Can Robots Improve Safety on Construction Sites? Here’s How

Raaz Robotics is an early-stage startup based in San Francisco, and we’ve been around for about three years now. As you can see, we create robots to do construction work, and we’re really focused on using robots to help move people out of harm’s way on construction sites by having our machines do very highly repetitive and very challenging work.

Today we primarily work on the building envelope as well as the structure. These are two scopes of work that are basically the largest line items of cost on any building, and they’re also where people have to be put in situations that are really hard for them—for example, working at height where there’s danger of falling, or doing tasks that require a lot of strength day after day, which can lead to repetitive strain injuries.

This is called Metal Man. Right now, Metal Man does three functions: it installs brackets on the edge of the building, it does layout, and it drills holes in the concrete. These are three applications that a lot of our customers have requested, so we built them out for them. In the future, we aim to turn Metal Man into more of a multi-purpose platform. It’s basically a dual-arm mobile manipulator, and our goal is to reuse the arms to perform a variety of different tasks by developing a suite of end-effectors—essentially a toolbox for the robot.

Today we mainly work with building envelope subcontractors and structural retrofit subcontractors, as well as some self-performing general contractors. These are the people who typically do the work on-site today. I’ve always been really interested in building buildings, and it’s very clear that the world needs a lot more of them. But today, construction work is still very hard, and it’s a challenging industry to work in. I thought it would be a great opportunity to take some of my skills from working on robots for unstructured environments and apply them to this industry to help people get their jobs done much safer and much more quickly.

I think one of the really interesting things about robotics is that some of the advancements are actually very simple on paper. One of the biggest things is that robots are just cheaper today—it’s much easier to get hold of very complex robotic equipment like these arms, as well as high-quality cameras and LiDAR sensors. All of this has really contributed to making something like Metal Man very cost-effective for our customers. On top of that, we’re also taking advantage of the latest advances in AI. We’ve done a lot of work on our vision and control systems, developing proprietary methods that allow our robots to see, feel, and work safely in these very challenging environments.

A lot of our customers really like being able to take people away from the edge of the building—that’s one of the primary advantages of using a machine like this. It removes people from situations where they’re at physical risk. Customers are also really excited about the ability to use a robot that can perform tasks very consistently. People are very good at their jobs, but they get tired as they work, and when you get tired, there’s a higher chance of being careless or making mistakes. With a machine, we can guarantee very consistent and accurate work, which can save our customers a lot of headaches because they don’t have to worry about mistakes they might not discover until it’s too late.

One of the biggest things we’ve focused on is making sure we tackle applications on construction sites where we can really improve people’s day-to-day working environment—not just save on the bottom line. That priority has translated directly into the design of the robot. Our robots have arms that can reach much farther than a human can, allowing them to work from much safer positions than a person would need to—for example, instead of hanging off the edge of a building or standing on a ladder to reach high into the ceiling.

In the beginning, we had a lot of challenges. Our first prototype had to be disassembled and reassembled on-site. But today, with our latest version of the robot, that’s mostly solved. Deploying a robot is quite simple: we load it onto a pallet, ship it to the site, unload it, and the setup or pack-up procedure takes about half an hour. We don’t need any additional information beyond two-dimensional floor plans for our robots to operate on construction sites, which makes it very easy for our customers to set up and operate. They don’t need full 3D BIM models, and we can integrate directly into existing construction environments with very little effort.

If I had one piece of advice for any startup, it’s to really talk to your customers and understand their problems. A lot of times, things that may seem obvious on the surface turn out to be much more complicated when you get into the details. Spending a lot of time with your customers and really listening to their pain points is the most important thing. I’d also say, don’t be afraid of taking a big swing at problems. Some of the most important problems are often the ones that seem very hard to tackle but end up being very achievable if you take the time to truly understand them.

Construction tech is in a really exciting place right now. There are a lot of new projects coming down the pipeline, and there’s significant investment in building America’s physical infrastructure. I’d encourage anyone who’s interested—especially new roboticists—to look for opportunities to work in this space.