LitePoint, a US-based wireless test solution provider, offers turnkey over-the-air test solutions from development to production, combined with software automation that makes testing of FR2 devices easy in both characterization as well as volume manufacturing.In a recent interview, Lite-Point president Brad Robbins, talks about the development of the 5G market, and what it means for the wireless test solutions sector.Q: As more 5G phone models become available in the market, what do you see the development of 5G market today?A: Despite COVID-19's unprecedented disruption on a global scale, in 2020 we saw 5G adoption rates grow dramatically. While the total number of mobile shipments decreased in 2020, 5G mobile shipments grew from just over 10 million units in 2019 to 180-200 million units in 2020. We are very early in the global deployment, but 5G is living up to its promise of better mobile access. In existing sub-6GHz (FR1) spectrum, 5G operates more efficiently, providing more capacity for operators in their existing spectrum. This results in a better user-experience for the end-user, particularly in densely-populated urban areas where LTE capacity is strained. As mobile operators enable more "mid-band" spectrum and MIMO, 5G devices really start to deliver on the promise of faster connectivity.The most dramatic demonstration of the power of 5G is in the mmWave (FR2) 5G spectrum, where >1 Gbps data connections are being achieved. Apple's recent product introduction shows that mmWave is practical to add to a handset, though the deployment of mmWave networks is currently contained to somewhat narrow areas of urban centers. As the build out of mmWave networks continues, this will provide the next major leap in available data capacity, allowing carriers to move beyond the mobile phone and broadly deploy fixed-wireless to offices and residences. We are still early in the roll out of 5G, but the adoption rate is growing faster than what we saw with 4G.Q: There are several testing solution companies with general purpose test and measurement solutions. What makes LitePoint testing solutions so unique compared to others?A: Achieving optimized production economics requires a test system that is optimized for production. 5G device production requires both an optimized solution for the sub-6GHz (FR1) band, where the calibration and testing is primarily performed in a conducted (cabled) environment, as well as the mmWave (FR2) band, where testing is performed over-the-air (OTA). For FR1 devices, LitePoint's production solutions provide industry-leading performance in power accuracy, signal quality, and spectrum bandwidth.LitePoint's offerings additionally cover the upcoming 5G NR-U and Wi-Fi 6E frequencies up to 7.3GHz, with a future proof 200 MHz of bandwidth to cover mid-band carrier aggregation scenarios. For FR2 mmWave devices, LitePoint provides the simplest solution for manufacturing compared to other offerings, with the entire solution integrated into a single box with a calibrated interface. This greatly simplifies the setup, maintenance, and reliability on the production floor, enabling device manufacturers to ensure that they deliver high yield with high up-time in manufacturing. Additionally, LitePoint offers turnkey over-the-air test solutions from development to production, including test chambers, antennas, and switching accessories, combined with software automation that makes testing of FR2 devices easy in both characterization as well as volume manufacturing.Q: As mid-price or lower cost sub-6 GHz 5G phones get more popular in the market, manufacturers are facing the challenges of testing time increasing and slow down the big volume production, what do you think how to help them to resolve the impacts?A: In addition to ensuring shipment qualify, test time is, of course, one of the next biggest focuses for production testing. One technique that we have been pioneering at LitePoint is providing solutions that test multiple devices at the same time with a single tester. This effectively enables the per-device test time to reduce, providing a lower cost per unit. Additionally, as a leader in production testing, we work very closely with the chipset providers to help develop or deploy new techniques to optimize test times. An example of this is the trend towards Sequence Based Test, which dramatically reduces the number of required communication transactions between the device under test (DUT) and the test solution. Instead of having one setup command per test, which occurs hundreds of times, a pre-defined list of setups can be loaded and setup before performing hundreds of measurements. Reducing this communication time provides an impressive reduction in the test time, potentially delivering up to 70% improvement over traditional test methodologies.Q: On the other hand, the OTA test for 5G phone in the mass production lines, will the testing chambers require the new arrangement of production lines? What do you see the OTA test challenges in big volume production?A: The concept of performing OTA testing on the production line is not new. In 4G LTE devices, it is common to perform some simple OTA testing of the phone at the end of the line. This test ensures that the final assembly of the phone, including the antenna and embedded software, have been correctly installed. This testing is performed in the near field and typically has relatively loose PASS/FAIL limits. For 5G, the addition of mmWave technology requires phone makers to provide a higher-quality OTA test with tighter PASS/FAIL limits. One advantage to mmWave technology is that the far field distances are significantly shorter than for sub-6GHz frequencies, making the size of production test chambers practical for the manufacturing floor. For example, the far field measuring distance for a typical mmWave antenna array for a phone can be <10 cm, which enables the placement of an appropriately designed measurement antenna very close to the DUT. It is common for the DUT to have multiple mmWave antenna arrays, so the production test chamber must be designed to support flexible placement of multiple measurement antennas.There are technical challenges that must be addressed to efficiently perform high-quality OTA testing in the manufacturing environment. First, the setup and calibration of the OTA chamber must be