After selling 3 million robots in various markets, the people behind Sphero have launched a new business, Misty Robotics.
In this exclusive interview, Robotics and Automation News speaks to the CEO of Misty Robotics, Tim Enwall, whose background includes senior jobs at Apple and Nest Labs.
A long time ago, in a galaxy far, far away
In today’s fragmented society, with so many communities and choices, it’s becoming increasingly rare to find things everyone instantly recognises, and the Sphero BB-8 is one of those rarities.
There’s something in the human psyche that is drawn to spherical objects, especially ones that move. And if it seems to be moving autonomously, all the more fascinating it seems.
And if it’s got a dome on top with a camera on it that makes it look humanoid and robotic at the same time, as the BB-8 does, it’s almost guaranteed that it will become a craze, which it was.
Sphero’s BB-8 was featured in one of the Star Wars movies, but in reality, the design of the robot – reminiscent of the old R2D2 robot from the first Star Wars films – is probably good enough to have found a similar level of success without such stellar product placement.
Enwall explains that it was Ian Bernstein, founder of Misty Robotics and of Sphero, who was behind the creation of BB-8.
“Ian was the founder of Sphero,” says Enwall. “He’s been building robots since he was 12 years old.
“The things that Ian has learned include the intense importance of personality and character in a robotic experience and interaction.
“The difference between the Sphero ball and the BB-8 was just gigantic in terms of the way people would engage with and spend time with robots. That’s one thing.
“Another thing is, from the very beginning, they made their robots programmable, and graphically programmable by fourth-graders.
“He’s got a ton of experience on how to make something complex, and then translate it into something that’s super-easy to program, even for a fourth-grader.
“Of the seven years and 3 million robots Ian has shipped, two-and-a-half of those years he has spent in China with our manufacturing partner developing great relationships.
“So he has experienced first-hand all of the challenges and joys of mass production and manufacturing.
“Speaking for myself, I’ve learned a number of things. Certainly the joys and difficulties of mass producing consumer electronics products, at companies like Revolv and Tendril and even before that with my background at Apple.
“So I know, from experience, that hardware is hard, and you just run into all kinds of problems that you didn’t know would exist.
“With a robot, you get into even more problems because of the physics. Consumer electronics is mostly about digital and electronics, but with a robot, you’re talking about all kinds of physics problems.
“So I’ve learned a lot about the difficulties and how you need to bring in expertise, and the planning cycles, and the difficulties of mass producing something.
“And also I’ve really learned through those experiences the importance of being very clear about the customer you’re going after, and not talk yourself into trying to serve a number of different customer personas.”
Misty Robotics is a spin-out from Sphero. The startup has raised almost $12 million in funding since last year to follow through on its vision of putting “a robot in every home”.
Is that realistic? Who’s to doubt it? When you consider the ubiquity of personal computing devices, it doesn’t sound like an impossible dream.
Enwall says the Misty developer robot is specifically aimed at people who want to begin programming robots.
“We believe the world is at an inflection point,” says Enwall. “Whereas it used to be about roboticists, now it’s about robots for everyone.
“And Misty Robotics’ concept and strategy with robots is not a direct consumer strategy – like iRobot, or Amazon, or SoftBank – our strategy is drawing from history.
“The history of the web, the history of the personal computer, both started with a concentration on programs. Web technology, personal computer technology, for the first few years, was sold exclusively to programmers because that was the only thing you could do with web technology at the beginning; or personal computer technology at the beginning – programming.
“And it turns out that tens of thousands of developers were unleashed as inventors and they invented all of the applications for us as consumers, whether on the web as e-commerce applications, magazines, and transaction engines; or in the personal computer, with spreadsheets, databases and word processors and accounting packages and the like.
“So, we are bringing out a magnificently built, well-designed robot that does most of what you want a robot to do, is affordable, at the price of a good programmer laptop, and is super-easy to program.
“And almost every programmer we talked to, who is not a roboticist, said, ‘Wow, that would be amazing’.”
2018: A robotic space odyssey
While many might believe there will come a day when there will indeed be a robot in every home and business, most probably would say it’s a couple of decades away.
It’s inevitable, but it’s a prospect that may unsettle some and excite others.
From a business point of view, the potential market is just so huge that it’s worth developing products for it now.
The global robotic vacuum cleaner market is said by TechSci Research as being worth around $3.4 billion. If that’s anything to go by, then Misty Robotics has got a great chance at making a fortune.
But the specific market segment in which Misty perhaps belongs is still very new and not much is known about it.
Interestingly, though, Enwall makes it clear that Misty is targeting the developer community – that is its market more so than the home market itself.
“In one sense, there is no market because there’s nothing like this for a web or mobile or enterprise programmer,” says Enwall.
“Yes, there are enterprise programmers that are going into buying expensive Nao robots, and they’re programming those Nao robots, and there are some who will go buy TurtleBot and strap a laptop to it so they can get a camera and speaker, but that’s a heck of a lot of friction.
“So I think the market is very nascent.
“We’re going after three types of programmers.
“The first is the enterprise programmer who knows that a robot might be able to help them in some way shape or form but are not quite sure how.
“The second market we’re going after is the entrepreneur who sees the opportunity that a robot could fulfil and is going to write their own software and sell the combination of their software and our robot into that space. So, for example, elder care, telepresence, security patrol bots and so on.
“The third group of programmers are those who want to improve their livelihoods in the future by learning how to program robots. It’s also parents who might want to enable their teenage children to enter the programming or robotics sector as perhaps a career.
“If you look at those three groups, there are some enterprise and entrepreneurial programmers who face a lot of friction and expense getting those other solutions to operate autonomously and might want to choose the Misty Robotics solution instead.
“In the third group or market, Lego Mindstorms has done really well. There are some other platforms for people to learn about robots.
