Wednesday, October 16, 2019

Will Automation Always be a Distant Dream for Modular Housing?

I recently heard, once again, a speaker comparing modular housing with automobile production lines.

Let me put this to rest for hopefully the last time. Modular housing is not anything like the auto industry. There are two very important reasons this is not a true comparison.

The first is the idea that houses can be produced on an automated assembly line with few production workers. Even if a factory automated the wall and floor panels and roof trusses, automation would end at that point in the factory.

From assembling those components into a volumetric module through electrical, plumbing, insulation, drywall and interior and exterior finishing, no robot or automated machinery will ever be able to take the place of human interaction with the module.

Having watched this agony video of a robot picking up a sheet of drywall, squaring it on a jig so it could place it on a wall panel and then screwing it to the studs, I can’t believe any factory would want to spend a million dollars or more to have drywall put up infinitely slower than any person that has ever worked that station on the production line.

And let’s not even try to figure out how long it would take a robot to measure, pick up a sheet of drywall, square it and then cut it to size before fastening it to the wall. My God, please put that robot out of its misery already!

And forget about Robbie applying tape and two coats of compound and final sanding. It would look like seagulls used it for their bathroom after just one wall.

The second reason we can’t compare the two production systems is as simple as looking at the ultimate purchaser of each one.

Ford builds Mustangs. They may have a hundred options including slight body line changes, engines, interiors and paint schemes but in the end there is really only a finite number of actions robots need to make to produce a Mustang. In 99% of those cars, the dealer ordered it and the end consumer simply bought it from the dealer’s lot.

A new house on the other hand is almost always drawn to the new home buyer’s tastes and desires. They move walls on paper faster than a toddler running away while you’re shopping. Choosing doors, windows, trim, countertops and kitchen and bath cabinetry are what make this the perfect home for the customer.

Literally thousands of standard and special order options are chosen for almost every new home going onto the modular housing production line. On top of that come Seismic codes for this state, snow load construction for that and wind speed calculations for yet another, all of which are built into homes on the same production line in the modular home factory.

Unlike the auto industry, we don't have a factory that builds only F150 pickups, another that builds Escapes and yet another that builds Ford tractors.

No, everything for the home is built on one assembly by human production workers that know how to read the print and how to make the cuts and add the additional materials needed to produce each of those unique homes.

Even BIM has a hard time keeping up with all those myriad of options, codes and regulations required for each individual customer. Two identical home floor plans, one for NY and one for VA built on the same production line will be built differently.
Modular housing is one of the last industries in the United States that will ever be fully automated like the auto industry.

As Mark Yost, the President of Skyline Champion, said last week in Pittsburgh, we have to create our own pathway.

That means not relying on people with little real time expertise in modular production trying to force automation on the industry.

Gary Fleisher is a housing veteran, editor/writer of Modular Home Builder blog and industry speaker/consultant.


Tom Hardiman said...

Were we at the same show? I heard the keynote speaker say that successful companies are the ones who have made incremental advances and continuous improvements over time vs investing heavily in automation and expecting quick results. I didn't hear anyone suggest that robots will soon be building homes.

That said, we are all complete idiots if we don't try to learn the best practices and what we can apply from other industries, including the auto industry. Certainly not the degree of automation. But lean manufacturing techniques, just in time inventory ordering, SOME degree of standardization, and reducing defects through a process of continuous improvement.

Ever heard of a company called Toyota. They also build a lot of modular homes in Japan. IT is not our enemy.

Anonymous said...

Has anyone seen the robots at the Giant/Giant Eagle grocery chain? They look for spills and report it back to the humans. It should end there and they are super creepy..Haha

Anonymous said...

You haven't put anything to rest with this blog. The only thing you've settled is the small debate as to whether or not you're willing to accept change. Spoiler: it seems that you're not.

If you cannot envision a future where Automated hoisting arms are lifting drywall sheets from their pile and delivering to a long wall jig to be screwed in place by some sort of robotic/automated delivery system, then I worry about the other opportunities you're not seeing in this industry. are we ever going to have full robots w/ arms and legs delivering and installing drywall in boxes? NO. But we will, as an industry, get to the point, probably out of necessity, where we automate the rough framing and heavy install portions of our manufacturing process (Read: the floor framing/decking, long and short wall, and roof/ceiling jigs).

