the young & the restless

Terry answers a question on technology entrepreneurship – understanding and embracing the difference between Raspberry Pi + AWS/Azure prototypes to commercialized products is critical to your success.

Speaker: Terry West  


In this podcast, I'd like to address a question I got from a fellow I've known for about six years. Naga is young Engineer, that I met right out of College. He impressed me right away. He's smart and very articulate. He is an embedded software, hardware designer. He sent me a question this week and said, "can you talk about Entrepreneurship". He said, "given that I'm a young person, I'd like to go off and do something, change something, do something big." It's a good topic. I have some interesting perspectives on it, given that I started my career at RIM BlackBerry, back in the 1980s, as their first employee. I got to watch Mike and Doug (the founders of RIM BlackBerry) grow their business over the first five years. Although I wasn't the entrepreneur, I would never claim that. Mike and Doug really were on the front lines trying to make payroll every month. I was only peripherally aware of those things, as I was a young Engineer, just doing the job. I loved it. I loved the joy of being part of a small company and getting things done every single day. Here I am, quite a bit older, doing the entrepreneurship thing again. I think I have some perspective on it, having started Serious and left my day job 10 years ago. I think we are at a time where Entrepreneurship is a huge opportunity. Actually, I'm more bullish on the opportunities for innovation and entrepreneurship now, than I have been in years. I would say this includes a combination of several factors.

First of all, as I mentioned a couple of podcasts ago when I talked about meeting a very smart banker. There's about two trillion dollars worth of cash coming back into this economy and it has to go somewhere. I think that cash is going to drive lots of mergers and acquisitions at the big company level. We just saw this past week, Microchip announcing they're going to acquire/merge with Microsemi, which is another large company that has been in business for a long time. This announcement came after the acquisition or actually, more like a merger, with Atmel, where they again, merged together two large companies. You can see this pattern with NXP merging with Motorola Freescale. The obvious first step is the merger of big companies. They add their revenue together, they cut the waste, that would be there if they left two complete organizations in place, or remove redundancies which improves profitability.

However, to me, while that's a normal and obvious thing for these large companies to do, I think of that as aggregating, not innovating. You take two companies, aggregate them, add the revenue, and if you're lucky, you get a sum of one plus one equals more than two (1+1=>2). Generally speaking, for the first several years, one plus one is barely two (1+1= ?<2), given the turmoil and churn between the two companies. I think I have some perspective on that, as I was part of the Intel team that was buying companies back in the late 1990s and many of those acquisitions didn't work out the way expected. Some of them drove innovations within the company but it was a challenge bringing in technology and innovating. I think companies like Intel and Cisco learned how to do that better over time. They became experts on how to bring in a smaller company (sometimes a very small company) in a way that it can actually impact the large company. I think that's really where the opportunity for innovation comes in for these large companies that are bringing back 2 trillion dollars into the U.S. economy. The challenge for those companies is how to take that money and grow the company. of course, they can certainly grow by acquisition, by aggregation of revenue streams. However, in the long term, at some point, when all the big guys buy other the big guys and they're all settled down, you get these behemoths looking at each other wondering what's next. There's nobody else to merge with, nobody else to buy at the large level. There is an inevitability that small companies will become a target for investment and acquisition by large companies. To me, it looks absolutely inevitable that we're on the verge, within one or two years, of entering an incredible phase like we haven't seen since the late 1990s. Hopefully, we don't end up in the same .com bubble that we had in the late 1990s. Nevertheless, I think we are going to see an incredible hiccup in funding, investment, then the subsequent acquisition of small companies, that are very innovative, can run faster and can drive growth for these larger companies. We've seen examples over the last 20 years, since the dotcom bubble burst, of a few things that were done extremely well from young people. For example, Facebook, Amazon and similar types of technology platforms. Whether you like them or not, they were innovative ideas that did extraordinarily well by themselves, in an environment where not many things were growing and innovating by themselves.

