Capturing interior spaces in digital form
Imagine your kitchen has been wrecked by fire, flood, or another disaster. How can your insurance company quickly and accurately assess what repairs are needed?
Today, an insurance adjuster must visit your site and painstakingly come up with an estimate—and each adjuster will figure things a bit differently. Software from Hosta Labs, a STEX25 startup, aims to dramatically speed and standardize this process.
With Hosta Labs’ Hopper AI structural assessment software, when you buy an insurance policy, you will be able to take a few pictures or videos of your damaged kitchen and upload them to your insurance firm. The software can then generate a detailed plan of each room, with floor plans, measurements and material takeoffs. When and if disaster strikes, you can grab an updated set of photos or videos, and the software will create an updated plan of the space that includes the required material takeoffs. Hopper AI will save you, the insurance customer, a lot of time and worry, and will save your insurance firm a lot of time and expense.
Hopper AI does not just understand objects separately, but understands them in a contextual relationship.
“Our Hopper AI solution provides an in-depth view from simple images or videos of interior spaces on risks, replacement values, and conditions, providing measurements, material takeoffs, object detection as well as technical drawings and CAD models,” says Henriette Fleischmann, Hosta Labs’ co-founder and chief operating officer.
As its name suggests, the software brings an artificial intelligence-powered platform to structurally assess interior spaces. “Hopper AI does not just understand objects separately, but understands them in a contextual relationship,” Fleischmann says. “It understands that a window belongs to a wall, and understands that while the wall can have a door opening, it should not have a crack or a hole in it.”
Hopper AI is based on patented technology invented by Rachelle Villalon, Hosta Labs co-founder and chief executive officer. Founded in 2020 and named after a hardy New England plant, the company has closed a seed round of financing. It is now finalizing a beta version of its software and negotiating agreements with major insurance companies.
While practicing as an architect and managing strategic initiatives at large architectural firms on the West Coast, Villalon looked unsuccessfully for a software solution to recreate interior spaces digitally. Deciding to build such a solution herself, she went to MIT to master the necessary computer vision technology frameworks. After completing her PhD in 2017, Villalon began putting together a startup to take on the problem of automating structural assessment. In 2019 she recruited Fleischmann, an experienced project manager with 12 years of relevant industry experience who was then completing her mid-career MBA at the Sloan School of Management.
While Hopper AI can be applied to many jobs in structurally assessing interior spaces, its market focus today is property insurance, where interior space represents a black box for both home and commercial insurance firms, Fleischmann says.
Hopper AI provides an in-depth view from simple images or videos to detect accurate replacement values and risk associated with the space, so that insurance can be priced appropriately and damage estimates can be generated more quickly and ultimately automatically. “The goal must be that the policyholder effortlessly feels safe about the coverage, and even more so if they experience damage,” Fleischmann says.
The software incorporates an application program interface (API) architecture designed to integrate smoothly with a company's existing software. “Our intention is to make it as simple as possible on all ends,” says Fleischmann. “Hopper AI is fast, effortless and reliable.”
“Our solution can be integrated wherever structural assessment of interior spaces is required, and the use cases for Hopper AI are plentiful,” Fleischmann says.
In home improvement, for example, customers can estimate the materials required to renovate a room from simple images. In robotics, again using only a camera, Hopper AI can provide a contextual understanding of the space so that a vacuum robot can navigate a room much more smoothly. In maintenance, the software can deliver the latest configuration of the built space. In fact, Hosta Labs won a startup competition with a global manufacturer in which Hopper AI converted images of robotic systems in factories into CAD models.
Hosta Labs aims to become the objective source of truth in the structural assessment of the built environment.
“That competition helped us further define our technology, and collaborating with large companies is an invaluable experience that you have to go through several times when you begin a startup,” Fleischmann notes.
Hosta Labs continues to draw on MIT's entrepreneurial ecosystem for crucial network opportunities, access to talent and intellectual alignment. “The STEX25 program has been a fantastic experience for us, increasing our exposure tremendously,” she says.
Based in Kendall Square, the startup has eight full-time employees and is hiring. “Hosta Labs aims to become the objective source of truth in the structural assessment of the built environment,” Fleischmann sums up. “We are already superior in terms of the output that we can create, so it's a matter of building out that competitive advantage.”