For as much as technology costs, humans cost more. A factory could spend $100k on a robotic arm that moves drywall from one place to another. They could spend $10k per year servicing that equipment. That single piece of equipment, on a low estimate, could replace 3 factory employees. Let's assume an average pay rate of $18/hour, 40 hour work weeks, and one week paid vacation for each employee - that means the three are worth: $112,320 per year. That figure doesn't even take into account the burden costs of the employee like FICA, Futa, W/C, etc. Your investment of $110k in year one already paid for itself.

What factory owner, large corp or one of the few small family businesses left, is going to say no to paying off a capital improvement investment in year one?

You're zeroed in on the wrong part of this debate.

Anonymous said...

I am normally all for innovation, but technology for the sake of technology is useless. Automation in a plant sounds like a great idea and I am sure there are potential applications, but I don't think it gives a plant a significant advantage. The home building industry (stick and mod) is seemingly under constant criticism for lack of innovation. I think this industry is filled with risk takers. The very act of buying parcels of land and building homes for unknown buyers in very risky. I think that there just isn't any technology in existence that is better than current methods in serving the buyer's demands. If I am wrong, or if the technology is improving, I have high confidence that this industry will adopt it.

Steve Murphy said...

Rarely have I read a post that I so strongly agree with and vehemently disagree with at the same time!

As someone who cut his robotics teeth as a production engineer in the automotive industry in the 1980’s, I agree completely that the modular industry’s attempts at automation are ill conceived and way wide of the mark. This, in my view, is not because the automation is not up to it. The technology is not being used to solve the right problems.

While the offsite sector evolves, the flexibility and imaginative contribution of a human workforce is exactly what’s needed. We need to use tech to help them make that contribution, not to replace them.

So what are the problems we need to solve? What does mainstream construction get wrong?

They build poor quality products and they kill and injure too many of the workforce!

So we will use technology to keep our workforce safe (lifting devices as an example), to help improve quality (AI engines behind image processing for quality control) and to increase dependable throughput (CNC pre-cutting of boards and timbers. Multi-head drivers to speed up fixing).

We will automate. We absolutely must learn from automotive, who got a lot wrong before they ended up with what you see now.

Greg Thomsen said...

Incredibly well said! There’s still a craft and artistic aspect to assembling homes. While I do believe robotics and AI will catch up eventually, the ramp period in this industry is SUBSTANTIALLY larger than an automotive assembly line.

Daniel Grundy said...

Why would you spend that sort of money on a Robot like that for dry walling then give it a single driver to do one screw at a time?
The mind boggles, if you want a human form, a exoskeleton is the way to go. Enhance a human operative, give them a frame which allows for many simultaneous fixings and allows that operative to work solo quickly and safely for walls and ceilings.
If you want automation don't go a using bipedal design that copies human motions.
I've got concepts of at least four better robots for this job and two of those I've come up with while writing this.
Automation of the less skilled repetitive tasks will come for the factories, I've done enough time studies at swan to see how to use it but robots like this? No thanks it's an interesting example if what robotics can do but that's it.
For serious dedicated solutions designers need get down on a factory floor and find the 'non value add' opportunities and solve them.

Josh Margulies said...

Outstanding response to your article, Gary. You got the boys all riled up and thinking! Yummy!

Anyone who has seen homes built in field or factory for 40 or 50 years knows how imprecise we are in reality compared to an assembly line. BUT a younger generation of engineers does not have that long and myopic perspective!! I cannot imagine automating drywall finishing, house setting, or deck building, or masonry, or hvac, or plumbing as something that can be automated. BUT THEY CAN!

You see, coach old boy, sometimes an “elderly” perspective leaves us a bit rutted in our outlook. We are plainly limited by what we know up to now, the present. We cannot see the future because we don’t have as much of it left as do younger, more visionary people.

I cannot imagine!! But some of these young sounding, Jules Verne boys do!!

Talk more with them.

Anonymous said...

