AI we will be able to create more eco-friendly and healthy homes in the future
Your house will be outfitted with native landscapes and solar cells thanks to computer vision and GANs.
In the book SuperSight: ‘What Augmented Reality Means for Our Lives, Our Work, and the Way We Imagine the Future,’ author David Rose explores how augmented reality is already transforming various industries. Covering food service, medicine, education, construction, and architecture, he explores what the technology could accomplish in the not-too-distant future.
Everyone should use solar panels. Since the average cost of a sustainable energy system has dropped from $5.86/watt to $1.50/watt over the last decade, it’s a no-brainer from a financial standpoint. If you live in the sun-saturated South, you’ll save even more money in the first month of financing an installation because of the sun’s rays.
So what’s stopping us? You have no idea what you’re getting into! Math, logistics, taxation, and aesthetics all play a part in the equation. The Tin Man from the Wizard of Oz has inspired many homeowners to fear that their homes will become shiny and reflecting as a result. Learning to “speak solar” in new terms like kilowatt-hours is a part of the process of calculating out how many panels you need. Even if the risk is only perceived, it is still a risk.
Energy Sage, a Boston-based green business, is on a mission to encourage people to go electric in their homes. That includes solar panels on your roof, an electric car, a home battery system, automatic blinds, and a smart thermostat that pre-cools or pre-heats your home as you drive. As part of this effort, they’ve teamed up with us at Continuum to show potential customers what an electrified house might look like. For our clients, we use public Google Home satellite imagery to size solar panels, digitally overlay them on their rooftop and then present the results from both the street and their neighbor’s fence. As a result, we combine these photographs with data from Google’s Project Sunroof, which helps you estimate your roof’s solar savings potential. The decision to switch to an electric home is a simple one once you’ve seen the stunning images of your new home and realized how much money you’ll save over time.
A similar SuperSight-envisioned strategy will assist other home renovation projects. Looking at landscaping as an example, we’ll see that it’s a complex, possibly expensive, and risky endeavor that necessitates pre-project visualization.
When I first met landscape designer Julie Moir-Messervy at an MIT pitch competition, I was immediately drawn to her mission: empowering homeowners with the confidence and resources necessary to transform their backyards into outdoor living areas. Her startup, HomeOutside, uses AI and computer vision to help consumers see new possibilities for their backyards. With a compelling image of their yard in mind, the company helps clients make it happen by hiring an installation, delivering goods, as well as helping spread the payments over time.
Greenscapes not only increase the resale value of a home but also reduce asthma triggers, speed up recovery from illness, lower summer temperatures, and even prevent crime. The bees and birds benefit from proper native landscaping, which pollinates trees and reseeds vegetation. Trees in the southwest can minimize the demand for air-conditioning, while hedges in the northeast can reduce the cold weather’s wind and heating costs. Each tree may absorb a ton of carbon dioxide throughout its life, lowering runoff and erosion at the same time.
Because most people don’t know where to begin, Julie informed me, “most people do nothing in their yards.” There is a lack of knowledge about which plants to choose and how to arrange them, or how to build a landscape design and maintain it over time.” I took place on her board and began working on the problem.
An artificial neural network (GAN) developed by HomeOutside is being trained on the hundreds of designs (think of them as recipes) it has created for clients over the last two decades to create beautiful and sustainable landscape designs automatically. Starting with a 3D view of any address, the corporation employs Google Earth Engine and photogrammetry (US only, currently). Afterward, the GAN architecture uses a generator and a discriminator to create a new design and evaluate or score the work. Until the discriminator decides that the landscape has a good composition: shade trees, natural pollinators, grass for playing, hardscapes/decks and furnishings for gathering areas, plant diversity, and so on, these two networks will continue their iterative game.
