How does Misty compare to other robots?

“Personal Robots” are an entirely new category of robot

We get asked this question a lot: “What’s the closest competitor to the Misty robots?”

It’s a good question, because there’s not a straightforward answer.

In the broadest sense, our competition is the mundane and boring human tasks that we perform all the time — so the competition is each human’s capacity for pain tolerance. For centuries now, some humans at the upper end of the earning scale have paid other humans to perform these often boring and rote tasks, especially within the home (butlers, maids, nannies, nurses, etc…) or in the office (inventory cycle counters, greeters, tour guides, etc…). Those of us at the“regular” end of the earning spectrum tend to perform more of these tasks on our own, and rob ourselves of the time to pursue our passions, our joys, and our relationships. We don’t exactly have the tens of thousands of dollars to employ those other humans, nor the millions required to buy a Boston Dynamics’ robot.

In a narrower sense, our competitors are “consumer” robots — robots meant for wider consumption by a broader audience of user. These differ from “industrial” robots designed for fairly specialized tasks, like moving material within a warehouse (Kiva) or cooperatively assisting a manufacturing line worker (Rethink). There’s certainly a wide variety of consumer robots out there — robots that are doing some pretty amazing things. Robots like Kiki, Aibo, UBTech’s Lynx, Cozmo (may it rest in peace), and our spin-off siblings, Sphero.

What separates Misty robots from those other consumer robots is three things: the variety of use cases performed, where those use cases are performed, and the market targeting.

Let’s start with market targeting. Most of the consumer robots are targeted at, well, regular consumers. If you go to their web sites, you’ll see families and children in bright white environments. These robots appeal to every-day users who have no specific specialization other than being a family or being children in avid search of a toy or education experience. The companies who make these robots work hard to figure out, “What single thing would the average family want in a robot?” and seek to deliver it — and deliver it to very exact consumer specifications. They sell the products in consumer retail environments. They’re taking their cues from very successful consumer robot companies, like iRobot with their Roomba product. They’re not targeting the office or office workers. You’ll rarely see a picture of someone at work next to one of those robots above. They’re strictly for the home. In fact, most of them position themselves in the category of “home robot”.

Misty, on the other hand, is targeting a very different and very specific market: software developers (and those who employ them) and teachers/parents of STEM students. Inventors, creators, and dreamers who have never before had an advanced and easily programmable robot. Somewhere lurking inside the heart of the 20 million programmers in the world is a young 12-year-old who has fantasized about what they could make a robot do. Or, there’s an actual 12-year-old dreaming about “Robots Meet Fortnight” and who is ready to make that happen. We’re here to begin to help fulfill those dreams. We’re still a long ways away from ultimately fulfilling those dreams — robots need hands, they need more electrical power consistently, they need more intelligence, they need more language skills — but we’re here to get a start, with robots who have agency, can navigate themselves, and can see, speak, and hear. When you see our advertising or come to our web site you won’t see families and K-8 children in the picture — that’s not who we’re serving. We’re serving the creators who will make thousands of uses for the rest of the consumer marketplace over time.

Speaking of thousands of uses, Misty is built to create thousands of use cases, whereas each of those consumer robots listed above is solving a particular use case, whether it’s replacing Fido the dog (Aibo); being a jovial companion (Kiki); providing remote eyes, candid photos and information retrieval (Lynx); or serving entertainment and education to K-8 students (Cozmo and Sphero). In contrast, Misty is built to give access to the foundational elements of robots — self-navigation, personality, guided navigation, sound input and output, visual input and output, and AI capabilities — to enable use cases across the office and the home. By offering up the robot as an advanced and easily programmed platform, we enable one robot to perform thousands of duties. Consumer robot companies, by contrast, put a significant amount of effort into studying each duty, coding that duty into their robot, and offering it in the next revision of the robot. Everything is within their control, and because each “duty” has to be built to consumer specifications, it takes a lot of engineering time and a lot of calendar time to get it right.

Which brings us to where are those use cases performed. Misty positions itself as a “personal robot” for the office and the home. We believe people have as many mundane and rote tasks to perform in the office as they do in the home. And frequently, if freed from those mundane/rote office tasks, people can focus on more productive and profit-making uses of their time. So Misty robots are useful in the office and the home. Why “personal”? Because it’s affordable. It’s the price of a software developer’s laptop. Any individual in the office or the home could have their own robot perform a set of tasks that are time-freeing for them.

Which, finally, brings us to the narrowest sense: what other multi-use-case, built-for-developers-and-makers robots exist? I can hear you say, “because, surely, if there’s a 12-year-old robot dreamer inside those 20 million programmers, some of them must have been pursuing their dreams!”.

Absolutely true.

Today, the state of the art for those dreamers is either the iRobot Create or the Turtlebot3.

The iRobot Create is a non-vacuum cleaner version of iRobot’s very successful, very high-quality vacuum cleaner. It includes a robust autonomous navigation capability and is very affordable ($199 as of a search on April 18th, 2018). If you want sight, sound and eyes, however, you’ll have to literally duct-tape a laptop to it — which is, in fact, what many robot-explorers do. And then you’ll need to spend 3–6 months writing foundational software that enables you to command the robot to see, speak and hear. You’re not programming it within 30 minutes to do something meaningful. This robot is very functional — but not exactly something a corporate software developer would deploy into their office environment to solve a problem routinely. Not something they’d use to greet and escort visitors to a conference room. Nor something that an entrepreneur who wanted to build an eldercare business around robots would find appealing to sell for use in an elderly person’s home.

The Turtlebot3 (as of July 16th, 2019) is $1,399, and is a pretty sophisticated academic research platform. It comes with open source Robot Operating System (ROS) software capability, autonomous navigation via a 2D-plane LIDAR system, and an HD camera for visual input. Installing the software on a laptop, becoming familiar with ROS, and starting to program the Turtlebot3 takes a good week. ROS itself is very powerful for extensive robot exploration, but also not very performance oriented, and it’s difficult to place into more scalable, production types of applications. Even with ROS available, the average JavaScript or Python developer isn’t able to concentrate on and build their specific dream for several weeks (or months), because it takes a long time just to get the “core platform” to a level where it can easily perform the functions the programmer wants. Turtletbot3 has no visual output (no screen), no audio input/output, and no personality expression — to get those, you’ll have to — again — duct tape a tablet or laptop to the robot, and then write the integration software between the external computer and the Turtlebot3. Visually speaking, as with the iRobot Create, it’s a bit hard to imagine these cruising around the office or home by the millions.

Some might argue that Lego’s Mindstorms is a competitive product — which is true for the middle school crowd. But once kids get into high school and start to really get into JavaScript and Python, and then head to college to get into deep neural networks and machine learning, Lego Mindstorms quickly leaves the picture.

Which brings us back to Misty robots.

The advanced personal robot that’s easy to program.

By delivering a great looking product that one can see going into millions of offices or homes, that has more capability than either of those platforms, and that’s easily programmable in 30 minutes to achieve powerful skills, we believe we’ll appeal to a much wider array of software developers and makers who dream of building skills for robots. Like many disruptive technologies at the beginning, there is a very small market of innovators. But as happened with forbears like the personal computer and web browser industries — both of which started with software developers creating thousands of use cases for office workers and consumers — we are confident the market will expand rapidly as we deliver the ability for robot dreamers everywhere to begin to fulfill their dreams.