Peter Semmelhack

Peter is a card-carrying geek with the pictures to prove it. Having spent the last 18 years of his career in software, Peter suddenly rediscovered hardware two years ago in a bid to relive his Heathkit days. He is a very happy husband and father of two wonderful children.

April 2nd, 2013

Artificial Scale

There is growing developer interest in what is popularly called the Internet of Things (IoT) – in essence, anything connected to the Internet that’s not a laptop, tablet or smartphone.  As the price of microcontrollers, CPUs and networking chips continue to fall, it becomes easier to incorporate them in just about any type of product.  3-D printing has made it possible to easily construct physical assemblies and Kickstarter has helped fire the imaginations of entrepreneurs everywhere.  And since all these connected devices are useless without related applications and services, the developers of the world are taking notice of the opportunities.

But the simple-to-understand “Internet of Things” phrase, which seems to imply a huge new world to explore, belies a major obstacle to its success — scale.  Regardless of what you read in the media – billions, trillions of connected devices! – there are currently few, if any, markets ready-made to buy millions of new devices, let alone applications that could run on them.  As a result, large scale IoT installations that would be of interest to independent developers do not exist.

Unlike the desktop and smartphone markets which are comprised of millions of users all working on similar hardware/software platforms (think WinTel and iOS), the IoT markets are extraordinarily fragmented.  Not only are there dozens of verticals – health, retail, energy, home, etc. – each one is a collection of dozens more.  For example, health can be further broken down into fitness/wellness, drug compliance, disease management, telemedicine, remote monitoring, and many more.  The result is a huge conglomeration of highly specialized markets needing equally customized technology and applications to support them.  There are currently no large IoT markets.  There is no scale.

The absence of scale obstructs the most critical building block for attracting world-class developers — a clear path to revenue.  Today, if I write an application for iOS, I know that, if I’m successful, I could potentially sell it millions of times.  It’s simple arithmetic to figure out how to get rich in this world.  But absent that large installed base of users to sell to, developers are unlikely to devote the resources necessary to build, deploy and support the killer apps needed to drive adoption.  There’s a reason Angry Birds came out on iOS first.

One way to solve this problem is via something I call “artificial scale”.  Network technologies, by their very nature, provide a layer of abstraction that helps smooth communication between nodes.  For example, when I use my iPhone to call my friend in London, there is no need for it to know that it is traversing two carrier networks and establishing a connection with a Nokia Lumia running Windows Mobile.  The network handles all the details.  Likewise, when I point my browser at a website, it does not care whether the web server is running Linux, Windows or OSX.  The network handles all the interfacing.  So, networks are good at hiding hardware heterogeneity.

Let’s apply this idea to a typical IoT market vertical – say telematics.  There are currently over 250 million cars on the road in the US.  That’s a lot of vehicles.  As a developer, if I could sell my $0.99 app to some percentage of that population I could make a good living.  But the problem is all those vehicles are divided up by manufacturer, model, year, etc.  It’s fragmented.  But if I could somehow get all those automobiles on a network, I could use that abstraction to make them all “look” the same to an application running elsewhere on the network (e.g. the cloud).  If I could make all the Ford Fiestas, Chevy Impalas, Toyota Camrys and Nissan Sentras all look like one type of car then I could potentially lure the developers of the world to pay attention.  This trick – using a network API as a way to combine and aggregate fragments of an IoT vertical market – is what I’m referring to when I say “artificial scale”.  Applications running in the cloud can now access potentially huge bases of heterogeneous hardware platforms using networking technologies to handle the abstraction.  The market now has scale, albeit at the network level.

This approach is possible today and represents a potentially disruptive new way to approach IoT markets.  The alternative is what we’ve been seeing now for many years –proprietary, vertically integrated, winner-takes-all, solutions.  For example, Nest may win the war for the home.   Maybe Nike+ will dominate the world of connected footwear.   But their market strategies are old.  I’m suggesting there are new ways to achieve similar market success using more open, community-based approaches to innovation.   The concept of “artificial scale” may be one such opportunity.

Bug Labs develops a cloud-based IoT application platform called Swarm that helps enable artificial scale and companies are using it to build developer ecosystems that will provide a single API set for collections of heterogeneous connected hardware types.  Click on the link to learn more and stay tuned for some exciting news coming soon!

October 12th, 2012

The Future of Cable Innovation?