“Imagination engine” technology is clearly of interest to companies that sell plants, furniture, lighting, and hardscapes since it bridges the conceptual gap between the existing condition of someone’s garden and what may be—and so motivates many more people. To make the dream real.” Outdoor enthusiasts and homeowners alike will benefit from this new trend—and so will the environment as well. This project’s environmental-minded investors, though, find the possibility to reshape entire neighborhoods the most exciting part. It’s possible that we could construct a new national park by connecting millions of backyards and creating a habitat for birds and bees in the process. Approximately 2.5 tons of carbon are absorbed per acre of forest each year. What if we made a considerable amount of carbon sequestration possible in our neighborhoods?
As part of HomeOutside’s great aim to proactively rebuild seventy million front yards, I worked with Julie and her team to develop a 3D redesign of their customers’ yards, which they subsequently emailed to their customers. You don’t even have to leave your house to experience the app’s rich animated landscape, thanks to its usage of physical world anchors. The edible garden’s location may be seen in a time-lapse video from dawn to dusk. This winter-themed illustration was created to understand better why they chose the neighbor’s yard over their own. This time of year is a veritable explosion of color.
Adding new shade trees and pollination shrubs to one’s yard is a novel idea, but will it frighten people? Google Street View has made it so that your front yard is no longer private. And if you’re trying to sell your house, you may not want to hire a landscaper and instead publish photographs of HomeOutside’s makeover version to boost your curb appeal.
Many diverse fields will benefit from this visioning technology once it becomes more widely available. For example, Home Depot recently invested in a startup called Hover, which, after digitizing your home in 3D, visualizes and costs new paint, siding, and roofing materials. Soon, SuperSight will show a real-life painter finishing the final touches to work, so you may feel the joy of seeing a job well done. With your new Volkswagen Passat, Volkswagen may put the kayaks and mountain bikes you enjoy on top. Who is the insurance agency attempting to sell you? Your new Passat will be blasted by hail, your solar panels will be blown off, and your shade tree will be struck by lightning, all in the course of their disaster scenario. Repaint first, then obtain insurance.
Is there a way for us to interact with these designs? Is it better to point and plant trees or paint flowers from a palette of options, like a 3D version of Photoshop, with our SuperSight glasses? What will we do if we don’t know what plants we like? Will we tell the system what we want to learn about our tastes, then present a single solution that we’ll love? “Expertly crafted” solutions, like those offered by an architect, interior designer, or wedding planner, are what we’ll prefer in the long run, I believe.
Over-specification is familiar among experts since they are so proficient in their field. Don’t tell an architect or interior designer, for example, that you want a window in this exact spot or that you want this particular chair in this color in this corner. As an alternative, you can convey your ideas at a higher level of abstraction (e.g., “I want the space to feel more connected to the environment”) or through the description of a required function (e.g., “We want a vegetable garden”), and let them handle the specifics.
Our interactions with SuperSight AIs will follow the same expert-guided model. As for the landscape, we may request a formal French garden with rectilinear patterns and exotic species or a curvy organic design that prioritizes privacy from our neighbors. For example, we could express our choice for an open area for play or a filled-in plan with more room for a fruitful garden. Our 3D landscape design will constantly recalculate as we convey these higher-level desires. When we wear SuperSight glasses, we’ll be able to test our hunches more quickly since we’ll be able to see reconfigurations in context, superimposed on our existing home.
There is no guarantee that HomeOutside’s technology can persuade millions of households to invest considerably in a sustainable environment. There has been a lot of positive feedback from consumers, who are thrilled to see their yards transformed. HomeOutside will use Google Earth and street view imagery in a generative AI tool to design tens of millions of landscapes, including sustainable flora, shade trees, natural pollinators, and bird-friendly berries in the next five years. As a result of its success, a million households will plant at least 3 million shade trees, such as oaks and beeches, that will each collect 48 pounds of carbon a year as they mature. These trees will have absorbed 14 billion tons of carbon dioxide throughout their lives.
“You are establishing the equivalent of a new national park—the National Park of us!” one of the HomeOutside experts said. With tools like HomeOutside, homeowners may be able to change the landscape of the United States.”
To help people envisage and imagine a better future for themselves and the planet is the ultimate potential power of SuperSight.