be strong.

Other Engineering Design Killers

What’s your experience? I’d love to hear about what you’ve seen lead to poor engineering designs in your practice. Tell the E&L community about it in the comments section below.