Next-Generation IoT Computing Modules: An Interview with Onion's Zheng Han

IoT computing modules are in high demand. In 2014, Onion Corp. set out to develop a small, wireless hardware development platform specifically for software developers. In just four years, the company has launched three remarkably successful crowdfunding campaigns featuring products that appeal to wide range of IoT innovators with diverse skill sets from makers and pro engineers. Onion's CEO Zheng Han talks about his company's IoT computing modules, running successful crowdfunding campaigns, and more. 

Onion's Origins 

C. J. Abate: Before we discuss Onion Corp., your IoT computing modules, and your crowdfunding campaigns, tell us about your background. I understand you worked as a software engineer at Siemens. You also interned at Analog Devices.

Zheng Han: My background is in electrical engineering with a focus on telecommunication and information systems. The first intern job I took was in a small Canadian company called RuggedCom, building network management systems. During my work there, I witnessed the rapid growth and expansion of this young network equipment company, which eventually led to an acquisition by Siemens, and this is why I was invited back to work at Siemens after graduation. The whole experience at RuggedCom/Siemens really planted the entrepreneurial seed in my head, so not long after I decided to quit my engineering job, I gave my first try at starting a company.

C. J.: What led you to focus on hardware and launch Onion Corp. in 2014? Was it a difficult decision to start your own business?

Zheng: Onion first started as a SAAS company. The initial plan was to build an IoT cloud that connects Internet-enabled devices. Later, we realized that, at the time, there weren't too many hardware modules with the capability we envisioned for our cloud services. So, we then we started building our ideal IoT module, and this was the start of the Onion Omega. We did a Kickstarter campaign on the idea and it became a success. We then made a pivot in the business plan to focus more on the hardware side, which lead to more hardware products, continuing to the Omega2. As with most other entrepreneurs, I found the most difficult decision was to start a business and to let go of a good paying job and jump into the unknown world of startups.

C. J.: Can you share some details about your experience with TechStars?

Zheng: My time at TechStars is best described as accelerated learning on steroids. There were times when we had to meet and talk to five to eight mentors every day for weeks. At the end of it, I would have so many different ideas, sometimes conflicting ideas, running in the back of my mind. After a while, the good ones tended to stay, and the bad ones were forgotten. Suddenly, running a startup company started to somewhat make sense.

C. J.: Tell us about Onion today. How big is your team? What is your business model?

Zheng: Onion has actually undergone quite a few changes in the past 18 months team-wise. We're leaner than we were before and now have all of the right people in the right places to boldly stride forward. From very early on, we approached building the team with flexibility in geography; the realities of being a hardware company almost force your hand in this respect.

We currently have a manufacturing and logistics team of four people in Shenzhen, China, a management and software development team of three in Toronto, Canada, and a VP of Sales in Boston, Massachusetts, USA. Unlike most other Internet startups, our core business model is very simple. We sell differentiated IoT computing modules, and we help our customers to get better and faster at making IoT products using Onion technology.

Onion's IoT Computing Modules 

C. J.: The Onion Omega IoT computing module was your first product. It was a tiny, Linux-based computer with built-in Wi-Fi. Tell us more. How did the product come about? 

Zheng: The Onion Omega (Figure 1) was our very first hardware product. We started working on the Omega in 2014 and launched it in April 2015 on Kickstarter, with a goal of providing hardware for people more familiar with the software world. Back then, the world of hobbyist electronics was dominated by two major players: Arduino and Raspberry Pi. (They're still the major players today, but the ecosystem has really expanded.) We had been using both of these platforms extensively and found ourselves wishing there was something in between that combined the best of both worlds.  The Arduino boards — while very useful for interacting with other hardware and the automation of straight-forward, repetitive applications — didn't make it easy to create any sort of networked projects, effectively creating a ceiling in terms of the projects you could build. The Raspberry Pi, on the other hand, with its Linux operating system and powerful processor, opened the door to using programming languages other than C and made connected projects possible. But at the same time, it was quite large, very power hungry, and needed some add-ons to even get started with it. We saw a gap in between these platforms, for something that was small, that had wireless networking built-in, and had a processor with enough punch to run modern programming languages — a device for the IoT.

And that's how the Omega came to be. It features a small microprocessor that runs an embedded Linux operating system, making it effectively a microcomputer. The storage, memory, and Wi-Fi radio are all built-in, so all that's required is power, and it's ready for building IoT applications.

