Over the past 20 to 30 years, the exoskeleton market has seen a variety of changes, from the prosthetics space to military applications that support soldiers in logistics or those who have been injured. These days, many companies are moving into manufacturing and industrial applications, utilizing robotics to help workers get their jobs done more effectively and safely.
One company on the forefront of the space is Sarcos Robotics, which creates a number of exoskeleton and robot platforms for customers across manufacturing, industrial and construction markets. The company has raised more than $56 million in funding, most recently a $30 million Series B round in September 2018.
Earlier this week, the company announced a partnership with the U.S. Navy’s Puget Sound Naval Shipyard & Intermediate Maintenance Facility to evaluate and deploy robotic technologies, including full-body, powered exoskeletons and man-portable inspection robots, for use in naval shipyards. The shipyard will evaluate Sarcos’ Guardian XO exoskeleton and Guardian S inspection robot across several unstructured, challenging work environments and tasks, including those that support the maintenance, modernization and retirement of ships and submarines.
Robotics Business Review recently spoke with the company’s CEO, Ben Wolff, about the evolution of the company, the problems the company is attempting to solve, and why exoskeletons are poised for big growth in the next few years.
Q: Talk about some of the early days of the company and its evolution towards exoskeletons.
Wolff: The company has its origins as a spinout from the University of Utah going all the way back to the early 1980s. At that time the team was focused on human prosthetics, and we created what was the first electrically actuated prosthetic arm. So you could call it in effect, kind of a robotic arm, as it had actuation. So that was our initial foray into this notion of creating electrically actuated or other types of actuated machines. That prosthetic arm wound up becoming the number one best-selling prosthetic arm in the world. Our successors are still among the leaders of electrically actuated arms.
What that experience did for us though was it kind of ingrained in our team a sense of biomechanics and understanding how the human body moves, and how humans might intimately interface with machines in a safe and productive manner. So that was our beginnings. If you kind of fast forward to the 1990s, we translated our prosthetic work and our work with sensors into robotics work, and we started designing and producing humanoid robots. We did those for government agencies. We’ve produced them for research institutions around the world. We produce them for private industry. In each case they were highly customized, highly capable humanoid robots. We also produced them for theme parks around the world — different types of animatronic robots. So we are certainly, if not the top, among the top producers of humanoid form factor robots in the world.
We continued to go down that path until 2007, when Raytheon bought us. From 2007 until 2014, we were the robotics division of Raytheon, with an almost exclusive focus on DOD customers.
Q: How did the company start working on exoskeletons?
Wolff: Back in the 2000-2001 timeframe, we started working with DARPA on exoskeletons. And we’ve been working on them ever since. DARPA provided a lot of the initial funding to get our exoskeleton effort underway. During our time with Raytheon, exoskeletons were a significant area of focus. In 2009, the Pentagon issued an initial requirements document that talked about the need for full-body powered exoskeletons for use in the military, primarily for logistics applications. So we had back in those days, a high hope of participating in that program.
But then Congress imposed sequestrations, I think it was 2010, and all new programs, not just the exoskeleton program, but all new development programs were cut because of the lack of funding. As a result, while we continued to make progress, it wasn’t funded, and we didn’t really have a customer.
So ultimately, that resulted in the opportunity at the end of 2014 for me to get involved to partner with the management team, and complete a management buyout. We spun back out of Raytheon, and became an entirely separate independent company.
If you look at where we were back in January 2015 when we completed the separation from Raytheon, we were a whopping 13 employees, but with a very rich history and a tremendous amount of intellectual property. The team that developed – those 13 people that came with the team in January 2015 had an average tenure of working together at the time of 20 years.
Q: That’s a pretty rare thing, isn’t it?
Wolff: It’s rare in any industry, but then you think about it in the robotics industry to have a team that has been, been through so much, seen, so much developed, so much new technology and innovated so well, but to have a core group of engineers that have been together for 20 years is something pretty special. That’s certainly what I saw, and this got me interested in getting involved in working with this team.
Fast forward to where we are today, with 110 employees. We’ve got employees in about four or five different states. We are focused, unlike in our past where we did a lot of custom development work for customers. We are now very much focused on our core products and bringing them to market – these are core products that we’ve worked on for a lot of years, and invested millions and millions of dollars in.
