More Hacker Spaces! Techshop coming to Pittsburgh

Techshop is coming to Pittsburgh.  This will be a great addition to Pittsburgh’s DIY / Hacker culture–which has a slightly different flavor in Pittsburgh because-unlike the big coastal cities that are ‘rediscovering’ the idea of building stuff-Pittsburgh is a city that never stopped thinking of itself as working, industrial city.   On a personal level, I’m excited that Techshop is coming to Pittsburgh with a focus on veterans.

I’m not sure how we should view these kinds of DIY/Hacker spaces in terms of the robotics ecosystem.  Off the top of my head, I can’t think of any successful start-ups that got their start in these kinds spaces.  If you look at the hacker space websites, the kinds of projects that they tout as commercial successes are more in the consumer device space (e.g. artistic iPhone docks)  as opposed to commercial robotics.  On the other hand, they seem to be a good marker of the kind of culture that builds robots.  So whether this is indicative or causative of a great robotics scene, welcome to Pittsburgh, Techshop.

As an aside: I’ve updated the cluster comparison with a few of these developments and more DIY/Hackerspaces.  There are links in the cluster comparison to several resources in this arena.

What cluster does a company with HQ in Boston but more offices in Silicon Valley belong to?

I’ve got more comprehensive data on public robotics companies due to some updates suggested over at hizook.  However, I’m at a loss as to how to classify Brooks Automation and Cognex.  They both make automation components for various kinds of industrial applications and they both have corporate HQ outside of Boston with two offices each (probably the legacy of acquisitions) in Silicon Valley.

At a loss as to how to classify them, I’ve made a new category for them on my charts.  If you have thoughts about how to get good acquisition data–especially as a lot robotics companies can be acquired in a transaction that is ‘immaterial’ to a 10-K/Q for public company–I’d love to hear them.

And here is the raw data.  Not all market caps were taken on the same day.

Pittsburgh has a robotics meet-up!

Time to update the cluster comparison statistics, Pittsburgh has a robotics meet-up!  Join the AUVSI crew for some whiskey tasting.  I’ll be disappointed to be out in the Valley on Monday night.

http://www.meetup.com/Pittsburgh-Robotics/

East Coast Chauvinism in Robotics: Time to Face Facts, Silicon Valley is Kicking Our Ass

A cleaned-up version of this article became my first post on Hizook.  http://www.hizook.com/blog/2012/06/25/east-coast-chauvinism-robotics-time-face-facts-silicon-valley-kicking-our-butt#comment-971

_______

I have lots of love for Pittsburgh in particular, but it really pisses me off when people on the East Coast repeat a bunch of falsehoods (See #8) about how Boston and Pittsburgh compare to Silicon Valley and the rest of the world.  Many people in Pittsburgh and Boston—including people I call friends and mentors—smugly think that the MIT and CMU centered robotics clusters are leading the world in robotics.  This is demonstrably false.

If leadership in robotics means forming companies, making money, or employing people, then Silicon Valley is crushing everyone—no matter what the Wall Street Journal editorial page says about their business climate.  I’ve previously published an analysis of the Hizook 2011 VC Funding in Robotics data that shows that the Valley gets 49% of total VC robotics investment worldwide.

I’d now like to add an analysis of U.S. public companies (see bottom of the page).  Basically, the ‘Pittsburgh and Boston are the center of the robotics world’ story is even more ridiculous if you look at where public robotics companies are located.  Silicon Valley is crushing the other clusters in the U.S. at creating value in robotics and in building a robotics workforce in public companies.  (A forthcoming analysis will show that this true worldwide and if you include robotics divisions of public companies not principally engaged in robotics such as Boeing and Textron.)

77% of the workforce at public robotics pure plays is in Silicon Valley companies.  An astounding 93% of the market capitalization is headquartered in Silicon Valley and even if you exclude Intuitive Surgical (NASDAQ:ISRG) as an outlier, the Silicon Valley cluster still has twice as much market capitalization as Boston.

The public companies that I deemed to meet the criteria of being principally engaged in robotics, that they had to make and sell a robot, and not have substantial value creating revenues from businesses not related to robotics are listed in the table below.

The one company that I believe might be controversial for being excluded from this list is Cognex (NASDAQ:CGNX).  However, while trying to do decide on whether to include them, I found their list of locations.  They have three locations in California including two in Silicon Valley.  That means that this ‘Boston’ company has more offices in Silicon Valley than in Boston.  I’m not an advanced (or motivated) enough analyst to find out what the exact employee breakdown is, but combined with the fact that they make vision systems and supply components rather than robots, I elected to exclude them. I acknowledge that a similar case could be made about Adept (NASDAQ:ADEP) that just made a New Hampshire acquisition, but I have decided to include them and count them towards Silicon Valley.   I do not believe that either of these decisions, substantively impact my finding that Silicon Valley is the leading cluster when it comes to public company workforce and value creation.