Comcast logo - square

Working with huge companies like Comcast has many advantages for a small firm like Bug Labs.  But one of the downsides is they typically like to keep innovative new ideas and initiatives close to the vest.   Their dead serious interest in IP and the institutional protections they place on it permeates everything they do.  It’s not a problem normally – unless you like the idea of openness (as we do) and prefer involving the community in the innovation process.

For us, our most influential and meaningful marketing results come from positive customer references.  Getting the word out on a successful project is HUGE for us.  So I am very excited to finally get the chance to unveil the great work we’ve done with Comcast over the past year.  Check out this link to read all about it on Comcast’s web site.

In a nutshell, Comcast used the BUG system to build and launch a modular, product development platform for 3rd party developers called CHIP – Comcast Hardware Innovation Platform.   It’s a unique new way for developers to design, develop and deploy applications that take advantage of some of the unique qualities offered by the Comcast platform.   Forget everything you think you know about Comcast and how they perceive the developer community.   Based on the success of CHIP this summer it looks like there will be plenty of incentives – both from Comcast and from the strength of the platform itself – to get developers excited.  In my opinion, every network carrier in the world, wired or wireless, will need to do something similar.  Capturing the hearts and minds of the best developers who can then bring additional value to the platform via applications will take on existential importance in the coming years.  We are thrilled to be helping Comcast explore the possibilities.

May 23rd, 2012

TechCrunch Disrupt NYC!

Screen Shot 2012-05-23 at 7.15.43 AM

We’ll be presenting and demonstrating our gear at the TechCrunch Disrupt NYC 2012 conference today.  Last time we did this was with TechCrunch Disrupt in SF last September with Ford showing off our OpenXC platform.  It was a great event.  TCD NYC is hosting a “Hardware Alley” in response to the growing interest in DIY and open source hardware, especially in the Big Apple, where there are a number of interesting start ups, hackers groups and university interest.  The venue is located at Pier 94 on 12 avenue at 55th st. Come down and check it out!

Update – here’s some coverage on the panel I was on.  Great event!

January 16th, 2012

2011 in Review

CES 2012 logoCES always marks a point of reflection for me – a time to look back at the prior year and see how we measured up.  It was at CES 2008 that we officially launched Bug Labs and won the CNET “Best of Show Award for Emerging Technology”.  It was a heady week.  It’s hard to beat that performance.  But it’s always fun to go back.

When I look back over 2011 and review our accomplishments, I feel good about how far we’ve come.  We felt the sky was the limit four years ago and we still feel that way.  Open source hardware and the Eric von Hippel’s concept of “democratizing innovation” have become much more mainstream.  Joi Ito recently said as much in an interview.  We’re proud of our participation in these movements and look forward to continuing to actively contribute.

2011 had some very memorable achievements.  As a company, our revenues again grew 100% year over year.  We added a host of new customers and struck some important partnerships (Arrow, Pitney Bowes).  We were invited to display our unique modular hardware designs in the MoMA’s “Talk to Me” exhibit, which was a particular thrill for me personally. We hosted the 2nd Annual Open Hardware Summit in Queens to another sell out crowd.  As a car lover, the announcement that we signed a Joint Development Agreement with Ford Motor Company to start building an open source hardware and software-based automotive platform was a dream come true.  Lastly, we formally announced BUGswarm, our new cloud-based IoT application platform, and launched our official Developer Preview.   Great stuff.

I believe 2012 will be a great year for us.  In addition to some very cool projects that we’ll soon be talking about publicly, we have a bunch of other announcements coming up so stay tuned!  Looks like we’ll be very busy :)

October 18th, 2011

Social Machines

connected imageIf you’re like me, you start to reflexively roll your eyes when you see some new, previously unmolested noun get saddled with the term “social”.  We have social everything these days.  And the trend doesn’t seem to be going anywhere.  The reason I think, though it doesn’t lessen the annoyance factor, is that the term does convey a new type of meaning, a new way of gauging the value of something, usually in relation to something else.  It’s one thing to say it’s “connected”.  It’s something different (better?) when you describe it as “social”.   At least, that’s the goal.  If you believe Gartner, social is one of the big trends for 2012 – http://www.zdnet.com/blog/btl/mobile-big-data-social-dominate-gartners-top-tech-trends-for-2012/60991

Lately, we’ve been using it in our battle against the antiquated term “M2M”.  Maybe not surprisingly, it’s been helping.  At the very least, it helps to re-orient the discussion around where we see a whole new class of applications.  Traditionally, M2M applications have focused on helping organizations become more efficient.  The goal was to save money and resources.  The ROI was based on reducing costs.  These types of applications will continue to drive a significant percentage of activity in the years ahead.  But what we see happening more and more is interest in M2M-type applications that help companies *make* more money, not save it.  The ROI is top-line based.