Enhancing workers with exoskeletons
Q: What are the products that the team is now working on?
Wolff: Our first product is the snake robot that you’ve probably seen that has actually started shipping to our commercial customers now. Then there’s the exoskeleton the line of products, which includes a small, medium and large version, not from a size perspective, but from a lifting and payload perspective. And those are machines that we expect to have out in the market by the end of this year.
We’ll have our first units in alpha customers’ hands in 2019 and then we expect to ramp up manufacturing and to start producing at scale and shipping and scale in the first quarter of 2020.
Q: Talk about the Guardian GT exoskeleton and its purpose.
Wolff: It’s an offshoot of the Guardian XO. It is a supersized exoskeleton, if you will. GT stands for giant tele-operator. It’s kinematically equivalent to a human in terms of the distances and ratios of the human body, just on a supersized scale. So the arms of the GT are seven feet long, and you can lift 500 pounds in each arm and you can perform all of the same kind of dexterous maneuverability and movement that you can with your human body, but in effect in this very large supersized machine.
Q: Do you consider this an exoskeleton rather than a robot?
Wolff: I consider it both. The reason I consider it an exoskeleton is because you can ride inside the machine and manipulate the arms while you are inside of it. You can ride behind it, or you can tele-operate it at a distance. So it has the flexibility to be able to do any of those things.
Augmenting tasks for workers
Q: Does it have autonomy, or are you working on making this autonomous?
Wolff: Our overall thesis for all of our robotic platforms is that human intelligence is the primary driver, and we augment human intelligence with forms of artificial intelligence to lighten the load on the operator. In other words, to make use of the suit easier or use of the robot easier.
But major decisions we think will always be driven by a human operator or a human manager. Now you might see situations where the human operator or manager is managing a fleet of robots, but ultimately major decisions are going to be made by the human.
Let me be clear about why that is. Our robots are not designed to be used in environments or for jobs that can be automated. Our robots are all designed for environments or tasks that kind of defy automation – where there’s enough variability involved in moving from one task to the other, that human wisdom, judgment, reflexes, and intuition are an important part of getting the job done.
Q: Can you give us some examples of these types of tasks?
Wolff: Sure. Let’s talk about even a fairly sophisticated manufacturing environment like in the automobile industry, where some number of cars are made on a manufacturing line that requires some amount of rework. Every automobile factory has a rework section where the automation part on the assembly line didn’t quite get the job right. So you pull the car off the line and something manually has to be done to fix it. That might involve taking the muffler system off, it might involve taking the drivetrain out, it might involve all kinds of different things. That is a very flexible set of tasks that have to be done because it’s not the same thing every time, otherwise it would have been done on the line in the first place.
Another example, let’s take a construction job site. Every construction job site is different. The terrain is different. The structure that’s being built for the most part is different.
I’m not talking about laying the bricks, you know, those kinds of routines. Routine tasks can be programmed and automated. But when you’re talking about pulling the whole thing together, the logistics applications, moving product on and off the job site, moving it around to different floors, all of that kind of stuff is fairly unique on a construction job site. So that’s another example. I can come up with a laundry list of things where human intervention is needed because of the dynamics of the task.
Q: So is it fair to say that you’re doing routine or difficult tasks, but humans are going to be there when the robots are doing it?
Wolff: That is true for the exoskeleton, at least the light duty and standard duty ones. The GT, on the other hand, because the GT can be tele-operated, it is intended to go into very dangerous and difficult environments.
We’ve had a lot of interest out of the nuclear decommissioning space, for example, because that machine can go in, it can use off-the-shelf power tools, just like a human would, but it can manipulate them in very dangerous and difficult environments where you wouldn’t necessarily want humans to go. And it also gives you the equivalent of superhuman strength because it can lift up to 1,000 pounds. So that’s a great example of going into a dangerous and difficult environment with the GT.
On the other end of the spectrum, if you take a look at our Guardian S snake robot, it is absolutely designed to be able to go into confined spaces, dangerous, difficult, dirty environments where humans really ought not be going, and to provide better type of reconnaissance in inspection and surveillance information than what a human should or shouldn’t do on their own.
Q: So you do have a couple of different markets and applications where you can use a lot of these different products.