So, I would raise a glass to my young friend, who asked me this question, and say "hey, if you want to do innovation and entrepreneurship, I would heartily encourage you, as I think this is a great time". Given my experience with Serious, for the last ten years, as well as being a little older. I would caution that encouragement and provide a little guidance based on a couple of things that I've been talking to my friends about, over the last few weeks. In fact, a good friend of mine, Steve. Unfortunately, I can't name the company he works for, but Steve runs the I.T. department for a pretty large organization. He's really sharp guy and I find it fascinating talking to him. We were at the men's church conference over the weekend, eating very manly and probably horrible-for-you food (you know hot dogs, barbecue and all kinds of good stuff). Steve and I were sitting back when he began telling me about something he had been working on at the office that had more capability.

Steve now has the ability where a business unit, within his organization, can come to him and request to bring up a 40,000 node simulation of "this thing." You can think of it as an IoT set of nodes. They want it to have certain behaviors and want to see how those nodes are going to work together. In other words, how the hierarchy of the nodes and the management structure would work around that one firm. Steve said in the history of the universe he's never had the ability that he now has. This is not a multi-month planning cycle. This is something incredible. In a period of a couple of days, from the first request, thanks to standing start - He can go and write some lines of code that simulate each node. He can go onto Azure or AWS, light this stuff up, simulate 40,000 nodes simultaneously, doing all this stuff, all the management and all the things that they want to try. Then run the simulation for a few days, tear it down, as if it never happened, all within a period of a couple of weeks. I mean, it's just phenomenal what you can accomplish now with the scalability of the cloud and the off-the-shelf capabilities of Azure or AWS. Just phenomenal! We then started talking about Raspberry Pi, as well as, some of these other platforms. He's super intrigued with these things because the cost structure of those and their capabilities, even though they're educational platforms, make it very easy to do more than just simulations. Now you can create IoT devices, whip them up really quickly on a Raspberry Pi or a BeagleBone, demonstrate their functionality, connect it to the cloud, scale the cloud, and since he's part of a pretty big company, they could blow a hundred thousand dollars on raspberry pis and they would have rooms full of these things running as real nodes, not just simulations. Then tie them all to the cloud. He finds this whole concept of rapid deployment of prototypes for testing and all that goes on, fascinating.

Now, when I talk to Steve, he and I agree, even though he comes from the I.T. world and I come from the Embedded Industrial World. It's funny, that we very quickly got to this interesting conversation about Linux and about the vulnerabilities of Linux. He agreed with me, Linux is the most hacked platform on the planet, which is why they use windows in their organization and they are very secure. He also agreed with me the Raspberry Pi is not exactly a robust, commercializable platform, but for the sake of prototype, rapid deployment, and just getting something working - it's just phenomenal what you can accomplish. We got to the conversation of commercialization and the challenge of crossing that chasm between prototype and product. His company makes a lot of IoT devices of various kinds and also have the networks. Even though his approach was from an I.T. perspective, it was absolutely fascinating to me that his thought processes mirrored mine on the embedded industrial commercial side. We come from very different worlds but with the same thought processes.

This comes back to the entrepreneurship conversation. There has never been a time when we could easily create a device (any old IoT device) out of a dev kit from a one-chip company or slap together something with a Raspberry Pi and use AWS or Azure, write a few lines of code, scale it quickly (at least from a demo perspective), and show something to potential investors. That is just phenomenal. However, there is a massive difference between a prototype and a product. I guess that's one of the things that if you talk to angel investors and venture capitalists to invest in early-stage companies, one of the most important things that they're going to quiz you on is your commercialization plans. If you say you're going to take a Raspberry Pi and a few lines of AWS code, to a smart investor, they will say, "you are so naive, forget it." This is because there is a heck of a lot more to creating a product and calling it an application, than just writing a few lines of AWS code and slapping a few Amazon or Azure modules together. It would be like taking a demo platform, made of clone components, sold at cost, that you can't even get failure analysis done if there's a problem. There would be no consistency, year after year, in the product, or in the platform. Also, Linux has a very vulnerable security wise platform. You can't take something like that and say, "oh, I can just turn this into a product". However, some people have. We're starting to see the results of that. I think I mentioned in some of my prior podcasts the Chinese company that does webcams. Those webcams were based on Linux commercialized in that way, where you take a simple and Linux platform and you clone a raspberry pi or use it then turn it into a webcam, throw it out in the market. Yes, they made a ton of money; no question about it. Until the day when all of their webcams were hacked in the big Twitter attacks last year and their brand name was tarnished and left them so open. Of course, we've seen that in a few different cases, where companies have used what is inherent, extraordinarily, difficult to commercialize technology but great for prototypes, Raspberry Pi, Linux, BeagleBone. They just throw a few lines of code on AWS or Azure and say "that's my product" because they turned it into a product. We're starting to see things like government agencies say "no, no, no". You have to take responsibility for the outcome of your products if they're insecure or they're hackable.