C. J.: The Onion Omega’s Kickstarter was quite successful. I see that 4,459 backers pledged $267,851 for the campaign. Your original goal was $15,000. That’s amazing. Describe the process from launch to delivery.

Zheng: Thank you! As I said, we saw a need for an IoT-centric device that would make hardware development accessible for software and hardware developers alike. We were guessing there were lots of people out there just like us. So, we decided to find out! During the pre-campaign, we worked for months to define what exactly we wanted to create that would be useful. Then we prototyped these designs and started working on the software, which was a key component of the whole thing.

As for the Kickstarter campaign, we were coached a bit to set ourselves up for a real splash. Coming up with a target goal of funding can be a bit challenging for hardware companies. If you have an absolute minimum amount of funding that you need to raise to design the product, then you need to set a high goal. In our case, we didn’t have any large tooling costs, so we could target a low numbers and blow it out of the water. That gets you a lot of visibility in the Kickstarter community, which then feeds on itself.

C. J.: In 2016 you followed up with the Omega2 and a campaign that did even better — $672,801. Describe the computer’s specs and explain how it is different than the original IoT computing module.

Zheng: When we did the original Omega module campaign, we got lots of feedback from our Kickstarter backers. Some of this feedback had to do with functionality, performance, and price. That’s what lead us to start looking at other SoC processors that might offer some additional features, run at a higher frequency, and support a lower price. We found this with an SoC (system on a chip) made by MediaTek, and this is what we use in the Omega2 today (Figure 2). This SoC is very solid, and it is used broadly by large networking equipment suppliers, so that helps drive the volume high and the cost low.  To give the campaign some added boost, we increased the number of plug-in boards — “expansions” as we call them — to help customers create cool projects. And we came up with the idea of adding education kits as well. Many of our backers did not have much experience in electronics and coding — that’s the beauty of the Omega, you don’t need to be an engineer or computer scientist to use it and create things.

Oboo Smart Clock

C. J.: Onion’s Oboo Smart Clock (Figure 3) looks amazing. What’s the story behind it? Why a clock?

Zheng: With the Omega business, we were always coming up with project examples and tutorials to help inspire customers to build cool Internet-connected things. But those were typically small little “one-off” projects like an IoT thermometer or a robotic arm controllers or small device that would display the latest stock market prices. We decided that we wanted to take this a step further to make a complete end product ourselves to show the world what you could do with the Omega inside — and maybe expand into the consumer electronics at the same time.  We brainstormed for weeks, looking for an IoT product idea that was unique and could really take advantage of being Internet connected. You know everything is becoming Internet connected these days, from cars to refrigerators to coffee makers. In the end, we decided to build something that would meet our own needs, a product for ourselves. We wanted something that could wake you up and get you prepared for the day, without forcing you to immediately dive into your phone. And then we needed to name it, and found that these days, a two-syllable “non-word” product name is popular. We came up with Oboo. Plus, it starts with an “O,” so that made it fit into Onion and Omega, etc.

C. J.: From what I can tell, the Oboo clock was your third and most recent crowdfunding campaign. Correct? Tell us about it.

Zheng: Yes, we created the concept for Oboo, and got started working on the design with the plan to launch it on Kickstarter. Since we had already gone through two previous campaigns, we knew the recipe, the way to initiate, launch, and then nurture the campaign. As I said, the key with IoT is to be Internet connected, so with a clock there are lots of possibilities. For instance, you can always get accurate time from the Internet, so that makes Oboo always have the correct time. And when the time changes for Daylight Saving Time — bingo, your clock changes on its own, no more manual changes.

So now, with an Internet connection to your clock, that opens up a whole world of features that you can build into it. We had to be careful not to go too crazy with this, but we ended up in our campaign starting with a small feature set, and then getting feedback on which features were more popular than others. This is one of the key benefits of Kickstarter — to get that feedback from consumers before you finish the design. Some of the initial features that we did decided to offer were things like weather conditions, a local traffic report, and streaming Internet music. We came up with the concept of “cards” to display data from the Internet on the front panel of the clock next to the time. Each card would show information about a specific topic, like weather or traffic, and the user would switch between cards with a wave gesture.

