Adobe Stock / Mopic


Machine learning is making slow but sure inroads into the way building happens in the United States. Like many other opportunity areas, this subset of artificial intelligence that uses statistical techniques to give computers the ability to “learn” from data cuts two ways.

One executive who’s aggressively pushed at the front lines of technology’s role in the home building, design, and engineering process puts it this way.

“These programmable direct cutting machines work wonders in precision and accuracy and speed at a level humans could never approach,” he says. “But the machines won’t do what you don’t tell them to do. So, you have to start out with a level of accuracy in your architecture, engineering, and construction documents that doesn’t typically happen on projects, because typically a lot of those details get sorted out in the field on the building site.”

The good news is that most home building today takes place in blithe ignorance of what could be lost or gained were advanced materials procurement, cutting and assembly, and customer generating technologies brought to bear in the business as it currently exists.

However, that’s also the bad news.

Look at where the good people at McKinsey & Company rank construction in terms of industry sectors’ respective adoption and planned investment in artificial intelligence and robotics processes and solutions.

The McKinsey study has a note that echoes–on a more theoretical level–what our executive says about the double-edged sword of tech and data in building.

Of course, any AI algorithm is based on learning from the past. This means that AI needs a certain critical mass of data to deliver on its promise so scale will matter; as such, firms will need a significant amount of data (in this case projects) to train an AI algorithm.

McKinsey’s intimation here is that it’s only big companies that generated that amount of data, but it may be that the size of the organization is less important than the business and operational models in home construction themselves, at whatever the scale.

For if hard-and-fast assumptions on value generation in home building continue to orbit around builders’ ability to leverage relatively inexpensive land, inexpensive labor, or inexpensive capital to build homes and communities people want to buy and live in, then all the data in the world–or much less of it–will not steer builders into adoption and investment in technology.

The change coming is basic; it impacts the builder business model–how value and sustainable profitability get created in each home constructed–beyond the land arbitrage, beyond the materials sourcing arbitrage, and beyond the deltas in labor expenditure, and beyond the relative inexpensiveness of mortgage and finance.

It’s a classic remapping of houses–homes, really–as a consumer durable product.

That’s where real opportunity areas hide in plain sight today in home building, obscured by all the complexity, inefficiency, and counter-productive practices and processes so many people in building have lived with for so long that they believe they’re all necessary and fundamental to the process.

They’re not.

Homes–new, remodeled, rented, owned, resold–are a human need, a need that’s growing, not shrinking, and a need whose scope and scale are following a troubling trend of societal polarization, separating haves from will-never-haves.

Instead, if builders truly can look and learn from the tech community at the leaps and bounds they’re achieving in developing and evolving consumer-value-centric products and experiences, the mental and decision-making chain of events–the raw materials and data that pour into machine learning algorithms–can be simpler, more disciplined from the starting place, and less tolerant of wasted energy, time, money, and human talent.

This piece from Accenture managing directors Gary Hanifan and Kris Timmermans in the Harvard Business Review speaks specifically about “new skills”–human skills–that need to grow up around how artificial intelligence can favorably impact supply chains, an area of abiding interest among home builders of any size.

In a sense, though, Hanifan and Timmermans’ observations pertaining to where algorithms and human talent can blend to make the procurement puzzle make more sense, could apply to the entire value chain that leads to community and home development and construction.

In this new environment, both machines and humans are essential: By collaborating in roles such as supply chain planning and inventory management, the combined power of humans and machines will create new sources of value for businesses. We’ve explored the nature of the new value-enhancing roles that will emerge and identified three new categories of AI-driven jobs:

Trainers who help AI systems learn how to perform, which includes everything from helping natural language processors and language translators make fewer errors, to teaching AI algorithms how to mimic human behaviors.

Explainers who interpret the results of algorithms to improve transparency and accountability for AI decision making and processes.

Sustainers who ensure intelligent systems stay true to their original goals without crossing ethical lines or reinforcing bias.

AI, combined with advanced analytics, will enable supply chain planners to make more forward-looking, strategic decisions and spend less time on reactive problem solving.

Cut and paste “supply chain planners” in the passage above with “home building firms,” and we may be on to something.