In 2021 I worked on a surprisingly large number of construction jobs. This mostly came about because I was waiting on contracts or positions in my main line of work (tech) and they just never came in. I'm not a fan of sitting around doing nothing and I'd much rather be doing something productive with my time rather than nothing. So I ended up helping various other people with renovations and building work as something to do. This was great because I had some amazing mentoring from some people and learned a bunch of carpentry and general building skills. This has become exceptionally useful considering how absurdly hard it has been since the lockdown to get people in to work on things. Thanks Andy and Ross for sharing so much practical knowledge with me in 2021! I'm in the middle of writing an article about my recap of 2021 so I'll have some more details about all this in that post.
Something that got my attention when working on these jobs was the availability of engineered timber products like Laminated Veneer Lumber, often referred to by the acronym LVL. During the numerous days of lockdowns here I found myself watching a variety of videos about this stuff, for example this video about how LVLs are made or this video about how CLT is made.
Last year in particular I've was doing some research into various options with structural timber products. This is such a classic example of a situation where an industry might look a lot less interesting on the surface than it really is. In the last few years a number of engineered timber products have been created, recently for example I was doing some work where we were using LVL beams which has some really good structural properties while being quite easy to work with. While searching for something completely unrelated today I learned that engineered timber can be used to build huge spans like large roof sections or bridges. Take for example the use of Glulam for making arches strong enough to hold up the weight of a bridge:
This turned out to be an advantage for this remote bridge in Quebec, because the temperature conditions are extreme there is a lot of thermal expansion to deal with. The wooden structure was able to maintain a high strength without needing the same sort of heavy expansion joints if the structure were made of concrete. Speaking of expansion joints, earlier in the year I was doing some work with a company that installs expansion joints. We were looking at how different factors impacted the profitability of choices in building materials, this ended up being an interesting job in terms of the UI/UX for the data collection (if it's not easy to use something like this on a building site people won't use it1) and also made an interesting data science project.
Something I've noticed across these two projects is that there's a distinct lack of open data in this space. Given that construction is such a huge industry I know that this data exists somewhere but access is hard to come by. Because of technological progress there's now more materials choices than ever before but the data about these is hard to come by. Maybe you want to make a decision on using glulam vs LVL beams or choose between one expansion joint product and another, what's the best choice? Figuring this out isn't that easy, the cost per meter is probably the easiest thing to find and sometimes even pricing information is hard to come by. In the case of the timber/lumber industry you have to now contend with the fact that the prices of materials has changed dramatically in the last few months. You get a sense of this price volatility in raw materials when you look at the lumber futures lately:
Anyone who works professionally in construction knows that the amount of cost of the raw materials is only part of the decision, the installation costs and the impact this has on the schedule are things that you can only really know with some more information about the context in which the work is being done. Because prices haven't been stable on just about anything lately, I think there's even more value than before for a system that helps people to choose between construction products. Some automation here would be very valuable given the rapid changes happening across the market both in the prices of existing goods and the introduction of completely new product lines.
Choosing the best product between various alternatives all presumes that you know what the options are, product discovery is also frequently not straightforward. After hours of looking around and not finding all that much I was pleasantly surprised to find the open source wood initiative that's explicitly designed to make it easy for people to share information about wood construction designs. Considerations about the cost of building have resulted in a huge explosion of prefabricated materials being used in jobs, in part because there can be the opportunity to save considerable amounts of money by using prefabricated materials. The savings can come in many forms but often are the result of lower labor costs. You see an example of this with prefabricated roofing being used more often these days and stick framing roofs being less common.
Coming back to the situation with the beam I'm trying to source, the reason I'm looking into beams in the first place is that I have to construct a span which might will have a specific loading on the middle of it. I'd really like to know what the flexural strength properties of these products are so that I could make a better and safer choice of materials. But this data is exceedingly hard to find in any manner that's actually easy to work with and compare.
If I was a building materials supplier I'd likely invest in creating this sort of data capability, I think this would be a tremendously good investment because it would make it that much easier for customers engineers to get to know what materials they could use of the materials that are in stock and can be ordered in.
something I really appreciated with the construction industry is that if you make a truly shit application then people just won't use it. Something I've noticed in enterprise software, especially in a corporate office environment is just how much bullshit people will put up with in their software choices. The differences are almost certainly cultural differences between these groups of people and frankly I think a lot of people in the corporate space put up with far too many crappy software systems. ↩