There are a couple reasons why this trend is interesting to us:

1 – The budgets for these types of projects come from marketing and sales organizations, not operations and IT.  This usually equates to faster time to results.

2 – The technical skills necessary for building the apps for these types of opportunities are, in many cases, employed by the marketing department itself –> web developers.  This means there is less reliance on corporate IT staff which means faster execution.

3 – Both of these factors help companies do what’s absolutely necessary these days – innovate continuously, rapidly and at low cost.

What are some examples of this type of application?  Think about “social” vending machines and kiosks.  Think about social rental systems – cars being a great current example.  Think about social health devices.  The list goes on.  The point is – we’re not talking about remotely monitoring a coffee maker in a 7-Eleven to detect the need for maintenance.  That’s old school M2M.  The future belongs to something more interesting – social machines.  And perhaps, finally, a way to incorporate their existence, and the information they hold, into our lives more seamlessly and meaningfully.

August 18th, 2011

Machine2Cloud

I have a deep seated ambivalence for the term “M2M”.  I see it as an anachronism, too rooted in the past to be useful to us now.  But today everyone is using it to describe all kinds of markets/technologies.  Between the term’s history and the current hype I can’t help but think it’s confusing the heck out of everyone (including me).   So I conjured up the term “Machine2Cloud” (M2C!) while thinking about decent alternatives though I’m not suggesting it’s much better.  But at least it has in its make up the components of what I think is most exciting.

Salesforce.com came to market in 2000 with a (cloud-based) $999 package that competed admirably with the then best-in-class, behind-the-firewall Siebel SFA system costing hundreds of thousands more.  Even better, there was nothing to install, no IT manager to contact/coerce, no boss to convince.  It was easy to try with little to no risk if it failed.  It made the SFA experiment affordable.  The “I” in the ROI equation was very low and potential “R” disproportionately high.  It had the added benefit of making the sales manager look like a hero to his/her boss.  Salesforce.com used this approach spectacularly well.  They are now a huge, successful business and Siebel is no more.

The same approach that brought the value of SFA to the world’s millions of sales people, can now be used identically to bring value to the growing millions (no, trillions! thx Ericcson) of net-connected machines around the globe.  “Machine2Cloud” applications will bring low cost, low risk, high impact solutions to organizations that may have eyed and admired the perceived benefits of real-time information but were concerned about costs.  M2C architectures will help take the pain out of experimenting and trying new things.  It will provide the opportunity to scale quickly without the attendant, risky investments in IT infrastructure and IP licensing.  In short, it will bring all the benefits of real-time information – namely better, more timely decision making – to anyone interested in giving it a shot.

Supporting my belief that M2C will soon get a lot of attention is the rapid increase in interest of the world’s wireless carriers in M2M.  All the US providers have organized dedicated M2M sales units.  The major carriers in EMEA, Latin America and Asia are also announcing programs.  Why?  It’s simple.  They have to sell significantly more data at substantially higher rates of growth (if they’re to keep Wall st. happy) and the M2M/M2C worlds are a natural.

Low cost, easy to setup/administer wireless access to machines, I strongly believe, is a cornerstone for next the wave of M2M innovation.  M2C is a natural derivative.  Imagine setting up a complete telematics solution in a week instead of 3 months.  Or installing a group of innovative POS kiosks or vending machines over a weekend.  The tools coming available today allow for a literal “drop & go” approach.  This will radically change the landscape for M2M/M2C innovators.

This type of thinking fits neatly into what I suggested in my earlier post.  M2M will get absorbed into the more general, and meaningful, enterprise apps space.  Which, frankly, is better for everyone in the industry, and most importantly, more valuable to customers.