accurate and repeatable. mmWave antennas are highly directional, requiring the mechanical alignment to be very precise. For the measurement antenna setup, this can involve the use of a laser alignment tool to ensure that the antenna is focused on the center of the DUT antenna array. To ensure repeatable placement of the DUT antenna array, custom milled DUT holders are required which enable subsequent placements of devices to reliably return to the expected location in the test chamber. Another challenge in 5G OTA testing is overcoming the significant path loss - the decrease in the signal power between the DUT antenna and the tester connection. It is very common to see >50 dB of path loss in a 5G OTA setup - compare this to <10 dB in a conducted (cabled) test setup. This large amount of path loss needs to be accurately calibrated in the setup so that it can be corrected in the measurement of the DUT.Q: Both Qualcomm and MediaTek launched silicon chips for Millimeter Wave (mmWave) 5G devices. And both are investing more resources in ramping up mmWave development as they plan to roll out their respective next-generation 5G chips. What do you see the further challenges for mmWave testing solution?A: Though mmWave is new to consumer products, the technology is well-understood from its military and aerospace roots. The main challenge is scaling from "boutique" low volume to mass-market high volume. To enable this, an area of focus for LitePoint has been simplicity. Multi-component test setups in the lab can be powerful and flexible, but they are complicated and time consuming to setup, require specialized knowledge, and are very problematic to repeatably deploy on a large number of test stations on a manufacturing floor. The LitePoint approach was to develop an optimized tool for the production environment: fully-integrated, simple to setup, and delivers repeatable results without the need for highly-skilled technicians. For mmWave technology to become widely deployed in consumer products, the manufacture and test of these products needs to be as simple as the sub-6GHz wireless technologies. Our DNA is simple, robust, and repeatable equipment for manufacturing.Brad Robbins, president of LitePoint

While many might be deterred by the hefty costs of starting a semiconductor company, there are still many other entrepreneurs with a silicon dream knocking on the door of Silicon Catalyst (SC), an incubator and accelerator for semiconductor startups.Digitimes recently talked to SC's managing partners Richard Curtin and Tarun Verma, as well as partner Lance Bell over Zoom trying to better understand the company and their insights into the latest industry trends.Q: Please give a brief introduction to the accelerator and the background of the core team. Semiconductor is a very sophisticated tech and involves very complicated process. Do you have any specific aspect that you focus on?A: We are Silicon Catalyst, the world's only incubator exclusively focused on accelerating semiconductor solutions, including IP, MEMs, and sensors. We started operation in 2015, co-founded by Rick Lazansky, Dan Armbrust and Mike Noonen, against the backdrop of a falling growth in R&D in the industry, due to massive consolidation in recent years. Our executives and partners all have long careers in the semiconductor industry. We understood there is still significant room for improvement in industries developed by silicon. Based on that we try to address the largest problem silicon ventures have, mainly to lower infrastructure costs to make it happen. We put together an ecosystem of in-kind partners, that provide goods and services at either no cost or at a significantly reduced cost, spanning the pre-silicon and post-silicon stages of development, as well as an ecosystem of advisors, strategic-partners, and investors.We started this vision in 2015, at that time we had 18 in-kind partners, and a score of advisors; today we have 42 in-kind partners, almost 200 advisors, 300 investors, and seven strategic partners. We have expanded our footprints to other areas in the world. We have a JV in China, operations in Israel, and have been in discussion to expand to EU, India, Taiwan, and Korea as well. While we've done that, we are strengthening our partnerships with Global Semiconductor Alliance (GSA) and SEMI. We have recently expanded our university ambassador program on a worldwide basis.Q: Out of the 350 startups that you have engaged with since 2015, you have only admitted 35 of them over the years. And your Chengdu joint venture has accepted 16 Chinese startups since 2019. How do you pick startups? What are the criteria?A: We look at all the factors that venture capitalists would look at. But we bring real experts to bear - some of whom have built companies, been CEOs, or have been at the forefront of the industry for decades. That is part of our secret sauce. The seasoned semiconductor veterans in our ecosystem see their engagement with Silicon Catalyst as a "give back" to the industry.How do we screen the startups? For example, first we look at the team: is it a capable team? We look at the market, what pain they are solving? We look for evidence of customer traction. What is the value proposition? Are customers willing to pay for it? Finally, can they execute their plans to hit the market window? Do they have access to the right channels and go-to-market strategies? And then to get to the right design, and realize the revenues? We are very selective. We try to help every startup that chooses to engage with us, providing feedbacks to them. Some of them come back and get admitted later. We use the same formulae in China. China is a market with huge demand, with thousands of startups. It's been almost two years since we started operation there, and our joint venture will finish this year with a dozen startups in their incubator.When we first started, we only look at early-stage companies. As we have grown our ecosystem, we now have later-stage companies that have joined us. People that have raised may be US$5-10 million, may be more. We don't focus on any specific area. We are really looking for a business model that needs a piece of silicon to differentiate itself. So whether it's 5G, data-harvesting, AI, it's ok. We don't focus on a specific market segment at this point.Q: You mentioned that you have