Software jobs in the construction world
So I’m at an inflection point in my career as a programmer and I’m wondering if there’s a way to combine my enthusiasm for construction/building science with my marketable programming skills.
What are some of the interesting areas or opportunities out there for software to make construction greener or just easier? I’m not really looking to start a new company, just want to find an area where I can be useful and get a bit closer to the industry.
Some ideas I’ve had:
1. Materials markets like supplyhouse.com
2. Revit or all the different modeling software
3. Buildzoom or all the other contractor markets
I’m hoping to find companies that produce products people like or make people’s jobs easier.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Modeling is big, and could probably use more optimization and design effort towards friendly user interfaces. Scheduling software can help the trades.
There is also automation. Automation can help with energy efficiency by more intelligently coordinating what runs when. An example would be to run the dryer AFTER the air conditioner finishes, but not at the SAME time when you'd show a bigger peak load to the utility. In the utility world, we call this kind of thing 'load leveling', and it involves coordinating loads so that they are staggered and don't all try to run at the same time. This is more of an embedded type of thing though, so small controllers and not large computers, but there is certainly a future for this type of smart automation.
Bill
Thanks Bill. I think the sort of "smart automation" work you're talking about could help mitigate some of the worse occupant behaviors that I've heard most builders bemoan. One of the biggest headaches in this area has been interoperability between the different vendors and the lack of a single standard. In theory the new standard "Matter" will allow google and amazon and all the other universes to play nicely together, but I guess the next step would be layering on the know-how sauce to help correct for bad behavior like you suggest.
I thought this was a very interesting article in the Economist: "Efficiency eludes the construction industry"
https://www.economist.com/business/2017/08/19/efficiency-eludes-the-construction-industry
Their premise is that while computerization has led to dramatic increases in productivity across the economy, productivity in the construction business is lower than it was 50 years ago.
So there's low-hanging fruit in bringing technology into the construction industry. I think the holy grail would be a platform where everyone involved with the project -- architect, engineers, general contractor, subcontractors, permit official, inspectors -- works on the same model (Revit or Revit-like) and each gets a view specific to their needs. Integrated into it is all the functionality they need to do their jobs -- scheduling, bidding, purchasing, billing, inspecting, payroll and accounts payable.
There would need to be firewalls between each party, so that one party can't see another's private information about costs and profitability. But everyone can see the schedule and status for the whole project. in order for that to work no one working on the project can own the whole project, there has to be an intermediary who plays that role. That's a natural role for a software service provider.
Deleted
Here's another idea, more of a software-hardware product: Local Positioning System (as compared to GPS, Global Position System). A system where you put out a set of beacons around a construction site and can use them to fix your location in three dimensions to within 1/16" of an inch. At the start of the project you carefully set out the beacons, and then for the rest of the project you use them as your primary measurement tool.
Once you have LPS on a site there are all kinds off automation opportunities. A very simple but very useful one would be a carrying robot -- carry this thing to this location.
There's a company already doing site layouts with a robot (like a roomba but a printer). Undoubtedly they are working on other tasks that capitalize on what they've already developed.
The issue I can see here is that construction projects are typicall one-offs, especially custom builds. Robots are great in places like auto assembly plants, because you have one design for a particular model vehicle, and then the plant churns out bazillions of the same thing. Construction is different, which makes it harder to automate since every project is different. You end up in a situation where setting up the automated labor saving devices takes as much, or maybe more, labor than it would take to just do the same work the old fashioned way, so you end up not saving anything. That's not to say there aren't ways to improve things, I'm just saying it's harder to do on custom projects compared to factories that make lots of identitical products.
The local positioning system would probably be most useful as a sort of survey aide. Framing would be easier if the foundation was EXACT, for example -- perfectly level and square. Things like that would save time further down the line, and the same goes for walls and the like.
Bill
That's basically the idea, get everyone working off of the same measuring system. Then get everyone working off of the same set of plans...
I was talking to a friend who's a mechanical engineer. He typically works on mid-sized municipal projects, where "mid-sized" means in the $20 million range. Think a small school or library. He draws up the HVAC system. The municipality insists on drawings that show every fitting and fastener, so for a section of ductwork he draws the duct, then a bracket to hold it to the wall and two screws to hold the bracket to the wall. (He actually just does a rough drawing and sends it to India where a low-paid technician puts in all of the details.)
Then the plans get approved and eventually they go to a mechanical contractor. And he does whatever he feels like that day when he's installing. Or more accurately, he uses the plan as a rough guide then does what he thinks best from there. In fairness to him, by the time the project gets to the stage where the HVAC is being installed, none of the preceding contractors have followed the plans exactly either. And the detailed list of fittings and fasteners? The contractor buys 20% more than he thinks he needs and throws out the rest. My friend tells me that on one of these jobs they don't pay for dumpster service, a salvage company pays them for their trash.
There's a lot of room for efficiency gains just from everyone working the same plan.
The detailed drawings are needed to get fair bids when the job goes out to bid, which is almost always the case with gov't projects (I've been involved in a lot of those). You have to put in all the details to make sure everyone bids the same thing, otherwise the contractors gouge you on change orders later (well, the drawings didn't show SCREWS in those brackets, so that's EXTRA!, never mind that the brackets won't do anything with a screw to fasten them to something...).