Change is inevitable and misery is optional. You might want to take a look at Ted Benson Homes operating as at Bensonwood in Keene, NH. in how he is changing the way homes are built.

Now this is not about Modular or site built, its about how to build better homes in a controlled environment in a more cost effective manor. Recently I read an article on this blog where someone commented about COST does not Equal Market for many Modular Builders.

Bensonwood is a panel plant using robotics and manual labor. Really need to look at the This Old House Episode S17 E 9 on this plant to understand the process better as it is not a typical panel plant or truss plant. Its a step up from the former Foremost Homes Plants in Greencastle PA.

I am not sure I understand or buy into the Champion/Skyline Presidents comments as that's merely doing the same thing and expecting a different result talk. I visited the Lillington, NC Champion Factory where they build both Manufactured Homes and Modular on the Same Assembly Line or at least that was what I was advised when I took the tour. So the speech doesn't match the reality of his operation if I was advised correctly. Skyline in PA is a different operation than NC factories.

Until we separate the old image of mobile homes from true modular nothing in consumers minds will change. As an appraiser I see fee appraisers that refuse to use site built homes as a comp to modular because they believe modular is inferior. Its a mindset that education and regs haven't changed.

Modular, Manufactured, Site, Panel, Kit, whatever; cost and quality don't equal market value. Building Officials don't measure Quality. Government regs, land cost and old thinking blame games might as well be a screen door on a submarine.

I don't claim to have answers given I am a government appraiser however before the modular industry becomes Sears a wake up call is required at a minimum and that I do agree with the Champion/Skyline Presidents speech.

Good luck guys, I still choose Modular over Site, the question to me is which modular company will overcome the cost/quality/image factors first to survive.

Keith Martin said...

Great comments Gary. There are far too many Management Consultants talking up automation and AI in an industry that they simply do not understand. Having said that, the industry needs to balance the realistic application of technology in modular offsite, not discard tech.

However, for this to even begin to gain traction we all need to see a revolutionary shift in procurement methodologies so that builders/manufacturers/operators can invest in the technology without being penalized for not being the cheapest price - we still reward cheap price over quality, reliability, sustainability, safety, technology and offsite modularization, with an absence of 'life-of-asset' evaluation.

Phil Woodman said...

I very much believe that automation of modular construction is out there, somewhere in the not too distant future. And, what we know already know about automated production of cars is not too dissimilar.

That said, it really comes down to the cookie cutter principle. If there was ever a time where a company would get an order to build 20,000, 50,000 or more houses, all the same dimensions, all the same specification, then spending millions on a fully automated factory would be worth the investment. However, I really can’t see this happening in the foreseeable future.

Therefore, the initial step change of getting house production from the job site to the factory was massive and still has very many critics in the industry. So the leap to everyone living in a new town or city, where the houses are all the same, is some way off. Automated manufacturing has it’s place in modular components, such as bathroom or kitchen pods where repetition is less of a problem and flexibility is not necessary.

Anonymous said...

Incredibly well said! There’s still a craft and artistic aspect to assembling homes. While I do believe robotics and AI will catch up eventually, the ramp period in this industry is SUBSTANTIALLY larger than an automotive assembly line.

The truly exciting part about the modular housing industry is its’ inherent agility by not having to have/wait on specific skilled labor (or a slower, million dollar robot) to get the job done!

Anonymous said...

And how may “experts” commenting have successfully delivered a multi level development from design inception to delivery.
Modular construction requires a defined process, but to compare it to automobile automation is a very long leap!

John Faller said...

I don’t really see automation in the modular industry, unless done by the suppliers. Consumers buying modular homes are tired of the cookie cutter products and are wanting customization, not only in available floor plans but also in the products themselves. The automotive industry is a process line with limited options. While there is quality control in the automotive industry, I strongly believe that the modular industry has much higher quality standards.

Anonymous said...

Trusses; wall panel;roof panels anything that can be standardized makes sense for the factory to consider. Final assembly will still require man hours both in the plant and in the field. One thing that could speed the process is builders looking more to personalizing not customizing three or four plans with alternate elevations. Small builders don't become middle size by doing everything custom.