established a joint venture in China. Can you elaborate on the details?A: We partnered with a company called Silicon Power Technologies. The primary ownership is a Sino-American company headquartered in Hong Kong. The Chengdu venture is focused on semiconductors for the power electronics market. There is a lot of semiconductor demand in China, and so the innovative products that the startups are developing can address the needs of the market.We help the generally techno-centric people to round out their value proposition, reduce the seed investment amounts and the time to prototype for them through our 24-month incubation. We provide in-kind partner goods and services. Everything from pre-silicon to post-silicon, all the way through to businesses support services, such as corporate attorneys, IP attorneys, business development consultants, advisors, etc. And this is done on a worldwide basis. We have an application cycle twice a year, and we are coming to our application deadline on January 11, 2021 for the next batch of applicants. We go through the screening processes, meet with the entrepreneurial team we select with an agreement, and bring them into the incubator. During those 24 months of incubation, all in-kind partner goods and services are provided free of charge, or at a dramatically reduced cost, so they don't need to raise the money just to get business off the ground. We take that burden of design tool and shuttle run cost completely off the table. We can take them with their concept, and everything through from design, IP, design services, foundry. We have been working with TSMC since the first days of our incubator, enabling the startups with rapid access to MPW shuttles. When a company comes into an incubator, they have full access to all these goods and services that are only available with Silicon Catalyst.Very clearly, TSMC is front and center for companies in our incubator, being able to provide shuttle runs for their designs. We have access to the process-design kits (PDKs) for all the models. They can immediately start designing and knowing they can get onto a shuttle when they are ready to tape out.Q: For the 16 startups of your JV incubator in China, do they use the local foundry or TSMC?A: Since they are focused on the power space, they have some specialty foundries in China applied on the power IC segment. When we set up a joint venture overseas, what they do is take our methodology and recipe and then localize it. They have experienced a very rapid learning curve because we have been at this for so long. We have a process that is second to none, and you need the support of an industry to get them from a concept on the back of a napkin all the way to have a cashflow-positive business that maybe ultimately ends up with an acquisition, whether domestic or international.Q: Due to the geopolitical competition, some people are saying that the G2 trend are going to last. What do you think?A: Silicon Catalyst is really looking to be "the Switzerland of the semiconductor industry." The semiconductor industry has been growing incredibly over the last few decades. This has always been a global innovation market. The inventors know no borders. Let's find a way to make that continue. There are too many problems to be solved. Some people focus their attention on 5G, but there are far more problems than 5G. Silicon Catalyst is a worldwide incubator, we stay agnostic and work with startups and partners from all countries. More than half of our startups are from outside of the United States. We find our incubator provides essentially an epicenter of synergy to connect those startups with industry experts, as well as access to all the resources they need. We have engaged with over 350 new applications to enter our incubator - we have incubated companies involved with diverse market segments, from memories to biotech, to 5G, to energy harvesting. Where do we see a startup that is applying to our incubator? How will they fit into our ecosystem of advisors, strategic partners, and in-kind partners? And that really is the secret sauce of our organization because there is so much support in each and different areas.We only accept a startup that we think will be enabled with the access that we provide. One thing we do is truly unique: we de-risk the investment. So you have a company like TSMC that offers us shuttle runs to companies in our incubator, with their knowledge and access to manufacturing experts. And Synopsis or Advantest, and other leading pre-silicon and post-silicon vendors of design tools and services, coupled with experienced semiconductor executives. Once accepted, the chances of succeeding for those startups become ultimately greater. They come from all over the world, which is very exciting for us.Q: You seem to have an interesting startup portfolio.A: Let me show you the landscape of applications: everything from life-sciences, to edge-computing, to artificial intelligence (AI), machine-learning, inferencing for edge-computing, energy harvesting, cyber-security, etc. Everything has to do with semiconductors. Just recently we expanded our footprint into the MEMS, sensors, and actuators markets. One of the recent partners is ST Microelectronics, which is the world's largest MEMS manufacturer. This is an exciting development for us because there are many applications such as microphones, all the way to life science, etc.Q: People are much concerned about the bottleneck of Moore's Law. But as you mentioned, there are emerging applications and developments in the industry. Do you foresee a period of "golden years" for semiconductor industry? Any new trends spotted?A: We are now down to 3nm process technology, enabling billions of transistors per design. But there is an interesting new segment we are seeing in our companies of portfolio, which is optical computing. Optical computing probably is going to be the next wave. So if you are thinking computing horse power is reaching its limit, because there are so many parallel machines and architecture. But the next bottleneck is interface communication, I/O. Now we are seeing enormous growth in optical interconnect. One of our portfolio companies, Ayar Labs, is doing optical connectivity. Many of the startups spawn from university environment. Ayar Labs had a foundation technology developed by a professor in UC