IOT Manufacturing: China, Taiwan or Elsewhere? Interview with Titoma CEO Case Engelen.
Where is IOT manufacturing headed as the US China trade war continues?
In the interview for Digital Metropolis podcast, Keesjan Engelen (Case) – CEO of Titoma – shared his extensive knowledge of the industry, and how he sees the future of smart city assembly.
Case also shares why he chooses to live in Taiwan.
By reading the interview below, you’ll get to know more about Titoma – a B2B electronics design and manufacturing firm located in Taipei specializing in IOT, manufacturing and consulting – its history, way of work and vision.
-Hi, will you please introduce yourself to that?
– Sure. My name is Case Engelen. I’m the CEO of Titoma, which stands for time to market. We are a product development firm in Taiwan, doing electronic product design and manufacturing.
– It is great to be speaking with you today, Case.
You already mentioned briefly the kind of work that you do, but just for our audience, could you explain more in detail
What is IOT manufacturing and how is Titoma different from other manufacturers in Taiwan?
– We are more like a consultancy.
Western companies generally come to us with an idea, a product that they need to have made, and then we take the whole project from end to end.
From napkin sketch all the way to a ship product. There’s quite a few companies in Taiwan that do what we do on a much larger scale.
There’s the big laptop manufacturers, they make 95% of the world’s laptops, but we focus on embedded devices
– Now I know you’ve spent more than 17 years in Taiwan, so this has become a new home for you
What was your journey to finding Titoma, and specifically to Taiwan from Netherlands?
– I came here originally to learn the language some 25 years ago.
Originally, I started working for a western Dutch industrial design firm, helping Taiwanese companies and American companies like Polaroid and Gateway with their industrial design, making sure their products looked better so they would sell more.
I did that with quite an international team which was really fun. I’ve always liked working in a very international context, but after some five years I wanted to have more control over the whole process, and just be a consultant.
So that’s when I started with a couple of friends Titoma some 17 years ago.
We’ve been going strong, ever since it’s been a bit of an up and down, but recent five years have been really good.
– That’s great to hear.
Has your Chinese also improved in the time?
– It has actually gone down. I’ve been here for a long time now. In the first three and a half years I have Chinese classes every day.
The first nine months was like full-time Chinese.
So, I speak quite a bit of Chinese, but to be very honest in my daily work with some of our people working here, I speak in Chinese just because it’s sometimes quicker, sometimes more fun.
It’s not really needed, because everybody here speaks English as well.
– That makes sense. That’s the double-edged sword of Taiwan. You come here to learn Chinese and you don’t always need to use it all the time.
– I tell people you should find a partner who’s not too ambitious about learning English.
– That’s great advice for our audience. So, we have specialized and we’ve talked to many companies that focus specifically on smart cities.
As you definitely know, IOT has become the central infrastructure of how people don’t just discuss urban infrastructure, but how they imagine it as a whole.
So how would you personally, through Titoma, define IOT?
– I don’t have a very enlightened definition for it. I would just say: connected devices. That’s how I see it.
– And has this idea of connected devices or even the way you manufacture them changed over the last 25 years that you worked in the field?
– It’s commoditized a lot. In the beginning everything was difficult.
It had to connect, and there had to be an app, and had to be a server, and everything.
Right now, a lot of the building blocks are there, and on the hardware side there’s lots of reference designs out there, so it’s become a lot faster to get something up and running.
So, I think the impact that has for a lot of hardware firms, is that hardware is no longer really something you can build your competitive advantage on, because any box with a sensor can be replicated pretty easily.
The competitive advantage is gonna be on everything that’s around that.
The app, the community, the user base that you build up, the database with all the data that you mine for useful results for the users, that is what it’s going to give a long-lasting advantage.
But yeah, we’re able to make a box with the sensor. It’s what we do, but to be honest there’s a lot of people who can do that, especially here in Asia.
– So, what have been the emerging opportunities in applying IOT to solve different problems?
– There are many different applications. The IOT adoption in general has not gone quite as fast as people had predicted that.
There are all these lofty numbers that were thrown around 2020. There’s been a lot of people that were just IOT guys like: “we need to do something with IOT, do an IOT project, we need some pilots”.
