The InsideMaps team, which is launching its app today on the Disrupt SF 2013 stage, is applying 3D mapping and capture technology to an interesting problem: How do you visualize furniture purchases in your home before you buy them?
According to InsideMaps, you do it by using your smartphone to scan the inside of your home, allowing them to create a 3D model in the cloud that you can then place virtual bits of furniture inside to see how it fits.
This may not seem like an earth shaker, and let’s be honest it is not. But as the son of an interior designer and fine finishes guy who worked in the trade every summer, I can appreciate it as a deceptively useful tool for some folks. Specifically, says CEO George Bolanos, there are a few segments that InsideMaps is aiming at: end-user customers, interior decorators and real estate agents.
Balanos is the founder of Wanadu, a web conferencing company that was acquired by Cisco’s Latitude. He went on to a role as Director of Technology at Sony Mobile. InsideMaps’ co-founder and CTO Jörgen Birkler comes from Sony Ericsson Silicon Valley where he served as Head of Applications & Services Development for the Xperia line.
The app launching today is a fairly simple affair. You fire it up, give it a few bits of information about your home and then begin scanning each room. You’re given the option to do a single or to zap the whole house at once and you’re turned loose to shoot what is essentially a series of panoramic shots of a room from the inside, culminating with a spin in the center. Those shots are then uploaded to the InsideMaps cloud for some special sauce application.
That consists of utilizing data captured by the gyroscope and magnetometer inside your device (which is why they have to be fairly new smartphones) to craft a model. That model is then checked against a series of visual touchstones in the images. The heights of doorknobs, window sills, etc. While they do this, they’re also gathering data about the average sizes of openings like doors and more, so they can apply machine learning to improve the quality of the models down the road.
Currently, the processing takes around 20 minutes for a room or 4 hours for a whole house. InsideMaps is using Amazon Compute Cloud to perform the model-making process, in part because of its instanced GPU setup, which comes in handy for jobs like these.
The app runs on Android for now, but the team is at work on an iOS version. As mentioned above, you need to have a device no more than around 2 years old to have the sensors needed to make the app do its magic.
Once the models are built they can be viewed in any browser using WebGL, with no additional plugins necessary. Users can access them via URL, which is sharable, or send ‘rendered’ 2D snapshots of a piece of furniture in place to friends.
The business of furniture
The monetary model of InsideMaps is refreshingly concise for an early stage startup. They’re currently working with large furniture distributors to cut deals that will provide a link on their site to the InsideMaps app. This will act as a funnel that allows users to scan their home and then utilize 3D models of the furniture of that maker to fit them into their homes.
I queried Bolanos about how they were going about creating the models of the furniture to use. It turns out that the team made a happy discovery once they started looking into this: most furniture makers already create 3D models of their wares. For ‘augmented reality’ or 3D rotation widgets, sure, but they had the models, and InsideMaps can tap into those without having to reinvent the wheel.
Once the pipeline is set up, InsideMaps is planning on utilizing a revenue setup that combines referral, affiliate and sponsorship models. This creates a ready-made meritocracy where results beget income. It’s a far more honest and straightforward system than a lot we tend to see these days.
Currently, InsideMaps is walking to big retailers including a deal with a subsidiary of the largest furniture retailer in the U.S.
But the direct sales market isn’t the only one that the team is excited about. Bolanos says that InsideMaps is equally intrigued with how the app will work with interior designers and real estate agents. They’ve already begun testing it with real estate agents actually, and it’s working out ‘well’. The idea is that an agent could gift a 3D model of a home to a new homeowner, or provide it in advance in the interest of ‘listing transparency’. In addition to helping people visualize the space in 3D, something a large number of people have a difficult time with, InsideMaps is also able to tap into the ‘new homeowner’, who is 10x more likely to buy furniture during their first 10 years of home ownership.
Eventually, InsideMaps hopes to get the processing down to about 2 minutes per room, with the hope that you’ll have that model processed before you’re even on to the next one.
The panel first asked if the app can do paint and wallpaper as well. Balanos replied that they can indeed, they just ran out of time.
The panel seems to be very interested in InsideMaps turning what they see as a tool into more of a service. A service which can then be used to build out a transactional model for people to have their homes scanned.