Berkeley, collaborating with two PhD students and one master's student. They developed a revolutionary I/O solution that they are able to push for significantly faster data transmission speeds. The key to their success is that they answered one of the questions we asked during those screening meetings: in order for your technology to be successful, does the entire industry need to lean towards your direction? Or are you actually a plug-and-play the existing ecosystem? In fact, they are manufacturable. They not only had a break-through technology, but also easy for the industry once they overcome the hurdles as a startup. This optical I/O is truly a game changer.We introduced them to various venture capitalists, guided them and helped to establish their strong foundation for business growth. They recently closed on a Series B round, raising US$35 million from investors. We would like to think that our very early engagement with them in our incubator aided them on their meteor path to success. We have multiple success stories like that. New technologies are coming everywhere, and over 50% of our startups are located outside the United States. It's our knowledge-based connection to a unique and growing ecosystem that is heart of our success, coupled with a comprehensive screening process to filter out early, those companies that have the greatest chance to succeed. With a willing group of participants as in-kind partners, they brought in with their services at no cost or low cost to our startups, because they, too, are in search of potential customers at this significant age of consolidation. Budgets for research and development had been slashed, so innovation from large companies have been limited.The future landscape could not be healthier. It has a worldwide footprint, and we are beyond delighted to explain our model and communicate to Digitimes readers. You cannot think about semiconductor without mentioning Taiwan. Silicon Catalyst is looking forward to even more semiconductor innovation in the years ahead.(Editor's note: Digitimes Research analysts Eric Chen and Jim Chien contributed ideass to the making of the interview.)Richard CurtinTarun VermaLance BellPhotos: Silicon Catalyst

Cyber security is a key issue for the era of big data, Internet of Things (IoT) and 5G. Cybersecurity Ventures has predicted that, globally, businesses in 2021 will fall victim to a ransomware attack every 11 seconds, and the cost will top US$20 billion.Digitimes recently had the opportunity to talk to hardware cyber security startup Dover Microsystems' CEO Jothy Rosenberg about how the company's solution - an oversight system - works and differs from others.Q: Could you briefly introduce Dover's solution?A: We are an IP licensing company. Our CoreGuard silicon IP resides on the silicon substrate next to the host processor and acts as an oversight system for the host processor - monitoring every instruction as it executes to ensure it complies with a set of security, safety, and privacy rules. If an instruction violates an existing rule, CoreGuard stops it from executing before any damage can be done. Updatable security, safety, and privacy rules, called micropolicies, maintain metadata about every piece of data and every instruction processed by the host processor, enabling CoreGuard to distinguish between good and bad instructions. Micropolicies are designed to stop entire classes of attacks, including buffer overflows, code injection, data exfiltration, safety violations and can even protect AI/ML systems from dangerous malfeasance.Q: What is the difference of the network security solutions between Dover and Synopsys from the early stages of a chip design?A: Synopsys and other EDA tool vendors are providing the SoC designer with tools to help prevent flaws in their HW design. Some have also acquired companies with technology to detect anomalies in bus traffic, such as the case with Siemens (parent company of Mentor Graphics) acquiring UltraSoC earlier this year. However, these tools do not protect any of the software running on the SoC - this is where CoreGuard is focused. CoreGuard is built on two critical observations: (1) today's processors are based on a simple architecture that prevents them from protecting themselves against attacks; and (2) all software has flaws that can be exploited by attackers coming in over a network. All processors sold today are still based on the 1945 von Neumann architecture. Although processors have advanced according to Moore's law and become faster, smaller, cheaper; they lack the ability to determine if each executed instruction is correct. Steve McConnell, author of Code Complete, states there are 15 to 50 bugs per thousand lines of source code on average, and according to the FBI approximately 2% of those bugs are exploitable. That means, in Android which has 15 million lines of code, there are at least 4,500 exploitable bugs. In Microsoft Office which has 40 million lines, there are at least 12,000 exploits. And a Ford F-150 which has 150 million lines contains potentially 45,000 different ways to take over the vehicle or steal private data from it. CoreGuard is what is classified by security researchers as an oversight system, watching each instruction a host processor is executing as it is happening and determining if that instruction is correct or not. And if not, then CoreGuard does not allow that instruction to complete, prevents any damage from being done, and notifies the host processor so it can take appropriate remedial action. The level of protection EDA vendors, like Synopsys, are providing does not overlap with CoreGuard. They are mutually beneficial as the goal is to protect our electronic systems with a defense-in-depth approach.Q: What is the level of acceptance of Dover's solution in fabless? Can you share with us some user cases/scenarios of your solutions?A: Our first publicly announced customer is NXP. NXP recognized the innovation and revolutionary approach of CoreGuard in protecting devices against network-based attacks right at the processor level. We are in discussion with several other fabless semiconductor companies, which have not yet been announced publicly.NXP is utilizing CoreGuard to create inherently secure processors for embedded devices. Additionally, we announced a partnership with Cadence Design Systems, to