Then they get some guy they know and ask “can you connect some sensors to us and just do a pilot?”.
Okay, that gets done, and then very often it’s what they call pilot purgatory.
Nothing happens, and of course what a lot of people forgot to do, is developing a product.
Especially something still fairly convoluted IOT product. You need to have a very strong business case.
How is it going to make money? Just like the internet will collect data and become rich with the underwear nicking gnomes, that’s not a very straight route to profitability.
A very solid business case, I think that’s what essential. Like how is this data gonna help your company to give better service to your client, to give faster response, more uptime or whatever.
– Have you seen any new business models emerge that have really transformed the way you do your work?
– I have one of our clients had an interesting model where they installed their equipment.
I need to be a bit anonymous here, but they installed it in 95% of the premises.
Then, they would do a revenue share with their clients, where they would get like 50% of the uptick.
It’s a bit of a seasonal business.
So, without that, there would always have been an argument on the weather, and the holiday season, but having this baseline of the 5% where the solution was not implemented, makes it very clear what the device is providing.
So, I thought that is a very smart way to measure your outcomes.
– How are they collecting revenue to keep them anonymous?
– They help sales of a seasonal business by advertising.
So basically, because it’s a bit like selling airplane seats, if you don’t sell them, they’re gone.
So almost 80 – 90 percent of all the extra revenue they create is margin.
So that’s why that’s very suitable for a revenue share.
– Do a lot of clients and your clients companies specialize in a particular industry or field?
– Most of them have their own core technology.
It can be RFID access systems or UVC curing like hardening glue, for example.
Actually, most of our clients. For them, we are enablers. They sell for example glue that needs to be hardened and they just want to sell a lot of glue.
They do not want their clients to have problem with the glue hardening devices. They just want to make very sure that they don’t fail.
Or a company that’s selling beer – they just want to make very sure that the beer stays at the right temperature, so they don’t get complaints.
We work with companies that are not necessarily hardware centric themselves.
We are like their partner. They just say: “you guys take care of all that hard work, we want to focus on sales and building that community, building that infrastructure, building the app”.
– That makes sense, and now it also seems as you mentioned about IOT.
There have been these auxiliary terms that are launched into the bucket of software street or marketing, like 5G, AI, big data processing.
On the surface it seems like these very revolutionary technology has a way to function through IOT to really transform the way people live.
Are you and your company prepared for these changes or do you see them as still being very far out?
– We do lots of different kinds of RF radio – from RFID to BLE to wi-fi 4G.
So, in a sense going to 5G, it’s just another module. It will enable a lot more data at lower cost.
5G when it’s really rolled out in a significant way, can work – for example – with a dash cam for cars.
Right now, if you want to send over full 4k data, that’s a lot of data. It’s a big charge in your data plan and very soon that is going to be a lot more affordable.
At the same time, edge computing is another trend that is actually going in the other direction, where a lot of the analysis is done closer to the users, or where the data is captured.
So, you don’t need to send as much raw data up to the cloud for analysis there. With microcontrollers becoming a lot cheaper these days, it becomes much more affordable to do analysis on the ground, without using RF so much.
– The way you’re making it sound is you’re going to have a lot more orders for IOT devices.
– I do have to say the whole COVID-19 thing has been quite an influence.
It’s impacted the economy worldwide, so naturally that also impacts us in the manufacturing orders.
Although for some reason there’s still a lot of people who want to do new things. Maybe a lot of people now have time to think what are we going to do next?
What is going to be the next episode? So that is going to be quite interesting, and we’re also supporting some in some small ways.
It will be interesting how we’re going to get all this under control.
Here in Taiwan we’re really good. I think we haven’t seen a new case in the last 200 days, but we’re a bit of a bubble here and the rest of the world is really gonna take some time to get it all sorted out.
– I agree that’s the elephant in the room. This is the only place in the world this is possible.
– We’re actually gonna be the ones that are the only ones in the world that do not have the herd immunity.
– Hopefully there’s a vaccine, so that’s not going to be a concern.
So, it seems like the way you’re describing it, over the past 20 years there has been a real decrease in the price of producing the hardware, and now we’re seeing a decrease in moving of data and then processing it.