Contractors then don't usually want to actually add everything up, so they just estimate, then add a percentage, then hope they have enough. If the contractors are efficient, they will keep their leftovers for use on other projects. I do a lot of projects where my bid is for design AND installation, so I standardize all my materials, which means I can reuse leftovers on other projects more easily. We always use the same type of cable, connectors, etc. If a customer specifically requests a particular thing (like a specific color of connector, for example), then we order those for that particular job.
Due to the nature of construction, it's difficult to get things "perfect". I like to jokingly say "construction is where it's all f'ed up, and then you have a building when you're done, if you're lucky". Construction projects tend to be a little chaotic, and I agree there is probably a lot of efficiency that COULD be gained, but getting there is tricky. If you standardize on materials too much, you limit what you can build, but if you "build the dream", you're probably going to have more wasted materials. I usually try for a sort of balance as best as I can.
Maybe we should be thinking about better job estimating software? Better estimates of materials requirements would help to limit waste at the end of the job. Then you just have difficulty of getting the crews to actually follow the plan. You could probably cut down on drywall waste, for example, by using software to show all the optimum ways to cut the panels. The problem is I'm not sure anyone would bother to follow that in practice, and the usual rule is to cut to fit anyway, to avoid measurement errors.
Bill
I am also a programmer. I have been getting more and more interested in building science, construction methods, electricity usage, and indoor air quality.
If I was going to jump ship into this industry, it would be integrating IAQ sensors with HVAC and ventilation, electrical usage management, and maybe some other home automation. None of this would help with the construction process.
I just don't have the time to take that on at a hobby level right now, and I don't have the desire to switch occupations.
Yeah, I think my own personal goal would be something akin on fully instrumenting a bunch of different metrics for my house. I know they're not the best, but I've got my place stuffed with Airthings sensors and I use them to help gauge the impact of changes I make to the house. How much does that insulation actually help? How much humidity build-up is there if I were to seal the vents to the crawlspace?
The sort of integration you're talking about seems possible with something like homeassistant, but setting it up and maintaining the suite of equipment you'd need would be a part-time job, at least initially.
You're right that it wouldn't drive much in the construction industry, but I think there's an element of education and information that would help builders justify making the right decisions to their clients.
Deleted
As a retired software engineer, I would combine HVAC, IOT, Building Automation and home ecosystems. Embed an enocean sensor into a roof assemble and switch on the dehumidifier if you reach a dew point. Could create a very nice residential PLC with CoDeSys.
Decide if your interest is in Commercial or Residential Construction
- Commercial: There are already too many software "solutions" but few of them solve anything without also creating an equal number of new issues. The industry is eager for solutions. For success stories, look at Revit, PlanGrid, Fieldwire, CxAlloy, Textura, MultiVista, Ebuilder, ProCore and years ago before Oracle ruined them, Expedition and P6 (which are both now old, but they were terrific for their time). All of these current solutions have aspects that are terrific for certain parts of the process, but all of them have significant issues, including that none of them talk to each other (but they all claim they do).
If you want to work in the residential world, your challenge is simply getting people to use software. They won't. Or they claim to then don't. Or garbage in, garbage out. There are certainly exceptions, but it is a smaller market.
Thanks! Autodesk seems to be producing some of the most widely used (though not the most loved) products right now, so I might take a look at some of their remote positions. The fact that software can't interoperate is both completed expected and pretty grim.
I really wish there was someone attempting to address the residential space, because my own experience as a homeowner has been so poor. I know that some platforms are trying to unify the bidding experience, but I think it's not as profitable as they'd hoped and it's harder to generate the capital needed to cover all the complexity of the industry. It's almost like you'd want something that was so compelling and useful that people would accommodate the tools instead of expecting them to cover every aspect of their workflow.
Here are a few ideas.
I had a computer programming job at Argonne National Lab right after grad school working for https://www.anl.gov/profile/ralph-t-muehleisen on a building performance simulation project. You might also check out the National Renewable Energy Lab (NREL) and the Pacific Northwest National Lab (PNNL) for these types of jobs.
In addition to working directly for Autodesk, Revit Plug-ins are a big world.
Take a look at this open-source plug-in if you are into that side of things: https://github.com/eirannejad/pyRevit The primary developer, Ehsan, is very friendly. He's an active participant in the PDX Revit User Group: https://www.meetup.com/Portland-Revit-User-Group/?_cookie-check=1nQ0bQpjhdduB5-J
Ideate is a heavy-weight in the Revit plug-ins world, so that would be a good place to look to see if they are hiring: https://ideatesoftware.com/
https://www.cove.tools/ is an energy modelling service that is growing and I wouldn't be surprised if they were hiring.
If you're actually looking for a job and not looking to start a company, Brendan has good advice.
Find someone who's doing something you think is cool and see if they are hiring.
Yeah, this is the kind of advice I was hoping to find. Thanks guys! Energy modeling seems like a deep hole to fall down, so cove seems super interesting.
If you're interested in the startup world, there's a big list of them here: https://constructionphysics.substack.com/p/a-brief-history-of-construction-startups
I found my current job via Climatebase, which has categories for Buildings and Materials & Manufacturing: https://climatebase.org/jobs?l=&q=§ors=Buildings§ors=Materials+%26+Manufacturing&categories=Engineering%3A+Software&p=0&remote=false (actually, I found Climatebase via the Hempwood site, which I discovered here on GBA!).