I’m hoping the people who are spreading the misinformation that Silicon Valley has to catch-up to Boston and Pittsburgh will publish corrections.  I believe that this is important, particularly because I want to see Pittsburgh reclaim its early lead in robotics.  So many robotic inventions can trace their heritage back to Pittsburgh, it is a real shame that Pittsburgh has not used this strength to create the kind of robotics business ecosystem that one would hope.

It is impossible for communities to take appropriate action if they do not understand where they stand.  I hope that this new data will inspire the Pittsburgh community to come together and address the challenges of culture, customer access, and capital availability that have been inhibiting the growth of Pittsburgh’s robotic ecosystem before they lose too many more aspiring young entrepreneurs—such as me—to the siren song of California.

Company (1) Ticker Employees (2) Market Cap $M (3) % of Employees % of Market Cap Robotics Cluster
Accuray NASDAQ:ARAY

                   1,100

  463

20%

2%

SV
Adept NASDAQ:ADEP

                       183

43

3%

0%

SV
Aerovironment NASDAQ:AVAV

                       768

  577

14%

2%

SV
Hansen NASDAQ:HNSN

                       174

 135

3%

1%

SV
Intuitive Surgical NASDAQ:ISRG

                   1,924

  21,840

36%

88%

SV
iRobot NASDAQ:IRBT

                       619

  606

12%

2%

BOS
MAKO Surgical NASDAQ:MAKO

                       429

  1,110

8%

4%

Other
Stereotaxis Inc. NASDAQ:STXS

                       171

 13

3%

0%

Other
Total

                   5,368

24,787

100%

100%

(1) Companies are U.S. public companies that have been identified by Frank Tobe’s or my own research as principally engaged in robotics
(2) Employee Count as of Last 10-K Filing
(3) Market Capitalization as of 6/24/2012

Who are the top 20 academic roboticists?

In trying to compare the clusters, one of the most important and difficult to measure factors is the quality of the academic pipeline in each of the three clusters.  I thought about looking at patent filings, but that seems too hard and not truly indicative enough of what we are trying to measure.

A single lab, without a single patent could potentially blow the doors off company formation and economic impact in robotics.  I’d like to propose a different measure, the top 20 (or some other number) roboticists in academia… then lets see where they are and where their knowledge is creating value.  On the Pareto principle, we expect most of the useful output to be from the top researchers.  Also, I’d like to call attention to the fact that I don’t have criteria for what makes a researcher “top.”  I promise it is less trying to curry influence than the RB50, but I fully admit to not having the full insight especially into Boston, Japan, and Switzerland.

So here’s the start of my list in no particular order with blatant bias towards CMU:

Sebastian Thrun; Stanford; http://robots.stanford.edu/ Corporate Work and Spinouts: Google Car, Google Glass, Udacity

Red Whittaker; CMU; http://www.ri.cmu.edu/person.html?person_id=339 Spinouts: Red Zone Robotics, Astrobotics

Rodney Brooks; MIT; http://people.csail.mit.edu/brooks/index.html Spinouts: iRobot, Heartland Robotics

Henrik Christensen; Georgia Tech; http://www.hichristensen.net/ Non-spinout: National Robotics Initiative

Homayoon Kazerooni;  UC Berkeley; http://www.me.berkeley.edu/faculty/kazerooni/ Spinouts: Ekso Bionics (Formerly Berkeley Bionics)

Rich Mahoney; SRI; http://www.sri.com/news/expertsources/mahoney.html License Arrangements: Intuitive Surgical, others

Sanjiv Singh; CMU; http://www.ri.cmu.edu/person.html?person_id=290; Spinouts: Sensible Machines

Hagen Schempf; CMU; http://www.ri.cmu.edu/person.html?person_id=267; Spinouts: Automatika (Acquired by QinetiQ-North America)

Howie Choset; CMU; http://www.ri.cmu.edu/person.html?person_id=47; Spinouts: Medorobotics

Behrokh Khoshnevis; USC; http://www-bcf.usc.edu/~khoshnev/; Spinouts: Contour Crafting

Incubation in the Clusters

Once again, Silicon Valley is showing the rest of us how its done (see “Incubation” for the data).  Robotics only feels like it is poorly incubated in the Valley, because it doesn’t have incubators with multiple branches in the Valley like biotech and software do.  At least traffic sucks so bad in the Valley that when robotics gets going in the Valley it will need multi-branch robotics incubators just so people won’t have to drive.