– If you look back to what the first PCs looked like… the first of the first supercomputers in terms of those capabilities, everybody carries a supercomputer in their pocket these days.
The smartphone has actually been a real driving force in making a lot of components such as cameras, and memory, and screens available at much lower prices.
Because of the huge quantities. So, the camera industry has really done a lot for the whole price curve.
For ourselves as a company, a big change that we have seen, is that before, we really worked a lot with – what I call – ODM suppliers, factories already making something.
For example, somebody needs a display to display advertisement, there you go.
I already know a factory in Taiwan making that, so let’s work with them.
Give it a nice industrial design, adjust the firmware a little bit, interact with them between the client and the factory, so everybody needs to know what needs to be done.
That was initially our business model at Titoma, and it worked okay, but it was just really slow.
We got a bit frustrated with that.
It took quite a while to say why it is not working. Because if you’re working with a factory already making, say… displays, they have the largest quantities because they’re already making a hundred thousand a month.
They exactly know how to design this kind of a product.
They have seen product returns, so they know what kind of problems you could have.
They have tested all the potential components, so they know which components have the best value for money.
They have designs 95% ready to go, so it’s the fastest way to market.
It’s a really compelling case, but it took a really long time to get it done.
The problem is that factories are not really focused on product development. Factories want to manufacture.
They want to keep their production lines running at 98% capacity.
So, if you have a new client with a new to the world product, that is going to have a first order of 2000 units.
At the same time Dell has an order for 50 000 units.
There’s a little issue that really needs to be solved, because otherwise those 50 000 units as the new product, you’re always going to be at the very bottom of their whiteboard of priorities.
So, it always took a really long time. ODM factories are optimizers, they are tweakers, but they’re not really these sort of inventor types that you need to make something that is really new.
– Behind me here, it’s an amazing team that you’ve found over the years.
What has been your process for differentiating yourself specifically on the industrial design, the speed, the quality?
– We first worked very much with the outside factories.
Got frustrated with that, and then we started to build our own teams.
Hardware, firmware, mechanical design – all in the house. In the beginning that was very tough, because people come to you and they say: I want a likes counter, how much will it cost to make that, and how long will it take, and what will the unit cost be?”.
It’s really hard to estimate something that you’ve never made before, so it took time to get good at that, but we learned.
Now we’re really at a stage where we have solid processes in place for everything.
I was very lucky that I found a really good CTO, Leo.
He really built up the teams. In the beginning he always had a very international crew, like half Taiwanese and half foreigners.
I’m from Holland, there’s people from El Salvador, from Colombia, from France, from South Africa.
It’s quite an international mix here, but if you compare it with our neighbor across the road – Acer – we’re still a pretty small company.
That was really hard to find. Leo is from Colombia. He has a really good relation with the university there.
It’s like the MIT of Colombia, it’s not very well known for electronics but there’s 48 million people.
There, we got a really good relation, because the professors there said: Wow, one of us has become CTO of an international company, can you do more of that?”.
So, we started the trainee program where every year we get the top 10 graduates.
Five at a time we bring them to Taiwan for half a year of training here, in our Taiwan office.
Then bring them back and continue in Colombia. That has been working out very well because everybody is trained in our system.
That is really nice, because it means that everybody works in the same way. It’s a bit like the Borg from Star Trek which is really helpful.
In the past we’ve had a case where we had a very complex project with the 16 custom PCB boards.
The key firmware engineer for that project was a bit older and he suddenly got an eye problem. His wife said: “You have to stop working now”, and he stopped within the week.
We had a really hard time picking up that project again and get it on the road again, because if somebody is working to its own standards, it’s really hard to dig into somebody else’s firmware unless he follows a very well-defined structured method.
So, we have done all that defining and structuring, and now people are almost interchangeable.
That means if suddenly a project is really urgent, then we can add more people to the team. If somebody goes on a honeymoon for three weeks, somebody else could take over, and so on.
So that has really made our whole process a lot more reliable
– Just for clarity, how does Titoma distinguish hardware, firmware from software?
– Hardware is for us. The PCBs, the printed circuit boards with the electronic components on them.
