- Two Grossly Undervalued Robot Stocks
- Dialing in to “Mission Control”
A potentially mischievous new species of robot is arriving which can teach itself about the world, learning how to do things on its own, forming learning networks with other robots, even using body language to ask other robots for help.
I think this new species will convince many investors that robots are finally living up to their promise.
The EU, for example, has just completed a project led by Sweden’s KTH Royal Institution of Technology that got a couple of NAO robots (pictured below) to use body language to ask for help in lifting something, and to recognise when its fellow robot needed help.
Off-the-shelf Robots Cooperating
Google’s robot car and AlphaGo <see my AI Report> are other dramatic examples of step change advances in the field of machine learning and robot autonomy.
Amazingly enough, these developments are under hyped, if anything, in the media. In this briefing I’ll point to the most promising investments in this field.
Learning from the Cloud
I like Brad Templeton’s observation that ”a robot will be truly autonomous when you instruct it to go to work, and it decides to go to the beach instead.”
We are some way yet from robots playing truant, but recent developments merit the attention of professional investors….
Arguably, the Google car remains the world’s pivotal robot. Google ‘gets’ the internet and IoT. The car uses the network to access Google’s enormous database of maps and satellite and street view images and combines it with streaming data from GPS, cameras and 3D sensors to monitor its own position within centimetres and with past and current traffic patterns to avoid collisions: after millions of miles traveled on public roads with only a handful of bumper scrapes.
In short, robots are beginning to live up to their potential because they can now have their heads in the Cloud, using wireless networking, sensors, big data, machine learning, open source and the internet of things to improve themselves, whether it is to heighten their driving ability — Autopilot’s temporary blindness coupled with driver error or enhance their housekeeping <iRobot‘s wi-fi connected vacuum cleaners armed with 3D sensors and VSLAM navigational computer> and surgical skills <Intuitive Surgical‘s da Vinci>.
Robot design is being transformed as robots don’t have to carry nearly so much onboard processing power and memory, but can outsource them to the Cloud — indeed they have to in this era of not just big but ginormous real time data analytics. A good example is Amazon’s digital assistant in a capsule, Echo/Alexa.
And just imagine the onboard computing required to make feasible the “Belief Spaces” that are crucial to evolving robot applications.
A belief space, despite the spiritual sound to it, is a mathematical term that describes the modelling of the environment, sensors and actions using distributed probabilities. As these become more complex finding solutions requires enormous amounts of computing power, which is why until Clouds of computing power were formed Belief Space was considered intractable.
iRobot and Teradyne look grossly undervalued
Robot companies, big and small, incumbent or insurgent are being forced to adapt to the new scene — whether the industrial robotics giant Fanuc, robot vacuum cleaner supremo iRobot or the new breed of collaborative robot makers Rethink Technology or Universal Robots.
Indeed, I reckon iRobot, which has been around since 1990 as a spin out from MIT and which created and controls nearly 70% of the global robot vacuum cleaner market, is only now poised for major growth as the addressable market for its products swells to at least $7 billion a year in a runaway home automation market.
The company has persisted in missing its targets and upset Wall Street in the process over the years but has a clean balance sheet, a healthy free cash flow, is full of patents and has a deep bench of roboticists. It looks like a growth stock, and is a clear takeover target: it goes without saying by a Chinese conglomerate.
And how shrewd of automatic testing company Teradyne to acquire Denmark’s Universal Robots for $285 million last year. UR has a commanding presence in the nascent market for ‘co-bots’ — robots that work side-by-side with featherless bipeds. While Teradyne’s tester business is growing by 5-7% a year, the global co-bot market is on a CAGR of 50% from its low revenue base of around $100 million and is on track to hit $1 billion a year by 2020. Barclays puts it at more $3 billion.
Teradyne’s big customers, such as Samsung, Qualcomm and IBM, had been asking for advice, in pre-UR days, on which third party integrators to turn to automate manual tasks. UR promises to be the tail that wags the Teradyne dog.
Dialing in to Mission Control…
Finally, a prime way of getting a handle on the direction advanced software and digital technology are taking the world is to attend Google’s annual I/0 conference for developers, or at least watch it live.
This year’s 1/0 conference in May was no exception: over 7000 developers and analysts attended it unvirtually, and several million ,on and off, via live streaming, including me and over a million from China.
Google is without doubt the world’s advanced software powerhouse, sports an insane computing resource, and is one of a relative handful of internet platform companies that are kind of mission controls to the future — Amazon, Microsoft , Facebook, Tencent.
Apart from tantalising reports at 1/0 of what Google is doing to turn the nascent virtual/augmented reality industry into perhaps the world’s biggest industry by 2030 and many other bits and pieces, a highlight of 1/0 for me was a video of a group of robot arms teaching themselves via a continuous feedback of networked failures and successes to pick up objects and move them from one box to another.
No pre-programmed rules had been involved. Taking a leaf out Deep Mind’s AlphaGo book, the Google team tasked with devoting 20% of their time on promising projects simply decided to leave the robots do the project by themselves.
That’s the way it’s going to go across the board, with, as is the case in this non-linear, runaway world, turmoil and opportunity ahead, and even the possibility of robots that decide to bunk off work.
ps: over the next week or two I’ll be working on an in-depth report for subscribers of the exploding world of virtual, augmented and mixed reality and the next computer platform it is yielding for the internet of ‘presence and experience’ and looking at the companies in the vanguard.
And if you haven’t already, do follow me on Twitter: @ormebytes and we can take some of the interesting debates we’ve been having online.
If you’d like to contact me about any of the exponential developments in genomics, robotics or AI, drop me a note here.