August 10th, 2011

Mobile M2M is the Next Wave for Mobile Enterprise

A few years ago I started a company called Antenna Software.  For those of you old enough to remember, the Mobile Internet was going-to-be-huge in 2000…  Eleven years later it’s now starting to live up to that promise and Antenna has benefitted from the upswell.  Antenna focuses on what’s called the Mobile Enterprise, a market category that is roughly comprised of mobile professionals (sales, service, etc) using smart phones, tablets and other wirelessly connected devices to conduct business, securely, in the field.   The revolution was started by the RIM Blackberry + email and has mushroomed to include just about every type of enterprise app you can imagine.  It’s big business now and getting bigger quickly.   The reasons are simple.   Mobile enterprise apps help get the right information, into the right hands at the right moment which helps improve customer responsiveness, worker productivity and operational efficiency.  The ROI is bulletproof.

Now let’s look at what I call Mobile M2M.

With the success that has come with enabling mobile workers with real-time information has come the realization that the same ROI benefits could also be collected from other types of mobile “assets”, for example cars, trucks, locomotives, golf carts, bicycles, shopping carts, etc.  A system that tracks things via a network connection is usually categorized as telematics and is a classic M2M market segment.  And this category is experiencing significant interest and growth right now, which makes complete sense.  Vehicle efficiency, safety, compliance, etc. is of vital interest to the business managers who operate them.

Keep in mind that “mobile” doesn’t have to mean constantly-in-motion.  It could be a kiosk that gets moved once a month.  Or a digital signage system that promotes a weekend sale and is then relocated to another part of the store.

The point is, everything we’ve learned from empowering mobile humans with the ability to produce/consume real-time business data is exactly applicable to mobile machines.  And rather than try and create a new way of defining and conceptualizing M2M as something *new* I think it is far easier to see it from the eyes of a business owner/manager/partner –> it’s an extension of an already well-understood, ROI-centric revolution called Mobile Enterprise.  Mobile Enterprise and Mobile M2M are the same thing because that’s how managers run their businesses.  The differences are purely technical.

This is good news for potential users because it becomes part of a seamless value proposition.  It’s good for M2M vendors because using this context helps fit solutions, that may have been pigeon-holed as a separate silo into a mainstream, well-understood and valued application category.

August 2nd, 2011

Top Ten M2M Misconceptions

Screen shot 2011-08-02 at 3.49.14 PM1 – It’s new – No.  The acronym “M2M” is at least 30 years old and so is the concept.  The good news is – it has a history of solid business value.
2 – It’s a market – It’s not.  It’s a heterogeneous collection of hundreds, if not thousands, of vertical niche markets (worldwide) all with their own requirements, costs and ROI models.  It’s not unusual to find M2M markets measured in thousands of users/units, not hundreds of thousand or millions.
3 – It’s gonna be huge! – Only in the aggregate.  But individually, there will be lots of small wins; some very small.
4 – The wireless carriers are easy – I wish it were true.  And while they are certainly improving, the carriers today represent a stubborn barrier to innovation in this space.  More on this in my next post.
5 – We just need a killer app. – Bad news guys.  There will be no #1 killer app; no “Angry Birds” of M2M.  But there will be lots of niche killer apps.
6 – The big guys are going to dominate the market – Nope.  Because there is no mass market opportunity the scale economies that come with being big do not apply.  Cisco, Philips, Ericsson and their ilk may think they will but they can’t.  M2M is not a winner-takes-all (ala St. Jobs) opportunity.  Speed to market and rapid, high quality, niche-focused product development will rule, not just price.
7 – There is revolutionary potential – Perhaps.  But I believe this is a critical evolution of two much more mature, longer term trends – mobility and mass customization.  First, everyone now in the mobile space is looking for the next green field and M2M is it.  And just about everything learned from mobilizing humans will apply to machines.  Second, the highly fragmented M2M market space will become the poster child for how new, bottoms up, open product platforms, technologies and movements (hi open source hardware!) can help drive significant levels of innovation at every level – hardware, software and cloud.
8 – Build it and they will come… – Everyone remember RFID??  It’s easy to believe that some of the new technologies becoming available will feed the fire.  But the opportunities now presenting themselves are not technical.  They are pure business.  New whiz bang technologies are not going to help anyone here.  We need to put in the hard work of demonstrating real business value and ROI.  Luckily there’s plenty to be had.
9 – We’re reaching a tipping point – Not yet.  I think 2012 will see real progress and some very public “wins”.  These will start to lay the foundation stones for even more rapid growth.  The tipping point will come once we’ve reached a critical mass of these proof points.
10 – The acronym “M2M” is sufficient – No way – see Point #1.  The term is antique and we desperately need some new lingo to adequately describe the opportunities becoming available.  In addition, the term M2M has substantial historical baggage (think POTS lines) that we need to jettison.  I wish I had a great answer but for the moment will leave it to others far wittier than I (though I’ll keep thinking about it!).