The firmware is the software that that runs on these PCBs. We’re working not with Pentium-like chips, but with microcontrollers.
In the smaller microcontrollers every of them has its own language, and you really need to optimize very much for that particular piece of microcontroller and the components that work with it.
That’s what we call embedded design. So, in embedded design you have a really close link between the firmware and the hardware.
The engineers sit side by side to make that work well.
The difference with a PC is that somebody in India can write the software, and somebody in Taiwan can make a PC box, and it will work together because Microsoft is in between.
That’s what they call the hardware abstraction layer. That is the big connector. Software, back-end software, that’s what you run on the servers on AWS Amazon.
In general, when we make an IOT device, we make a skeleton app, very basic bridge between the hardware device and the server.
And then our clients that are in Alabama or in Germany or in San Francisco, they do the app there themselves. They work with focus groups.
They tweak and tweak, but that is sort of a work they can do separately from our hardware.
– It seems like your company is very focused on the future. In terms of human capital, you’re working with all of these bright students from Colombia.
How are you conceptualizing the industrial future in a lot of conversations about smart cities?
People are envisioning this world in which every street is able to communicate with each other automatically: every light pole, every car.
There are obligatory questions of what kind of material are you using?
Will these devices be viable in 5 years? And 10 years? And 20 years? Are we going to have to throw them all out every decade and then rebuild them?
So how does Titoma think about sustainability and kind of designing for 20 years from now?
– That is a tough question because future proofing your hardware always sounds good, but it’s always been really hard.
We’ve done that in the past and somehow the technology always develops a bit different than you thought it would.
People suddenly find some feature really important, and future-proofing hardware is really difficult.
It’s the best to work in a modular approach where you can connect – for example – a different radio module or a different camera that is going to give you the more flexibility for the future.
In terms of housings, in general you can count on things becoming a bit smaller, so you don’t have to oversize your housings too much.
Anyway, it does make sense to leave some space for various kind of eventualities, because having to change out your phone every two years or so, it is a big waste of a lot of resources.
There’s a Fair phone – out now by a Dutch firm – that is actually modular.
As long until you just drop it, your display is probably still gonna work, but you may want to upgrade to 5G.
But then – on the other hand – people are going to say: “if you have 5G, you have so much more data” so you probably also want the bigger processor, and you also want a higher resolution screen.
So, a lot of it is still developing in sync and it’s hard to predict where we will be at 20 years out.
– Of course, and you’re not in the business of fortune telling, I suspect.
And there often is a certain fashion chasing in technology companies, both in terms of addressing consumer needs that are constantly evolving, but simultaneously creating new perceptions and new interactions with your devices.
Has Titoma evolved in the way you understand IOT and the way you design the hardware and the firmware?
– The big thing is we’re a service company.
People come to us and say: “hey this needs to be happened, this is what we feel needs to be done” and then we say okay, and we do a lot of suggestions so you can better do it this way or that way.
Then we find this amazing module that is really good value for money, so why don’t you use that instead of insisting on this one that is three times more expensive? but for us personally as a company what has changed more is that we did a lot of consumer projects, and quite a few startups.
In consumer there’s this relentless race to the bottom in terms of price.
It’s really hard to win that. Nobody wants to win a race to the bottom, and so that is a reason that we are now almost entirely focused on the B2B business where reliability is a lot more important.
In B2B businesses people do not want to have returns, because it’s really expensive to have stuff replaced.
– Now I’d love to ask about your choice of Taiwan.
You’ve written recently about the outflow or supposed outflow of companies from China and how it’s not actually panning out right.
Many companies are staying behind. There’s this conversation of a decoupling.
We’re talking as an election is happening in the United States. What is the evolving role of Taiwan in these, the companies across the world and their technical understanding?
– In electronics you’ve seen there’s a lot of animosity towards China recently.
A lot of clients come to us and say: “No, we really do not want to have anything made in China anymore” and I agree with many of the points. And China… frankly I’ve been in Taiwan for 25 years now.
A lot of my friends moved over to China and I’m very happy that I didn’t.
I feel it’s becoming a bit of a difficult place to live there with the whole social control and all that.