I think we're going to see that continue to rise. The entrepreneurs that are really going to be successful, I think, are going to have, amongst many things, two really, really important focuses. Firstly, how do I use this cool stuff like Raspberry Pi like AWS and Azure, to create amazingly powerful, quick, and cheap proof of concepts? "How do I do that, with very few resources, very few heads and not a lot of money" because you can't spend a ton of money getting to that first investment phase when starting to scale your business. At the same time, the successful entrepreneur has to focus on commercialization. The entrepreneur has to understand that what they've built with slapping a few lines of code together in the cloud and using an educational platform that was never meant to be a product. The distance between that and a commercialized product has to be understood. There is a way, there is a path and the path is not just by a commercialized Linux module, the path is much more related to the supply chain and the kinds of things that are sold into that environment.

Your idea can be innovative, your approach can be disruptive. Your product can win where many, many big companies have gotten stuck in the mud for the last 17 years. The challenge is you still have to live in the environment that these big companies sell into. Cisco does not release products that are flaky. Apple does not release products that are made with clone Chinese components. You have to participate in those markets, whatever that market is, whether it's an IoT node or whatever it is, with products that have the right characteristics. If you're selling into the industrial market or industrial kitchen market or commercial environment, you know maybe revolutionizing an HVAC controller system for large buildings. You can't sell unreliable product into that market. You won't get bought and you'll be out of business before you know it.

Your commercialization plan has to say, okay what technologies are available, that like the Raspberry Pi, like the few lines of AWS, can get me to market quickly. Are there modules, are there off-the-shelf solutions, are there partners? For example, Arrow Connect, the folks at Arrow have wrapped an enormous amount of intellectual property around AWS and Azure. They can build an incredible, scalable, secure platform including provisioning and user management, rights management, plugins, analytics and all that stuff and they build it on top of AWS and Azure. In fact, you can actually run it on AWS and Azure, simultaneously. Also, the application that you prototyped can be dropped into that environment and worked into something that is actually, incredibly robust and secure. However, you have to find a partner like Arrow Connect. This way, once you get your funding, your initial funding, they will make that transition for you and create a commercialized cloud presence.

Similarly, on the device side, whether (for example) you are using an HMI with a BeagleBone and you want to create something with a touchscreen. That's where, Serious, for instance, as well as others, can come in and provide industrial, robust technology for your HMI or your IoT coms and control, on the embedded side, on the device side.

You have to have a map and a clear understanding of prototype, which is great to do with these cool new technologies, to commercialization. I think that's where a lot of the big companies fail. They either think the prototype can be commercialized and then two or three years later they realize what a horrible mistake they've made or they're so fixated on the limitations of the commercialized products they have today that they dismiss the value of the rapid prototyping and concepts that are available with some of this off-the-shelf cool stuff like Raspberry Pi, AWS, and Azure. I think, here you are, a young entrepreneur with something unique, you understand and embrace this cool new technology of AWS, Azure, Raspberry Pi, BeagleBone. These cool rapid-prototyping capabilities.

Now, here's your challenge. Go build those things, with as little money as you can, to show off the capability of what's possible. Then map out the commercialization aspects with things that meet the needs of the target market, with real partners, with real capabilities, that you can actually sell into those target markets that will eventually get you bought for lots of money. Once you create that map, you'll have a recipe for success.

Return to Podcast Main Page


embedded technology blog

Visit Blog



captain's log embedded podcast


The challenges of HMI and IoT can be overwhelming. Listen weekly to our Captain – Terry West, CEO – on industry trends he sees and implications for the embedded systems designer. 

You'll find practical thought leadership conversations covering all facets of embedded technology, human machine interface evolution, architectural challenges, IoT readiness, and much more.


Podcast General App Image-100.jpg

You can now stream the Captain's Log on Podbean, iTunes and Google Play Music!

Available on Podbean.jpg


Listen on Google Play Music