The industries that have the biggest environmental impact are routinely framed as agriculture, including meat and dairy, and fossil fuels, namely oil.
In January 2020, however, The Guardian reported that “the construction industry accounts for 60% of all materials used while creating a third of all waste and generating 45% of all CO2 emissions in the process.”
The implication is, if we are to avert the climate crisis, we need to change the way we build things.
Silicon Roundabout sat down with Kasia Borowska, Managing Director of Brainpool AI, London’s most compelling Artificial Intelligence (AI) consultancy, to talk about AI’s role in rebuilding the construction industry.
DAISY is Brainpool’s first product for the Construction Design industry. The name, not actually reflecting the tendency to gender AI with a female bias, is an acronym.
Design AI SYstem.
In short, DAISY revolutionises construction design processes that have largely remained unchanged for decades. Contemporary buildings, which house complex electrical and plumbing infrastructures, are, to this day, planned by hand. Thus, even the most ordinary of rooms and staircases that we all live and work in take 3–4 hours of high-value planning and design time.
“The construction industry has been one of the slowest adopters of AI technologies,” Kasia says, “but it will be completely transformed in the next five years because there are so many processes that can be done efficiently.’’
The DAISY algorithm, built with a construction industry partner, Staircraft, uses genetic programming to find the optimal floor plan design. For new residential homes, rather than hours (and often days) of human endeavour, DAISY calculates over 100,000 variations in 5–10 minutes and creates the optimal design. For Staircraft, the biggest stair provider in the UK, who provide a third of new British homes with timber stairs and floors, the DAISY prototype is projected to save them 5–10% reduction in cost per floor.
Brainpool Co-founder, Kasia, is moving to Canada with one half of her core team of eight.
“For the time being, my role has changed: I’m DAISY’s project manager and I’ve tasked myself with heading to North America to figure out how to turn DAISY into a product that everyone [in construction] needs.”
Kasia has pedigree in problem-solving long before she graduated in Cognitive and Decision Sciences (MSc) from UCL in 2011.
She was only 15 when she set up her first venture, a Latin and Ballroom dance school, in her hometown of Sopot, Poland.
“My parents gave me courage”, she remembered. “Without that, it’s hard to build a business, because there are so many risks involved. But my parents gave me the comfort to think that no matter what happens it will be ok.”
That’s why she’s heading to Toronto. In simple terms, there are more opportunities in the timber construction industry in North America than in the UK.
But won’t there be many more AI companies competing for business in the US?
She agrees, there will be. However, very few organisations have access to the kind of AI resources that Brainpool has.
As the name implies, they are a “pool” of 400 of the most gifted AI and Machine Learning data scientists, or “brains”, on the planet. Almost all of Brainpool’s experts are PhD level, coming from the most prestigious universities in the world such as Oxford, Cambridge, UCL, NYU, Harvard. What began in 2016 as effectively a recruitment platform to place data scientists into companies, has transformed into a leading consultancy to guide client AI builds.
This change, Kasia explains, was essential because there is too big a knowledge gap between data scientists and people-in-business. Essentially, the two disparate parties found it difficult to communicate because code and money are two very different languages. So, now, Brainpool sits happily in the middle of the marketplace, somewhere between the academics and the corporations.
If it sounds like they’ve hit the sweet spot and in just the right time, it’s because they probably have.
There are still plenty of challenges, however.
“Such as finding “The Diamonds”!” exclaims Kasia.
“Diamonds” is AI-industry slang for the very few people who are data scientists, and therefore have a full understanding of algorithmic and machine learning theory, but who are also data engineers and have exceptional coding skills.
She continues: “CTO [and Brainpool co-founder], Peter Bebbington, is one of them. We have three more in the core team but we always need more! They don’t have to be London or US-based. It’s remote working. One of the guys is currently based in Bangkok.”
The Future Is Green
In December 2019, Brainpool completed their third investment round, overhitting their £120,000 target by 168%.
That means they can progress with their long term vision to produce AI across the industry-mix.
“Over time, our ambition is to nurture an ecosystem that allows PhD-level researchers to find real-life business applications for their work”
In Kasia’s view, the current misapplication of PhD research is a societal-problem that needs fixing.
“There is so much high-quality AI research out there that never sees the light of day. Most university research ends up in some professor’s drawer with no way of solving the actual business and social problems that we face.”
So what industry will they look to specialise in?
“Any and all…”, she says, “…so long as there is an ecological advantage to using AI.”
DAISY, it should be noted, has an important ecological element. The sub-optimal design that currently plagues the construction industry incurs up to40% wastage of wooden joints and beams. The cheapest way to dispose of this wood is to burn it. DAISY, once deployed across the design industry, will minimise material wastage on a grand scale.
Perhaps it turns out that DAISY’s name isn’t merely an acronym after all, and more a reflection of the natural processes the algorithm is in tune with.