But in terms of electronic manufacturing the thing is China over the last 25 years has built up such a mass of factories of electronic components.
Especially that is something that nobody is going to replace within the next 5 or 10 years. A lot of people have done their due diligence.
Some are asking: “Can you quote me this in Vietnam, and in Malaysia, and in Taiwan?”. But if they are already making something in China, and they would have to redo the injection molds very often, they don’t own the design.
The design is actually owned by a factory in China, so they would have to completely redesign that whole device.
That is a big investment. They do all that investment and then they go to a different country. The costs in that other country are going to be higher actually than they’re going to be in China.
So, there’s been a lot of talk and a lot of diligence and studies. But if you look at the big companies like Samsung, you can see that they have moved all their Samsung owned plans out of China.
They are going to come to India and to Indonesia, or wherever. Those are very high-profile cases, and they’re really big, so they have a big influence also on the export numbers.
The thing is, if you’re making a million phones per month or per week, anytime you set up a new factory in India, you can just say to your top 20 suppliers:
“Hey guys, we’re going to go to India, I need you to also build a plant next to our plant to give me your custom housings, and your customized displays, and your customized PCBs, and your customized cable trees” and so they got their top 20 factories to all come with them.
With that kind of business volume, you can just demand that of your suppliers, but if you’re producing 5 000 units of this or that a month, then there’s not many people who are going to come with you to Myanmar.
And I feel that the perception in the media is a bit cute, because of these big phone companies.
The majority of the companies now manufacturing in China are just sticking out there because it getting you tremendous value for money in China, and you continue to do it.
What may happen is that final assembly may move. It may move to Taiwan. A lot of it already has moved to Taiwan.
It may move to Malaysia or Thailand or Mexico or even the US, but that is final assembly.
That is basically taking that circuit board with the components on it and screwing it into a plastic box.
But still very often 70% of the value of a product – I mean an electronic product – is in the electronic components: all those PCBs and cable trees on this place.
And for that there’s just no better supplier than China.
So, we were talking about final assembly you can actually do anywhere.
It doesn’t matter. We have an office in Colombia where we’re actively looking at options. They are also closer to our US customers.
But still, the components that make up 70% of the value of a device in a majority is going to come from China.
That’s because in China for every component there are 10 different factories that make it, and especially in the design stage when you don’t know exactly yet which components you’re going to use, and from which suppliers, it’s really handy to be there and have 10 suppliers.
Here, around our office in Taiwan, or around our factory in Shenzhen, we can go to see what’s possible here.
If our favorite supplier suddenly turns out to double his prices, and we didn’t quite understand why, that’s not a problem, because next door there’s another guy making the exact very similar device.
And that what keeps everybody very relationship focused. They really want to do their best for you. They work very fast.
They work with really low minimum orders, and basically, they all are competing with each other, so they give you really good value for money.
And the big thing with electronics is that if one of your 187 components is missing, your line is blocked. For the next five years, I still think China will continue to play a really important role.
– What do you think makes Taiwan such an ideal location to move final assembly to? Or to move your company’s operation as a whole?
– Taiwan is a really nice place to live.
We have quite a lot of international staff, and they all like it here. I am very happy here for the last 25 years.
You don’t imagine that, but it’s actually very green. Behind me we have nice green mountains, there’s jungle with waterfalls and all that.
But even more importantly, the people are really friendly, really helpful. If you’re lost in the city they will take you by the hand and take you where you need to go.
It’s quite different from China, really. And in a business sense, Taiwanese companies are a lot more reliable to work with.
Of course, you have bad apples everywhere, but it’s really interesting. If you look at HP and Dell and Cisco and all the big brands, they are working with Taiwanese firms.
About Apple, you would expect that the procurement guy, the head of purchasing of Apple, is going to say: “I’m going to go factory direct”.
That’s not what he does. He goes to Taiwan. He says: “Dear mister Foxconn, please go build for me a factory with a million workers, build it for me in China and I want you to run it”.
And that’s because there’s a reason for that. They have had some bad experiences in the past.
If you go for this approach of having both: the design and the manufacturing done by one firm, which for many reasons makes a lot of sense because it’s a lot faster, you get better components, you get a lot better value for money anytime, there is an issue with any of the components.