deliver secure processing for aerospace and defense applications using a CoreGuard integration with the Tensilica LX7 processor. Through this collaboration, customers can leverage the runtime security monitoring capabilities of CoreGuard that identify security policy violations and stop them from executing before any damage is done.We also announced a similar partnership with Andes Technology, a tech company in Taiwan, to deliver professional network security for their RISC-V core. Andes recognized that network security was a major concern for many IoT applications, which is why they decided to partner with Dover. We have been talking to quite a few large chip manufacturers over the past one and a half years, and those talks are coming to fruition. We are confident we can close significant deals next year.Q: Who are Dover's partners of IC manufacturers?A: We do not make hardware. We are only dealing with IC manufacturers for the purpose of getting test chips made. We are a part of an organization called "Silicon Catalyst"; they are in Silicon Valley. They were created to help semiconductor-related startups succeed. They've got a deal with TSMC and some other organizations, to make it easy to get into silicon inexpensively. TSMC is waiving all fees for this, but we still have to pay for services. They not only have a partnership with TSMC, but also a really good deal with Arm. We can get several different Arm cores for free to make test chips.Now, we are pursuing getting CoreGuard into an ASIC through our participation in Silicon Catalyst. Being proven in silicon is an important validation step for many of our customers, which is why it is a top priority, and we plan to do this in the first half of 2021 to get a few hundred test chips.Q: Who are your major competitors in this industry?A: We mostly focus on companies which are starting to do cyber security in hardware. We do not compare ourselves to software vendors, because in any kind of software, there are bugs that bad guys can exploit. There are two major areas people are doing cyber security in hardware. One is called "compartmentalization". For instance, Arm has TrustZone, Intel has SGX, and SiFive has WorldGuard. Those companies are marketing extremely aggressively. It is a valuable technology for sure. It keeps one set of untrusted software from being able to corrupt a trusted compartment (where Android might be running for example).The other message out there is "secure processor" which is encryption. Encryption is important. It keeps data that is moving from one place to another from being intercepted, stolen, or even looked at. But there is confusion. You cannot say, I have a "secure processor," so everything is safe, I am done. Because that cannot stop somebody from attacking you with a buffer overflow, a common kind of attack. For example, ransomware has buffer overflow attacks in them to take over and encrypt your disk, and then make you pay to get it back.We are compared to those two products. But when we look at a publicly-maintained CVE database of all attacks, maintained by Mitre Corp, out of the 84,000+ attacks that they categorized, compartmentalization only protects against a little over 2% of those attacks. Encryption and all the crypto types of things only stopped about 10% of those attacks. CoreGuard is capable of stopping 95% of those attacks. It is a completely new paradigm, a game changer. CoreGuard came out of a DARPA research program from 2010-2015. DARPA does not invest in things that have already been done or are easy. In fact, they sometimes only invest in things that seem impossible. Our CoreGuard technology originated from the DARPA CRASH program. Some of the Dover team members were among the 14 performers. We spent 4.5 years trying to understand how to stop those attacks that are wreaking havoc in our systems. That is why we are here, because we came up with something really good. After the DARPA years, we incubated for two years in a not-for-profit laboratory here in Cambridge MA, and officially spun out and became an independent company in mid-2017.Q: Is there any difference between the traditional chip manufacturing process and process using Dover's solution?A: No. The customer building an SoC brings in a processor (Arm, RISC-V, MIPS, Tensilica, Arc, or other) into their EDA tools and combines that with other IP for I/O, security, memory and so forth and then adds their secret sauce that differentiates them in the market. As they bring in the processor IP they will also bring in CoreGuard and place them side-by-side on the substrate. The entire process is identical with CoreGuard or without it.Q: What is the current funding status and plans for funding next year? What are your objectives for raising capital?A: We are currently raising a Series A round now, which we are looking to close in the first half of 2021. This round will allow us to significantly scale the company, expand the product's capabilities and expand sales and marketing initiatives to build awareness and demand. Post A round, our goal is to get CoreGuard into as many SoCs as possible, as quickly as possible. To do this we are going to expand sales and marketing and execute several initiatives to build awareness and demand for our CoreGuard technology. To date, we've been constrained with what we've been able to do with a small team and limited budget. Also in support of this goal, we are going to invest in the development of furthering CoreGuard's capabilities - opening up new opportunities and markets for Dover. These types of capabilities include, support for A-core, multi-cores, and Linux, as well as completing the integration of CoreGuard with additional processor architectures. In addition, we will be expanding our micropolicy suite to address the security threats relevant to specific vertical markets and we will begin the process of pursuing industry certifications.Q: What are the verticals that you are working with?A: We look at different markets and prioritize those that have large companies and those which are building their own chips. Industrial IoT, 5G infrastructure are two important markets that we are focusing on in 2020-2021, in addition to large semiconductor players. There are also other sectors, such as medical devices, consumer IoT, automotive, infrastructure, transportation, which we will explore in the future.Q: Will you enter the smartphone market?A: Some day. That will be very challenging. It's a very mature market. They are very constrained on power and physical space and have very complex software stacks. Our current implementation of CoreGuard is optimized for embedded systems with smaller software stacks. At present, we're focused on protecting embedded systems, which accounts for the vast majority of processors used in today's market.(Editor's note: Digitimes Research analysts Eric Chen and Jim Chien contributed ideas to the making of this interview)Dover Microsystems CEO Jothy RosenbergPhoto: Company