June 15th, 2011

Bug Labs + Arrow Electronics – This is big.

Screen shot 2011-06-15 at 7.42.20 AMToday we announced an exclusive, new partnership with Arrow Electronics, (NYSE:ARW) an $18B, worldwide electronics parts distributor and solutions vendor.  It would be tempting to file this under the “no big deal” category, but you’d be making a mistake. This partnership could have far reaching implications for not only Bug Labs but for the Open Hardware movement and makers in general – here’s why.

Hardware is hard.  It’s a reality that our customers confront every day.  The fact is, working with atoms is far more problematic than fiddling with bits.   As a career software person, this basic fact continues to frustrate me.  I dream of the day when the gulf between the two, from an ease-of-innovation standpoint, collapses.

One of most basic issues with atoms is that their supply, by definition, is limited.  When it comes to electronics, the atoms that go into making all the components needed to build interesting things, are in vastly shorter supply.  We’ve all experienced it as consumers – six week waits for [fill in the blank popular gadget].  The flip side is equally troublesome.  Inventory, shipping, customs, accounting, etc – everything you need to worry about when it comes to selling your limited supply of atoms-as-gadget is a serious pain in the neck.  Trust me.

These special demons of hardware innovation – finding, buying, trading, storing, shipping “stuff” – stop many would-be Steve Jobs’s from ever getting out of the starting gate.  Wouldn’t it be nice if it were easier, more transparent/visible and affordable?  Wouldn’t it be cool if the whole process of going from hardware prototype to selling production devices looked more like the Apple Appstore model, where developers spend time developing and Apple handles the rest?

Yes it would –> enter Arrow Electronics.

Arrow is to BUG’s community of hardware innovators what Apple is to its software developers. They offer the supply, inventory management, manufacturing and worldwide distribution services needed to take a BUG prototype and make it a BUG-based production unit – in a fraction of the time and cost of “normal” hardware innovation processes.  You can now do in weeks what used to take months or even years.  They’re able to do this because they understand and support our platform across all our customers and that scale (horizontal not vertical) provides benefits to everyone.

Bottom line – You spend your time building and supporting your device/application.  BUG provides the IP (hardware, software) and Arrow provides the rest.  It’s a compelling combination – for everyone from individual innovators to Fortune 500 corporations.

Stay tuned for further updates in the very near future.  Everyone here is excited by the potential. Please let me know if you have any questions.

March 14th, 2011

Bug Labs + Pitney Bowes

PB logo_high_res_smallWe are thrilled to announce today a product we’ve been working on with Pitney Bowes for over a year now – a new BUGbase, we’re calling BUGsecure, that includes a special, tamper proof security chip.  We’ve issued a full press release here, but in summary this new chip, from Pitney Bowes, combined with their enterprise-class key management services and the Bug System gives our customers a completely new, high-security, trusted platform with which to prototype, pilot and produce innovative new wireless devices, services and applications.

The Pitney Bowes chip we’ve incorporated is the same one that lies at the heart of every Pitney Bowes mailing/postage machine whether it’s a desktop unit or a room size system and secures over $5B in postage transactions every year worldwide.  It’s called a tamper-proof, cryptographic key store.  In essence, it provides an ultra high security source for encryption keys with which to sign data.   This type of signature is essential for applications where the risks of fraud are high and/or the need for a bulletproof audit trail is required.  For example, many medical applications (think pharmaceuticals), financial, regulatory and safety use cases require this type of security.  Software-level security frameworks and protocols, while suitable for many uses, do not provide the level of certitude that a hardware key store does.

I heard a security expert recently say that “the mobile world today is about as secure as the desktop was ten years ago.”!  With so much attention being placed on mobility and empowering individuals and machines with new information and processing power, the need for the highest levels of security is exploding.  We’re excited to be working with Pitney Bowes on this exciting new opportunity.

We will be showing off the new system at CTIA in Orlando (booth #3659) all next week so if you’re attending the event please drop by and say hello!