The communication is just so much better here.
So, it makes all the sense of the world if you’re going to be a factory here, to also do the design here.
But you’re putting a lot of power in the hands of one factory, the hands of one company, so you need to be able to really trust that company and make sure they’re not going to do a Huawei on you.
– Speaking of what are your thoughts on this… very literal D-linking and Huawei now building their own factories to develop microchips.
– The thing with Huawei was that Cisco had their routers made by them before. And then they said: “Hey, you guys have all these 700 engineers sitting here, why don’t you also design the next generation for us and make sure that you pay attention to this and that?”.
Okay, that worked well for a couple of years, until Huawei suddenly came out with their own brand of routers.
And apparently in their first manual – which was silver beta copied – the helpline number was in the US.
If you rang that number it was answered by Cisco still. So, they didn’t appreciate that.
That was a case of where outsourcing went horribly wrong. So, you have to be very careful about who you pick.
Now, basically I think 70% of the world’s electronics is coming out of China. For those reasons of components and basically for the whole mass of not only components, but also the talent, the mold makers, the glass makers, the camera lens makers – all the expertise is there.
All so close together, and they all talk with each other.
It’s really the Silicon Valley of hardware in Shenzhen, and still also in around Taipei.
So about decoupling Huawei… the US has realized recently that actually 70% of the components come out of China, but ICs is the key factor where China is behind.
Of all the ICs that are being put into electronic devices in China, only about 15-16% are being put in there from local Chinese manufacturers, and maybe 60% comes out of Taiwan.
They’ve realized that ICs are really the key components and that’s where they can keep them in a chokehold.
So, they have forced TSMC – which is the world’s biggest manufacturer of ICs in the world – to stop selling ICs to, which is a very drastic and very aggressive move.
It’s very interventionist, but that’s what’s been done. Huawei, they’re now saying that they are actually building the 5G which they’re pretty good at.
They’re going to come with their own operating system which is some sort of a diffused computing system.
Because of the high data capacity of 5G, basically it will be like a hive of small phones that all compute together for you.
So, they don’t need those really strong processes anymore. That way they supposedly can sidestep the high-end processors that are being made by TSMC.
I’m not really sure how that is going to play out.
So far, for the last five years already, China has been really stimulating the local IC industry, with billions of investments, and it has barely moved the needle over there.
But on the other hand, they’re very smart people and they have managed to catch up in nearly everything else, so eventually they’re gonna get there.
You can well argue that by decoupling, the urge for them to build their own IC infrastructure is going to be even higher.
– Do you envision a very aggressive separation where you don’t have an iron curtain, but it’s just a silicon curtain where all of the Chinese supply management is within China proper?
– No. Because it’s actually a lot of clients that are saying: “Mister Wang, I really would like to continue to buy your fabulous gadgets, but I need you to have a factory in Indonesia or in Thailand or wherever, because I cannot afford to pay those 25% or 50%, whatever it’s gonna be”.
So that is going to be a powerful push to do final assembly in other countries. And final assembly, you cannot just do trend shipping, like just put a sticker “made in Vietnam” on it.
There is a legal criterium called significant transformation, so something really needs to happen in Vietnam or in Taiwan.
Some people argue that even putting a PCB inside a box is not enough of a transformation yet.
So, it’s well possible that you need to import all the electronic components, put them on a PCB here in Taiwan, put the housing around it, and then ship it out. Then you are completely safe.
– Thank you so much Case for taking the time to talk with us. This has been fascinating. Do you expect to stay in Taiwan?
– Yes. I am married here. I have the lovely children, lovely wife, and we are very happy here so… yes, I do, I think.
– Great! And just as the last question: what is your favorite Taiwanese dish?
– The gongbao jiding. It’s not very original. It’s chicken with the cashew nuts and a lot of red peppers. It looks really scary but it’s not all that spicy at all, and it’s actually really good.
– Fascinating! Thank you so much. It’s been a pleasure.
– Thank you.
This has been brought to you by the Taiwan Report and is available also in form of a youtube video. For more content like this, follow Digital Metropolis, a podcast about the future of cities and those building it, hosted by Roman Shemakov.