AImazing, founded in 2015, was groomed by Taiwan-based accelerator AppWorks. AImazing, which set up its headuqarters in Singapore and R&D base in Taiwan in 2017, was about to take off with its acoustic application when regulation changes in the Southeast Asian city state curbed its mobile payment services.Digitimes recently did a remote interview with AImazing founder Jun Ting to see how the startup has survived that setback in Singapore and how the coronavirus pandemic has changed its fortune.Q: How did AImazing survive the sudden regulatory changes that hit it hard in Singapore?A: Looking back at the footsteps of the company in the past five years, the acoustic application was AImazing's third product. We didn't really find the product-market fit until that third product, but this (regulation change) happened. Fortunately, these problems happened at the early stage. If it had been in the B or C round of fundraising, we would have had to face more investors and handle a bigger team.AImazing was a team of about 10-15 people at that time, and there was no way to continue because of the regulation change. The year 2018 was our low point, but we quickly found a new direction and now still have a team of six software development and data analyticsengineers in our R&D center in Taiwan to remotely support our services. Ou team members in Taiwan, Singapore and Malaysia total more than 30 people, and our market presence has expanded to the Philippines, where our local partners are running the business.Q: What's your latest products? What is the pain point that you want to solve?A: The idea of our latest product came from a meeting with Facebook in Singapore at the end of 2018. Facebook was exploring the possibility of working with AImazing, because we do transaction authentication for many electronic wallets. Visa and Master are the companies that have the best grasp of offline spending, but they can only know the total amount of money a user spent at a particular store, not the details of that spending. This inspired us to develop a new idea.AImazing's business model now is to integrate and analyze offline data to provide merchants with customer loyalty program solutions. Artificial intelligence (AI) and big data technologies are very mature and advanced, but the biggest problem is how to get the data offline. Offline is a very fragmented market, and in Taiwan and Singapore alone, there are probably hundreds of POS systems in operation. Our core technology is to get data from different POS systems and analyze it in a more efficient way.We built a platform and put all the offline data on it, and we also have our own app through which each company can get offline data. Of course, we still need to obtain the consent of the store first. Our service targets may even be restaurant chains with many outlets using different POS machines. They can put the data on our platform first, and then take the needed data from there. Customers can get clean and uniform data in a single docking. Even AC Nielsen is working with us because data on the cloud can also very confusing. We recently discovered by accident that 45% of our customers are using cloud POS systems.Q: It seems that AImazing has found a niche and a sweet spot, but in what ways has COVID-19 affected the company?A: Thanks to the COVID-19, we originally thought we were a "b2B" company and our customers should be big companies. Before the pandemic, we worked very hard to help banks and FMCG companies like Unilever, P&G and hypermarkets do online-to-offline data tracking in real time.After Singapore imposed a lockdown in April, we added a new feature to our product - a cashback loyalty program solution integrated with Facebook Messenger.Why do you need this service? Because many stores think they are already digital; after all, they are already using POS, and have joined many cloud-based delivery and cashback discount platforms, such as FoodPanda and Deliveroo. However, when the outbreak occurred, individual stores offered big discounts trying to attract customers to come and spend, but their promotional messages could not reach their customers.Each company in the market operates its own loyalty program independently. If a store wants to notify customers, they can only do so by phone, or they pay for their own advertising. Our loyalty program is integrated with Facebook's Messenger, which allows stores to advertise on Facebook and send directly to specific customers, enhancing the effectiveness of advertising.Q: What are your plans for future market expansion?The pandemic has allowed AImazing to be more aware of what customers really need. People working in the tech industry for a long time may tend to be carried away by their tech prowess, and they are prone to over-design and make products that are technologically advanced but ignore the real needs of customers. Data analysis may be useful for bigger companies whose operations teams can log on to get data and see reports. However, smaller stores may not be able to log on to see the data because of the digital gap or because they are simply too busy to do so.What this tells us is that we used to think that big data analysis is very powerful, but small stores find it very complicated and don't have the time to learn how to do it, thinking that they only need the simplest monthly report. When we put aside our own misconceptions and responded to the needs of our SME customers by sending simple monthly profit-and-loss reports through Whatsapp's API, the usage rate went straight up.The product has been so effective in boosting sales that hundreds of stores have joined our platform, and we've seen over 200% growth compared to the pre-pandemic period. We started the year 2020 