Which brings us to gesture control. We wanted Oboo to have a real futuristic “Wow” factor, something not really common — and so you can simply wave your hand in front of it to change cards or turn off the alarm, among other things. For the music, we wanted Oboo to sound good, so we added some nice speakers and enlisted mechanical design assistance from our manufacturing partner, who had lots of experience in small speaker enclosure design.

We also wanted Oboo to be able to communicate with a smartphone and play music over Bluetooth, so we added that as an additional subsystem. And, we also added a rechargeable battery inside, making it portable and also — drumroll please— to keep running if you lose power at night and need to wake up for that early morning flight.

C. J.: Briefly walk us through the general project timeline for a product like Oboo. How long does it take from concept to crowdfunding campaign to shipping?

Zheng: It was early August when we decided to go with the smart alarm clock project. We architected the hardware and software design and put our first initial prototype together in September. We attribute that to the ease of design that the Omega2 offers. With embedded Linux, we were able to quickly write JavaScript and some blocks of C code to begin to test out some of the features. In parallel to this, we worked on the physical design of the structure of the clock and did some renderings of what it would look like.

In late October, we settled on what Oboo should look like and what it would do, and worked with a local video contractor to put together a slick Youtube video that we could use for our one-month Kickstarter Campaign, which we ran during the month of November. From December to February, we finished up the hardware design and built our first full prototype unit. Now we started to get the confidence that this was going to come together.

Then during March, April, and May, we worked on the manufacturability of the hardware, and in parallel with that, focused more attention on software development. During the campaign, we established one “elite” class of backers who would get an early version of the Oboo, with the ability to connect to it, receive multiple versions of software updates, and give us feedback. We shipped our first units to these backers in late June, just over 10 months from when we got started on the project.

C. J.: Has it been easy to find manufacturers and then bring your products to market? Was it a trial-and-error process, or did you have contacts who helped make the manufacturing process easy?

Zheng: It has been an iterative, trial-and-error process, with lots of learning over the past four years. We’ve been manufacturing circuit boards for over three of those four years now, and during this time, we’ve worked with a variety of different contract manufacturers (CMs) and factories. There was a lot to learn, and we’ve made mistakes along the way. That’s OK as long as you’re drawing lessons from the mistakes and building relationships.

Right now, we’ve settled on three CMs that we use exclusively. Having strong relationships — where both the CM and you treat each other fairly and are happy with the work you do together — is helpful because then you can make use of their network. In the case of Oboo, we needed a factory that specialized in devices with speakers and metal enclosures. For us, this was not a problem, as we asked our trusted PCB CM and they were able to recommend and introduce us to other CMs who were experts in that field.

In terms of bringing products to market, we found that shipping can often be more complicated than manufacturing, and we learned a lot of lessons there. The key being: don’t be afraid to spend money on a good shipping provider. You will likely end up saving money in the end since you will have fewer lost orders that you have to ship again.

The Future

C. J.: Any advice for engineers and developers interested in bringing a product to market via a crowdfunding campaign?

Zheng: Crowdfunding can be a great vehicle for hardware companies to help get the necessary funds to create and then manufacture a product that is then “sold” to backers as a rewards from the campaign. Here are some of our recommendations: Do as much of the work you can afford to do ahead of time — but not too much. You want to keep the design open so you can make changes in the spec — add or delete things as you get that valuable feedback from backers. Be as aggressive as you can with the product features and price, but with the caveat that you want to exceed expectations and not disappoint your backers. Over-communicate with your backers. Send them regular status updates about where you are at, what your challenges are, and some background on any decisions you make.

We’ve found that as long as you are transparent and communicating, most (but not all) backers will stay happy and pleased with their support of your company. I said “most” backers, as you will have some that will use the crowdfunding forum to give you a hard time, to complain and threaten to cancel their pledge. As long as you expect some of this, and take it in stride, you will hopefully gain some constructive criticism and do just fine.

Here are links to our Kickstarter Campaigns:
 

• Omega1 (http://bit.ly/see-omega)
• Omega2 (http://bit.ly/see-omega2)
• Oboo (http://bit.ly/see-oboo). In the Update section, you can get a feel for the frequency that we were sending out updates to our backers.

C. J.: What is next for Onion? Any new products in the works that you can tell us about?

Zheng: We can’t tell you much right now. But we can say that we have great plans to continue to the support of the Omega2 for both the maker and commercial OEM community and to expand our presence as the invention platform for IoT. Stay tuned.
 

---

This article will also appear in the November 2018 edition of Elektor Industry magazine.