I am proud to announce that OpenSpace has partnered with some of the world’s best and biggest real estate and construction companies, as we close our Series A. Proud, to be sure, but also grateful. First and foremost, I want to thank everyone who decided to take a risk and join the team at OpenSpace. You are the company, quite literally. There is nothing without you. You are why we are here, and why I can write this post.
Thank you to the team at Lux Capital, especially Bilal Zuberi, for welcoming me as Lux’s first Entrepreneur-in-Residence, and for giving Michael, Philip and me the freedom to explore. You’ve supported us since Day 1, and we’re proud to have you as the lead investor in this round.
Thank you to our new investor partners who joined Lux in the Series A, almost all of whom started off as customers: JLL Spark, Navitas Capital, Suffolk Construction, Tishman Speyer, WeWork and Zigg Capital. OpenSpace exists to simplify the lives of your workforces, and your support means a lot. Thanks also to our seed investors who helped launch us on this journey: our angel investors, Lux Capital, Foundation Capital, Sterling, Baidu, Box, and the National Science Foundation.
How did we get here? In the case of OpenSpace, our story isn’t that complicated. There was a literal “a-ha” moment — no joke. I can simply tell you exactly what happened, where we are today, and at least hint at where we are going.
OpenSpace brings together a few observations.
First, we have a deep empathy for builders of physical stuff. I’ve personally been through the pain (and joy) of making real-world objects at both my first company and at 3D Robotics, the company that acquired us. We built a lot of hardware at both companies. We also built a software product for construction called Site Scan, in partnership with Autodesk. Throughout my work with Site Scan, I did my best to better understand the lives of the builders for whom the product was designed. I learned a lot, and couldn’t shake the feeling that there was something more we could do to make the lives of all kinds of people who work in the real, physical world — especially the men and women of the construction industry — a little easier and a little better.
Manufacturing is hard, I know that firsthand. From my experience at companies past and present, I know that construction is 10X harder.
But it’s not just about my personal wish to make a difference here. As a society, we really don’t have a choice. Our second observation is that the construction and real estate industries are at a critical point, and winners will separate themselves by evolving — whereas losers will be those that don’t change.
Why is there a need for change?
We’re facing many deep and intertwined problems, from the lack of affordable housing to a breakdown in critical transportation infrastructure, and it’s clear that building will be an important part of the solution. But while the list of projects that need to be done is growing, the number of people we have to do this work is shrinking. The U.S. alone has lost 1 million construction workers since 2008.
This affects us all, and we have to find more creative solutions. Luckily, there are two mutually-reinforcing technology trends that can be used to help tackle the problems I’ve described above.
Our third observation: cameras are becoming ubiquitous. To co-opt (and slightly butcher) a phrase my former colleague and 3D Robotics CEO Chris Anderson likes to say, “There are many peace dividends of the smartphone war.” One of these dividends is the ever-increasing power of camera technologies. Because of the heavy investment into smartphone cameras, all kinds of cameras are getting better and smaller, and can find their way into more places than ever before.
If you can put a camera where there wasn’t one before, there’s probably a business in there somewhere. Camera drones, doorbell cams — they are all part of this trend.
Fourth, computer vision systems are getting smarter. Investments in self-driving cars, autonomy, robotics and AI mean that we are building software that is getting better and better at parsing the physical environment based on video feeds and image data. The amount of images and video being generated by all these cameras is too much for a single person to make sense out of, but machine vision systems can sort, organize and compress this information into useful insights. This, in turn, helps people understand situations better and solve problems more efficiently.
All of the above observations are true. We knew that if we could harness these emerging technologies and apply them to pain points we saw in the building industry, we had the potential to create something really useful. After all, a picture is worth a thousand words — imagery data can provide an unbiased source of truth about a physical reality. But it’s 100% false to say Mike, Philip and I outlined these as axioms on a whiteboard one day, convinced ourselves they made sense in theory, and then started the company.
There was a lot more trial and error than that.
Tech companies typically fail not because they can’t get their tech to work, but because they build something that no one wants. Knowing this, we knew it was important to try out ideas and bootstrap before committing to go all-in. It also helps to go with what you know: we had gotten to know a number of builders at 3D Robotics, and they were nice enough to let us try out our early ideas on their sites. In particular , I am grateful to WEST Builders, Palisades and Lendlease for giving us this opportunity.
About that a-ha moment: I can remember very clearly testing our tool one day on a site overseen by Nick Mirkovich, a PM at WEST Builders. At that time, we had built a tool that could deliver something like the full visual record you see in our product today, but the imagery capture was a very manual process. I had used our software around Nick’s site right before a concrete pour, and showed him the results in the trailer. His first reaction was positive: he never had a record of what was under the concrete if any issues came up after the pour (and he had certainly seen his fair share of costly issues).
But, being a good PM — focused on cost, labor and schedule — he immediately asked me how long it took me to collect all those images. The truth was, it took a long time, because our tool at that point was pretty manual, and collecting the data was a time-consuming task.
Nick and his team don’t have time for new tasks. They, like the whole industry, are stretched thin enough already.
After I told Nick how long it took me to capture his site, Nick paused for a moment.
Then, he said, “You know, if you can take that camera, make it smaller, and just put it on my, Travis’ and Brian’s hardhats…we walk the site all day. That’s our job. If your little cameras could be ‘on’ while we did that, and if you could somehow take all those videos and turn that into what you are showing me here…well, we would buy that.”
Nick, I don’t know if you remember that day, but it was the literal a-ha moment for us. I knew intellectually that great products remove work for their users, they don’t add it. But Nick made it crystal clear that while this data was massively useful, it had to be incredibly easy — zero labor, even — to collect.
The technology to take a random string of images and accurately geolocate them, with no input from the user and no reliance on GPS or other external signals, is technically quite difficult. Creating a system that is reliable, fast and accurate presents a real computer vision challenge. Luckily, we specialize in solving them.
Since that day, we’ve focused on making simple, easy tools that are backed by powerful technology, designed to make builders’ lives easier. And I am happy to say that we have been growing very rapidly. Our team is now a mix of construction industry veterans (much of the team comes from industry, some with decades of experience) and some of the smartest computer vision and technology people out there.
From starting the company in August 2017 to coming out of stealth last summer (thanks in particular to Darin Peters at Hathaway Dinwiddie and Lincoln Wood of Turner for taking a chance on us while we were in stealth), we’ve done a lot.
We are now deployed on projects globally that total over $50B in construction value. Our tool is becoming part of customers’ daily workflows. We enable them to create “high frequency digital twins” of their job sites, making things like coordination between subcontractors, dispute resolution, and future renovations smoother and more efficient.
I’m confident that in a few years, having the type of image data we provide will be absolutely standard on construction sites. In the near future, buildings will have records of themselves that can be queried and explored just as easily as we search the web. We’re excited to be part of that.
So what’s next? Well, suffice to say we didn’t start OpenSpace with just our current product in mind. Great companies are stories told through a series of products and product families, and we have a lot of ideas in the works.
I will say this: the amount of data our customers are collecting is staggering. In just over a year, we have collected 150 million square feet of imagery data — enough to fill the Empire State Building 50+ times. Our current rate is about 7 Empire State Buildings of data per month, and growing. There is a lot we can do to organize and analyze that data to solve problems in ways that were previously impossible.
So, thank you all for helping us get this far. It’s a good time to take a breath, look back at how far we’ve climbed, and head up to higher heights.