Wolff: In the case of the Guardian S inspection robot, if you look at the kinds of tasks that our customers are using them for today, in a power plant, for example, doing an inspection of the power plant. Every power plant is different, every environment is different.
It’s not like a warehouse where a lot of modern warehouses can be constructed the same, and you’ve got a smooth floor, and you have shelves and stacks and just a fairly manageable environment of being able to have automated mobile robots with an autonomous function.
When you talk about inspecting the power plant that was built 30 years ago, none of them are the same. There’s no consistency there. That really illustrates our purpose of augmenting humans as opposed to replacing them.
Q: You recently had some news around the Guardian XO, can you dive into a little bit about what’s new there, what kind of feedback you got from industrial or manufacturing partners?
Wolff: That has been a labor of love. I mean, we’ve been working on it since 2000, and we’ve had almost a dozen different prototypes and iterations over the years. At a macro level, four years ago we couldn’t be thinking about doing what we’re doing today because it still didn’t have the cost, power, size, and performance of components at a point that would make it commercially viable.
What we’ve seen happen over the last four or five years in terms of the power consumption, the efficacy, the capabilities of the footprint, things like that. That’s why you see across the board, robotics are coming into prime time now. They’re no longer just on the manufacturing floor for assembling.
That’s one of the things we’ve been a beneficiary of, which makes this possible. But from the standpoint of our specific efforts, the three big things we’ve had to focus on over the last three years are power consumption – that problem was the single biggest deal – number two was having intuitive controls, and number three was having a machine that could move with you from a degree of freedom perspective.
Q: Can you explain more about the idea of improving your controls and freedom of movement?
Wolff: Essentially the idea is that you could just get in the suit, start moving it, and if you wind up having something go wrong, or if you have a challenging environment that requires you to really focus mentally on what you’re doing, you don’t want to be distracted with having to think about how you’re going to manipulate the machine. It just has to be an extension of your body, intuitive to use. So we put a lot of time into our control system and perfecting that.
The third big thing was having a machine that could move with you from a degree of freedom perspective, from a freedom of movement and motion for the human body so that you’re not restricted from moving in the ways that you’re used to. For example, trying to retrain the human body to do things, because the machine won’t move with you because it doesn’t give you the flexibility and freedom of movement that you’re used to.
So those are the three things that we’ve really focused on. I’m happy to say we’ve achieved a lot to the point that we’re now ready to start putting these machines into our customers’ hands next year.
Q: Would you say this is a big reason why a lot of people are predicting that the growth of exoskeletons in manufacturing or industrial use cases is going to grow in the next few years?
Wolff: There’s a culmination of factors for those projections. Number one is we’re finally seeing the technology get there, and the price points get there. Number two, if you look at some of the forecasts that are out there about labor shortages, not only in the U.S. but globally, particularly in Western Europe, Asia, Japan and Korea specifically. We have some serious problems with labor shortages, aging workforces and frankly an issue of new generations not being particularly receptive to going into jobs that are manually challenging, that are back-breaking jobs. When you take a look at some of the forecasts, whether it’s in the construction industry, the manufacturing industry, or logistics industry, there’s a lot of concern about how we continue growth and to enhance prosperity both at a corporate level and a national level, when you don’t have people that are willing or able to go into the workforce to do the jobs. So I think exoskeletons are going to become a major player in helping to solve that problem.
At the same time, you’ve got a real movement in environmental health and safety parts of large corporations, where they’re focused on how do we prevent injuries with more success than we’ve historically done. Not only because that’s the right thing to do, but because there’s also economic consequences to not having too many injuries and deaths on the job.
So when you’re looking at what we’re trying to achieve with exoskeletons, generally it’s how do you make up for the fact that we’ve got labor shortages in a lot of these industries, coupled with how do we start putting put a real dent in this [health and safety problem], where we spend $100 billion a year in the U.S. on occupational packages.
You put those two things together, there’s a confluence of factors and an attentiveness to those issues that are creating a huge opportunity for exoskeletons.
Q: One final question – we noticed that you’re the part owner of a winery – are there any similarities between running a winery and running a robotics company?
Wolff: In every case you’re dealing with people as your primary asset in your business. So understanding how to lead and inspire and grow a team that is the greatest team that you can possibly put together. But that’s in common with any, any business. If you kind of look at the product itself, we have a passion at both companies to delight our customers, and just do that in different ways.