All jealousy of California’s good fortune aside, robotics businesses are hard to start.  Not only do they have all the complexities of a software business (with a much more challenging test cycle), but they also have other parts that are equally challenging.  They are a hardware business, a manufacturer, and often a distribution or operations company as well.  I don’t see too many 22 year old college drop-outs running manufacturing and distribution businesses–they are too complex and require too much capital to just let them fail like a VC can do with a mobile app company.  Hence these kinds of companies are run by people who know what they are doing.  How do we create more entrepreneurs who ‘know what they are doing?’

For robotics to take off, we are going to have to find models that produce profitable companies with much less wasted capital than software venture capital does.  Incubation and mentorship are probably going to be really key to making this happen–good job to the Bay Area for getting on this.  If community leaders want to lay the foundation for something really extraordinary in their community, get a robotics incubator going in your community.

Cluster Activities (Continued)

The Massachusetts Tech Leadership Council is a really great organization.  I’m not sure how they get their members to pony up for the services that they provide (I’d like to know for my activities in Pittsburgh!), but having a professional cluster organizer like Elizabeth Newstadt and an organizational hub for promotion of the entire industry is fantastic.  I’ve heard that there are some frictions from the fact that the cluster crosses state lines and it is the “Mass TLC” as opposed to a New England-wide organization.  Still, the degree of organization that the cluster centered on Boston has is astounding.  A good deal of credit for this goes to the Mass TLC.  As an example, the survey they do of the robotic cluster is fantastic.  The other clusters should undertake similar surveys which would increase the value of Boston’s survey exponentially.

On the other coast, the San Francisco Bay Area is clamorous and still fairly ill defined–by which I mean there are a lot of people who may or may not be a part of the robotics industry.  Many robotics people think of themselves as being in the medical device industry, software, or electronic hardware–but not necessarily robotics per se.  On top of that, tons of people in the Bay who are not in robotics professionally provide the clamor and enthusiasm.  For example, all of my personal friends that build and fly drones for fun live in California.  I’m from back East, so the selection bias should run against the Bay.  They just love technology, nerdiness, and doing “your own thing” in the Bay–and robots fit the bill perfectly.  In fairly short order, I suspect that Andra Keay and the other folks behind the Silicon Valley Robotics Cluster and Robot Launch Pad will provide some of the rally flags to bring order to this energy–then the valley will be a sight to behold.  The Silicon Valley robotics people I’ve met think that their community needs to catch-up to Pittsburgh and Boston, but this probably only makes them dangerous since my data is starting to show that they are equal anyone.

Pittsburgh is a small community.  It is really great–everyone is super friendly and if you’re in robotics everyone knows everyone.  If you find yourself in Pittsburgh, I would be happy to introduce you to them and they will be nothing but good to you.  Things can happen really quickly because there is high degree of trust and community spirit.  My personal take on the robotics community in Pittsburgh is that there are things that need to be done collectively to get to the next level (VC education, a robotics incubator, more diversity of academic research, etc.).  The personal dealing model is going to be helpful, but not sufficient, to get the Allegheny robotics cluster to grow to the size that the region wants it too.  More formal organizations, supported by bottom-up enthusiasm for things like happy hours, meet-ups, and demos is going to be required for the Pittsburgh robotics cluster to scale.

Cluster Activities

How do we judge the quality of a robotics cluster’s activities?  The point isn’t actually to have a lot of meetings.  The point is to spark those interactions which can create or advance enterprises whether through ideas, collaborators, or resources.

Since we can’t measure that directly, I propose the Meetup.com test.  It is a crude metric, but should serve our purposes.  The idea is that the number meetup.com events coming up with relevant key words should serve as a rough proxy for how many people are out after work trying to create the next thing.

By this metric, Silicon Valley is crushing the rest of us.  Not only do they have more money, and better weather, but also they are out having more drinks.  Data is in the following the link.  https://robocosmist.wordpress.com/robotics-clusters/  Meetup.com test performed June 2, 2012.

Accordingly, San Francisco is the only place I’ve ever been served a martini made by a robot–by the Drinks Advanced Research Projects Agency no less.   Fortunately for getting girls back to my place, my gin martinis are still the best.  But the John Henry moment in robotic drinks is coming…

VC Activity

Travis Deyle at Hizook has some widely cited data about venture capital activity in robotics.  I’ve sorted his data to make a cluster comparison.

https://robocosmist.wordpress.com/robotics-clusters/

Observations:

  • Pittsburgh is further behind in the venture capital race than first appearances would indicate if the $25 M growth round from ABS Capital is not included in venture capital. No other deal in the data would qualify as “late stage growth capital.”
  • According to this data, the rest of the robotics world (not in Western PA, New England, or Silicon Valley) combined got $37.3M in venture capital funding– about 19% of total identified funding–or almost exactly what Boston got.
  • Silcon Valley got 49% of total identified robotics funding …and they have the best weather.  Not fair!