“So, you know, we have no misconceptions that this is the dawn of a new market, and, frequently, those are the largest markets that appear.”
Robotic lingua franca
The Robot Operating System is a set of frameworks to be used for programming robots, and it’s been catching on around the world, with many hundreds of thousands of users integrating it into their developments.
Enwall says ROS might have limitations when it comes to dealing with real-time systems, object- and face-recognition tasks, and what’s called SLAM, which is simultaneous localisation and mapping, something used by devices such as robotic vacuum cleaners to help them navigate a given environment like a home interior.
He adds that the Misty robot has two processors which are of a class used in Samsung and Apple smartphones, and the way they are set up makes the system more suitable for cloud connections.
Enwall says: “We believe that inspiration comes from a million different places.
“Some people are going to use the autonomous mobility and screen and microphone and speaker locally.
“But some people are going to do all that processing in the cloud, so they’ll move processing back and forth between the local processor and the cloud, especially where material in a cloud is oriented around the robot’s learning.
“It makes sense for the robot to spend a few seconds or a few minutes learning something because it’s got the time to take something back to the cloud and there’s a computer somewhere processing it.”
The million-dollar robot and the uncanny valley
Speculating on the future market potential of Misty Robotics, Enwall is of the contemporary technologist mindset which says that market testing is the way to go – it’s the only way really to know if you’re product will be successful.
Enwall says: “You never know what the response is going to be until you actually put a price on something, which is why we’ve put a price on it of $1,499 even while we’re selling a few dozen.
“So you have to validate what you think is a value proposition with a price and then gauge the response.”
And while Misty Robotics has no plans to develop the type of humanoids and androids seen in Star Wars, Ex Machina or any other science fiction film, Enwall indulges Robotics and Automation News with some speculative comments.
He believes humanoids and androids which pass the Turing Test, as well as the “uncanny valley” test, are decades away.
The “uncanny valley” refers to the feeling one might have when observing a robot which has some functions which are uncannily realistic, natural or humanistic – and yet, in its totality, it can be clearly seen as a robot.
The unsettling feeling that you might get when watching one of those scary Boston Dynamics robots is when you’re in the uncanny valley.
But anyway, Enwall says we have nothing to worry about and that we’re all safe – for now.
“On the mechanical side of the equation – motors, actuators, degrees of freedom and so on – there’s so far to go,” says Enwall.
“And on the artificial intelligence part of the equation, the software side, AI today is good at very targeted realms.
“But the notion of having an AI that responds well in all situations that you run into, that’s just so far away.
“If you look at idea of the ‘million-dollar robot’ that Boston Dynamics has produced, all it takes is to look at that robot and see just how far away we are from something like Ex Machina.
“I don’t know if we need to get there, or if we will get there.
“You might have heard of the uncanny valley, and there’s just no sense in having a robot that’s in the uncanny valley, so I wouldn’t be surprised if people tried to stay away from humanoids, or androids, until it can really cross over that uncanny valley.
“I just think that’s a long time in the future.”
While we all wait for super-realistic humanoids and androids to make investors billions of dollars in the future, the current investment climate is very good.
There is an excitement, a buzz, around robotics and automation that seems new and yet also taps into a decades-old understanding of the capabilities of the technology.
It’s just that now there are many advances in computing, electronics and engineering that seem to suggest all the components are available to create very powerful – and profitable – businesses.
Misty Robotics itself raised $11.5 million funding when it spun out from Sphero last year.
The funding was led by Venrock, whose investment in Intel was one of the venture capital firm’s first, in 1969. Venrock was also an early investor in Apple. The company is also an investor in Pearl Automation and Nest.
Misty is also supported by the Foundry Group, among others.
“We’re well funded for quite a long time right now,” says Enwall.
Moreover, across the entire robotics and automation industry – which also includes AI – there is more and more money coming in.
Stock market investment funds based on automation and AI are said to be outperforming traditional industrial sectors, including many tech segments.
But is there a real difference between investing in traditional companies and investing in robotics and automation companies?
“I don’t see a lot of difference,” says Enwall. “It depends on which investors you’re talking about.
“If you’re talking about early-stage investors – like venture capital firms, versus late-stage investors – like public investors and funds, then I think when it comes to venture investing, the reason robotics is hot is because – and I think this is true for the public markets as well – is that both the early stage and the public stage investors know, or at least have a very strong belief, that robots are going happen in the future.
“You know, automated cars and AI and machine learning – all these advancements – Amazon Alexa… People can see that the stars are aligning and, yes, you know it’s going to go through its ups and downs… Gartner has its hype cycle – the peak of hype and then the trough of disillusionment… so it will go up and down through its cycles like every market.
“But I do think most people believe that the robot market will be a trillion-dollar or more market globally.
“So I think that’s why you’ve got bullishness from investors both early stage and late stage.”
Solving robotics is people
For Misty Robotics and many other companies, Enwall says one of the keys to success is the people. Enwall himself has a 20-year background in building businesses; Bernstein has a tremendous track record in building robots; and a new hire – Ben Edwards – is an experienced expert in the internet of things as it applies to intelligent devices and networks.
“Ben’s the founder of SmartThings, the leading smart home hub, which was sold to Samsung a couple of years ago,” says Enwall.
“He was one of the seven founders of SmartThings, and he was specifically responsible for building their developer ecosystem.
“He built it from nothing to 50,000 developers and he’s super-experienced and knowledgeable about the tools that developers need and the communication and the style, and even where developers hang out and what they want to see from a company.
“So he’s had a lot of success building a developer community, and we’re really excited to have him join the company.
“And if you think about it, IoT developers are a pretty good segment of developers that would be interested in robots because they already have some experience of developing some physical things, so there’s a lot of holdover from Ben’s experience with IoT developers as well.”