Engineering Perspectives: “Optimizing” My Vacation

Twenty years ago when I went on vacation I would simply call a travel agent tell them where I was thinking of going and how much I could spend. Then, like magic, a vacation is planned. Of course, travel agents are still an option today, but now with the advent of travel websites and Yelp reviews, I can not resist the urge to spend weeks on planning my vacations. It starts with the flights: Tuesdays are the best days to book (not sure if that’s a rumor or fact). Combination packages on the travel websites combining airfare, rental cars and hotel are for the novice as they never save money and often limit choices (not sure if that’s rumor or fact). Lastly, the excursions, oh the excursions, which one to go to, for where, how long or how adventurous they are. So many details — it can be overwhelming.

I have developed my own system, which I won’t elaborate on for fear of boring everyone to death. Suffice it to say that when I’m planning a vacation, I am knee-deep in optimization.

As we previously posted, Syska Hennessy defines optimization as an iterative process of selecting a solution with a measurably superior outcome. For more information read our first blog post on optimization.

I’ll use vacation planning to highlight three optimization “lessons learned” that we think are relevant to the AEC industry. Here goes:

Hotel Selection in NYC vs. Tristan da Cunha
Lesson #1: Optimization requires more than one solution.

Tourism is a big deal in New York City, and we could come up with many ways to optimize hotel selection here. For example, we could create an equation that measures two variables – the cost of a hotel and the convenience of its location, accounting for transportation costs. Only hotels that attain a certain combined score would be considered. Of course, there are lots of other ways to make a pick and, if we were very intentional, we could make the claim that our selection was optimized.

Now imagine a vacation to Tristan da Cunha, a remote group of volcanic islands in the South Atlantic Ocean. You can’t optimize hotel selection because there are no hotels in Tristan da Cunha. There are places to stay, but not more than one option per village. This destination also has no airport, so you have to take a six-day boat from South Africa. Thus, Tristan da Cunha is clearly not a good candidate for travel optimization (but you can read more about vacations there in this NYTimes article.)

Relating to this professional practice, project teams must get to a solution before they can claim they are undergoing a process of optimization. If there’s only one clear answer, you’re also not doing optimization.

Mechanical Subcontractors, Camping, and Frozen Drinks
Lesson #2: You can optimize in more than one way.

My ideal vacation is likely different from your ideal vacation. Some people would love nothing more than a resort vacation, sitting by a pool for a week, sipping a frozen drink and eating lovely food. Others love camping, which is pretty much the opposite. This underscores that you can measure value in different ways, but you’ll still get the best outcomes by using optimization.

This is important for AEC because choices can be as stark as the difference between a resort and camping. If mechanical subcontractors were allowed to optimize exclusively for their installation and material cost, the resulting building system would certainly have a higher operational cost. Similarly, if a design engineer optimized only for operational efficiency, the upfront cost would be astronomical. We can’t forget about the architect who wants our ceiling heights as high as possible or the facilities folks who want easy access to all the equipment. How cool would it be to provide the owner with a set of ratings so we could rank a select group of outcomes according to their specific needs and preferences? That would be optimization.

Tiger Woods, Lost Golf Balls, and Multifamily Residential
Lesson #3: Evaluation criteria can defy optimization

Let’s say I’ve decided to take a golf vacation on the west coast. I’ve had a lifelong dream of golfing at Pebble Beach on the Monterrey Peninsula. If I were to go there, my entire trip would have to be planned around my tee time reservation and I’d want my experience to make me feel like Tiger Woods with his 2010 US Open fairway shot next to the Pacific Ocean. For lots of reasons, though, I won’t be golfing only at Pebble. The major reason is it’s too restrictive: You have to reserve a two-night stay at one of the very expensive course hotels. The next major problem is I have a family that I would like to bring along and a place like San Diego has a lot more options that are family-friendly. The final problem — which might not be a problem — is that there are a number of other outstanding courses in the San Diego area, in particular in Torrey Pines.

Thus, my evaluation criteria for a dream golf vacation to Monterey is complicated because I’m trying to find the best outcome of a total experience. It isn’t worth the trouble of creating a value function for my entire dream golf trip because it’s a one-off experience, full of competing dissimilar cost types, and mandatory requirements, like hitting at least one fairway shot on 18 at Pebble Beach.

This is relevant to AEC because big project decisions are also really complicated. Given a parcel of land zoned for multi-family residential, there are literally billions of combinations of building shapes all resulting in different costs, project stakeholder experiences, and efficiencies. If you doubt this, read about the number of combinations in chess, then count the combinations shown in any video by Developers for these types of building must use their skills and instinct to make decisions, which would likely be impossible to codify into a single optimization method.

Bringing it all back, vacations offer a convenient analogy to illustrate the challenges of optimization because they balance value, cost, and experience. So we leave you with the mental image of planning a dream golf trip to Tristan da Cunha with Tiger Woods and a bunch of mechanical contractors who like resorts. (What can go wrong?)

Written By Robert Ioanna