with a team of 10 and now we have over 30 people. Our cash flow is very strong and will be fine for the next two to three years, and we plan to raise Series A in 2021 to continue our international expansion. But we won't put too much thought into raising capital; we're only looking for strategic investors who can help open another door.We would have expanded to Thailand and Indonesia in 2020, but due to the pandemic, we are postponing all such plans to 2021. In the future, after the pandemic has subsided, our expansion will still focus on Southeast Asian countries. My goal for 2021 is to increase the number of storesusing our services by 10-fold, with 3-5-fold definitely achievable.Q: How did you find your partner in the Philippines?A: Our partner was looking for such a solution, and his father is the co-founder of a big supermarket chain in the Philippines. We spent nine months bonding and getting to know the local market and environment. I traveled to the Philippines for a week every month to determine how to implement our solution there and to develop the business together.It's worth mentioning that the Internet environment in the Philippines is the same to that in China, where most people skip the PCs and go straight to mobile phones, so their computers are very, very old. The power supply and the network are very unstable, so we had to solve the problem of sudden disconnection of the network service. First we did a proof of concept (POC) to make sure the solution was available locally, and he eventually became our investor. So not only is he running the Philippines operation, but also helping implement and deploy the solution and working with international data analytics organizations.AImazing founder Jun Ting (left) and two of his core team membersPhoto: Company

Taiwan's investors should change their mentality about entrepreneurships, giving more support to startups, according to Wistron chairman Simon Lin.Lin pointed out that many of Taiwan's venture capital (VC) firms would require startup founders to shoulder part of the investments in their businesses as a way to show their commitment. But this puts the startup founders in a make-or-break situation where failures would leave them broke, Lin said.In Israel, where Wistron is also investing in startups there, most entrepreneurs operate their startups with VC. They do not have to worry about financial support for their innovative ideas, and although many startups have ended in failure, local VC firms are still willing to inject more funds into supporting entrepreneurships, Lin said.Wistron, which was spun off from the Acer Group in 2000, started reorganizing its structure to enhance competitiveness and formed a VC team to invest in its upstream and downstream partners, Lin noted.With Wistron becoming a major enterprise in Taiwan's ICT industry in 2010, Lin said the company set up corporate venture capital (CVC) hoping that its investment could help the company overcome its bottleneck at the time.From 2010 to 2020, Wistron's CVC injected a total of US$4 million into 50-60 companies. In addition to having outside teams evaluate investment targets, Wistron's management team also participated in many of the investment projects. As a result, many of the investments became strategic moves to assist Wistron's existing projects, Lin stated.At the moment, Wistron's CVC is still mainly investing in strategic items that fit the company's future projects, but Lin said he is hoping to have CVC expand the targets to new categories of the future.Wistron is a hardware maker and combining hardware with the new economy is necessary for the company to create added value, said Lin.Wistron chairman Simon LinPhoto: Digitimes file photo

The key to whether something sells or not is not the superiority of the product technology, but whether consumers are willing to pay for it. That's a key message Laura Huang, associate professor of Harvard Business School, delivered at SparkLabs Taipei Demo Day 4 held recently.Digitimes later talked to Huang - who was named one of the 40 Best Business School Professors Under the Age of 40 by Poets & Quants - about the skills of perception that enterprises need to cultivate and what it takes for startups to succeed.Q: Why did you attend the SparkLabs Taipei Demo Day 4? Was it because of the new research you are currently doing?A: Edgar (Edgar Chiu, managing partner of SparkLabs Taipei) and I had mutual friends, which is how we got to know each other several years ago. When Edgar was in Boston last year, he invited me to attend this year's Demo Day, hoping that I could share some insights to startup founders. My research focuses on how investors make decisions to invest in early-stage startups, analyzing the soft variables, such as gut feeling and perception, and the hard variables, such as financials, competitors, and scalability. My research seeks to quantify the soft factors that investors rely on to help understand how these variables relate to decision making. I'm continuing to extend this research and am currently also focusing on the topic of mindset, which I hope will provide insight to Taiwanese companies.In many cases, the key to success is not the superiority of a product's specifications and features, but the willingness of consumers to buy it. It is certainly possible to develop the most powerful product, but if the price is too high or those who can afford it do not want all the features, then the value of the product is not recognized.In fact, many products that have achieved leadership positions in the marketplace are not the most powerful. As I mentioned in my book, "Edge: Turning Adversity to Advantage," the successful businessperson must be able to recognize an adverse market situation, such as economic recessions, and leverage the situation to their advantage. These advantages are of your own making. Therefore, success cannot be achieved by just putting your head down and working hard behind closed doors - something that the Taiwanese are exceptional at and very proud of.Taiwanese are taught from a very young age to "work hard" and "be good," which usually means to not have too many opinions. Often, the Taiwanese put double the effort into their work just to get the same results. The main theme of my research explains that the key to success is not hard work, but to understand what