Meet the 'How to Start a Start-up' panel for ESC Boston 2016 - Embedded.com

Meet the ‘How to Start a Start-up’ panel for ESC Boston 2016

As a follow-on to my previous column on this topic — Learn how to start your own start-up company at ESC Boston 2016 — I thought I'd share a few more details about the members of the panel who will be offering their hard-earned insights and answering questions from the audience.

First, let me introduce you to the members of the panel as follows:

  • Yours Truly (the moderator & ex-CEO of TechBites Interactive)
  • Jack Ganssle (Embedded Guru)
  • Jean Labrosse (CEO Micrium)
  • Jonny Doin (CEO GridVortex Systems)

After I've kicked-off the session and introduced the other members of the panel, we are each going to take a few minutes to share some of the topics we feel to be particularly important when you are founding your own start-up. After this we're going to open the floor to any questions.

Just to set the scene here, I thought it might be interesting if we — the members of the panel — were to share some background information about ourselves, including some of the topics we'll be discussing. So, in alphabetical order of first names (which, by some amazing quirk of fate, puts me on top), here we go…

Clive “Max” Maxfield (Trendsetter and leader of fashion)
I graduated with my BSc in Control Engineering from Sheffield Hallam University in 1980. My first position was at International Computers Limited (ICL) in Manchester, England, where I was a member of a team designing CPUs for mainframe computers. This was where I designed my first ASIC — it contained only 2,000 equivalent gates, and the design was captured at the gate/register-level using pencil and paper.


Yours truly circa 1980 (Source: Max Maxfield)

I subsequently moved into digital logic simulation and gained expertise with a wide variety of digital design tools and methodologies. Eventually I moved to America and joined a company that — based on my digital expertise — made me their analog marketing manager.

In the year 2000, I and two friends founded our own company — a high-tech marketing consultancy called TechBites Interactive. After about 10 years we had a brilliant idea and put all of our money into it. Sad to relate, it turned out to be a not-so-brilliant idea and we came to the point where we had to turn out the lights and close the doors.

My two partners got real jobs, while I struck out as a freelance consultant and writer. After a few years of this, I was offered a position at UBM (United Business Media), and I now find myself an Editor at EETimes.com, Editorial Director of Embedded.com, and the Technical Content Director of the Embedded Systems Conference (ESC). Meanwhile, in my spare time… (LOL)

Some of the topics I plan on mentioning in the How to Start Your Own Start-Up session are as follows:

  • Make sure both the company name and website URL you want are available before you spend time and money registering one or the other.
  • Get a real graphics guy to create your logo — not your brother's friend's girlfriend who considers herself to be an expert with PowerPoint and who creates something that looks great in color, but that changes to a big gray smudge when photocopied in black and white.
  • Make sure you look like a real company; for example, ensure the website, business cards, letterhead, PowerPoint templates, etc. all reflect the same look and feel (ideally use the same graphic artist from the previous point).
  • Make sure you act like a real company; for example, have ethics and suchlike policies in place from the start, like “No pirate software allowed on any of the firm's computers.”
  • Plan for long-term stuff right from the beginning. What's your exit strategy? What happens if one of the partners decides they want to sell up, or passes away?

Jean Labrosse (CEO of Micrium)
I fell in love with the idea of being able to program devices when I was in college. I was lucky to be there when 8-bit CPUs made their appearance (e.g., 8080, 6800, Z80). Back then, tools were almost non-existent; all you typically had was a Hex keypad, a hand assembler, and a cassette tape upon which to store your programs.

I was studying Electrical Engineering and I did a lot of digital electronics designs (TTL, CMOS, etc.). Microprocessors were a Godsend and I took a lot of classes from the Computer Science department (e.g., structured programming, data structures, high-level languages).

I started my career as an Embedded Engineer. My First job was in medical electronics designing things like 6801- and 8049-based T.E.N.S. units.


Jean Labrosse in a funny hat competition circa 1991 (Source: Jean Labrosse)

Later, I moved to California to work on industrial engine control systems for four years. This is where I was introduced to my first RTOS running on a 68000 CPU. Then I moved to Florida for 15 years, again working on industrial engine control systems.

On one of these projects I ended up using two different RTOS kernels. I started with the first because it was cheap, but basically it didn't work and I didn't have access to the source code. So I moved to a more expensive kernel and ran into a bug. Once again there was no source code and the support was terrible.

Eventually, around 1992, I created my own RTOS called µC/OS. This was free to use and the full source was included on a 3.5″ floppy disk. This was followed by µC/OS-II in 1998, at which time I started selling the rights to use the kernel in commercial applications while still supplying the full source on a CD. In 1999, I decided to quit my full time job and started Micrium. We released µC/OS-III in 2008. We bootstrapped ourselves and self-funded, and we've been profitable since our inception.

Some of the thoughts I'll be sharing at ESC Boston are as follows:

  • Have good principles — honesty and ethics — and develop for others, not for yourself.
  • Treat your customers as you'd like to be treated; listen to their needs and provide outstanding customer support.
  • Hopefully, you're not just after the money and/or the fame. I wanted to help the embedded industry; I wanted to provide quality products to the industry; and I wanted to share my knowledge instead of hiding it.

From the beginning I wanted to enjoy what I did, but I should also note that as your business grows, you find yourself doing less of the things you like and more of the things you must.

Jack Ganssle (Embedded Guru)
Jack started in the embedded business when the first 8-bit processors came out, which is a polite way of saying he's getting really old. Never a particularly reliable employee, he had to found several businesses to maintain a fiction of gainful employment. In 1981, he started a consulting business with his friend Scott. This business made a lot of money for everyone except the two founders, but they had a lot of fun and got involved in some amazing projects, including the White House security system.


Jack (right) and Scott (left) circa a few hundred years ago (Source: Jack Ganssle)

The above photo shows Jack and Scott a few hundred years ago enjoying a turkey rather than working on a Nova-minicomputer-based spectrometer they designed. Note that — as you can see in this image — the entire world back then was a much fuzzier place than it is now. The turkey was delivered by Jack and Scott's girlfriends who had grown despondent over the course of this four-day non-stop debugging session.

After a period of time, the partners had a falling out and elected to sell the business in order to preserve their friendship. Jack went on to found another consulting company, but that quickly morphed into a product organization. He wrote and sold a line of multitasking BASIC compilers for embedded systems (really! They were used on the Space Shuttle, even).

Eventually, the company designed and sold a number of in-circuit emulators for many different microprocessors. It grew to be one of the larger embedded tool companies with many (sometimes too many) employees. One day a friend came over and showed Jack the World Wide Web which was brand spanking new, and in a “Holy crap, this is cool!” reaction he started another company which became one of Maryland's first ISPs. Modems, T1 lines, ISDN connections, and UNIX boxes proliferated with wild abandon.

After 15 years of 70-hour weeks, Jack sold the two businesses and tried retiring, but after just one day he was bored to tears, so he started lecturing about building embedded systems, which turns out to be a lot easier than actually building them. On the basis that Jack never knows what will come out of his mouth until he opens it, the topics he may or may not touch upon may or may not include the following:

  • The profit motivation.
  • The follies of a partnership.
  • Working on — as opposed to in — the business.
  • Financing a startup.

Jonny Doin (CEO GridVortex)
I started working at 15, as a Z80 assembly programmer. My mother was getting nervous about the size of my hair and the volume of the Pink Floyd playing in my room, and when she saw the voluminous Leventhal's Z80 Assembly Language Programming book I was reading, she went through the classified ads in the local paper, found one that said “Hiring Z80 Assembly Programmers,” and sent me there. She also cut my hair short and made me “suit up” (I never gave up Pink Floyd, though).


Jonny's hair (top) and Jonny (underneath) deep in the mists of time
when he was 15 years old (Source: Jonny Doin)

I was very fortunate to work for several years at Hewlett Packard doing Embedded Design, UNIX programming and Pascal working on HP 1000 and HP 9000 systems. I worked at the Scientific Instruments Division and travelled the world solving instrument problems and being part of Product Design teams. It was a dream job. From HP, I went on to become VP of Technology at a Multimedia company that distributed broadcasting video editing equipment designed by a German startup, FAST Multimedia GmbH.

When I realized that my career was straying away from hands-on design, I left the company and took a sabbatical year to restart work on Embedded Systems. I got as many microcontrollers I could get my hands on — including PICs, 8051s, and Motorola devices — until I fell in love with the ARM Architecture. And then I got back working as an embedded engineer for a small biomedical company followed by an industrial instrumentation company.

I started my first company, Synapse Designs, in 1999 manufacturing power substation automation equipment. We installed over 6000 systems and automated the switch breakers of one of the largest cities in the world — the city of São Paulo. There, I learned the true meaning of concepts like “mission critical” and “liability.”

Unfortunately, that company went bankrupt after we lost our largest contract to another “friendly” company with which we had enjoyed a very close partnership. One lesson learned here was: “Do not hold on to an expensive payroll when you lose contracts — learn to let people go early.”

I started my current company, GridVortex, in 2013. Based in Sao Paulo, Brazil, the company designs IoT smart infrastructure systems. As soon as I started writing our core code, I saw that strong Cyber Security was the largest flaw in the smart infrastructure, so I set on to design Embedded Systems with proper crypto security, all in Bare-Metal. Today, our framework provides a robust codebase for fast applications that need Security and Safety from the bones up.

We organized GridVortex as a core holding company with compartmentalized business units. We were fortunate enough to attract highly-capable Board Members who help us with high-level advice and guidance. Thus far, we have bootstrapped ourselves all the way. My next mission is to bring GridVortex to the United States. Being the CEO, I have had to learn many new things, including how to talk to bankers, how to compute every tax and invoice, and how to fire people, but I still do hands-on embedded design, chip design, and hardware and firmware. I love this work!

Some of the topics I'll be touching on at the How to Start Your Own Start-Up session at ESC Boston are as follows:

  • If you are starting on an idea, or an application of technology, have a working, tangible proof of work — a working prototype or a working mathematical model — go beyond PowerPoint.
  • Work work work work work. You will learn the true definition of hard work.
  • Try to keep a core of talented people, especially in technology development. Having a group culture helps in creating and keeping direction.
  • Embedded development is extremely hard to assess and maintain. From the beginning, do peer reviews and design quality assessments. YOU are now liable for your own technology.

So, what are you waiting for?
I don’t know about you, but I think this is going to be one of the most interesting presentations at ESC Boston. This session will be held in the ESC Engineering Theater, which means it's open to all pass types — both full conference and FREE Expo-only — but either way, you need to be registered to attend. Hopefully I'll see you there.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.