people value and how they perceive you. This gives you a chance to flip any adversity to your advantage. If you can get twice the results with half the effort, why spend twice the effort to get the same results? Isn't the growth rate of 10X or 100X what startup companies are looking for?I have this understanding of Taiwanese culture because my parents taught me these values when I was a kid. However, growing up in the United States, I found Taiwanese values were very different than American values, which became a quick source of frustration for me. Both are valuable, and one is not better than the other. In American culture, self-awareness and social awareness are key, and you can't live in your own world if you want to be successful. You have to be able to understand others and know how others see you. There are dotted lines between these points, i.e., interpersonal relationships.By gaining a deeper understanding of how others perceive us and subtly subverting these views and stereotypes, we can create new advantages out of the limitations and obstacles created by an imperfect system. This is especially important for startups because the focus is not solely on products and services. In business school, we spend 80% of our time learning about marketing, products and services, financial accounting, competition, and so on, but spend very little time discussing people and perceptions. However, when we graduate and go on to start our own businesses, we actually spend 80% of our time discussing people, including customers, partners, and suppliers, to determine whether or not a product can be made and sold. There's a clear disconnect between what we learn in the classroom and what we actually do in the professional workplace. In Taiwan, I would like to see a transformation in the way we think about business education and the reality of the startup workspace as part of the new innovation ecosystem.Companies that embrace the mentality of believing superior technologies are key to success are prone to misinterpret results when doing market research. They often ask engineers to raise the bar for their products by focusing on pixels, lens resolution, and so on. However, I always tell my students that it is important to get customers' real perception of your product or service: you can't just rely on the data. For example, if you ask a consumer, "If my phone can take 30 photos or 99 photos in a row per second, which one would you prefer?" the consumer will most likely say 99 shots without having any idea of what that actually means. Then, the company will turn around and asks their product engineers to make a product that can take 99 shots without taking the time to see how the consumer will use the finished product. This market research does nothing to help the company understand what their consumers care about the most or what they want to achieve with the product. Companies rarely stop and consider if they really need that 99-shot series, or does the consumer actually care more about getting the best visual effect, regardless of how many shots can be taken? The factors that make consumers want to buy go far beyond technical specifications. We spend so much energy on specifications that we neglect the market feedback and user experience that entrepreneurs need to pay the most attention to, and how that feeds back into the design of our products and services.As we engage more with AI, big data, robots, AR and VR, it is easy to fall into the trap of technology and not realize it. Behind every VR goggle and IoT device, there is a user experience. "People" are the key, not technologies.Q: We have heard of examples of large corporations partnered with new startups in search of new applications for their existing products. What advice would you give to corporate venture capital (CVC)?A: CVC accelerators have a 95% failure rate because they don't use the right approach. These accelerators are usually well-funded, and they choose new ventures by cherry-picking: choosing Company A for its intellectual property (IP), Company B for its team's R&D capabilities, and Company C for a certain quality, and then they buy them all. This is because accelerators thought it would be more efficient to internalize innovation. They thought the startups, which have a certain understanding of their company's technology, would have a better chance of success if they used the company's technology and resources as a basis for their R&D. However, what they didn't see was that the startups they bought, such as the startups acquired by Google and Microsoft, recognized this cherry-picking strategy and tried to figure out what the big firms wanted, creating the product accordingly. When outsiders saw what they were doing, they thought it was uninspiring and not innovative because the product was often too similar to the parent company's already existing product. I believe that the struggle that many startups experience can actually be beneficial and lead them to making a product that will attract the most consumers in the market.This is because the products that come out of these in-house ventures will ultimately be tested in the marketplace. This is not to say that all CVCs are not effective, but they need to be designed carefully so that they don't end up catering to parent company's solutions. CVC accelerators are usually the latter, but innovation is often about trying to find a solution to a problem by drawing from multiple solutions that already exist, not the other way around. Innovation often starts with a technology application, and then someone goes back to find a problem and develop its solution.Q: Perhaps large corporations with a lot of past successes also need some level of reset and "learning to unlearn"? The pandemic has brought about many changes, perhaps their perceived successes will become a burden in the future?A: They don't need to unlearn everything, but they do need to know exactly what they need to "prune." This pruning is especially difficult for large companies because they are already successful, and it seems unreasonable for them to alter business practices that have already reaped rewards. While it's important to help them retain their core strengths, they must also continue to push innovation and learn new skills. As I said at Demo Day 4, large companies can be compared to trees in the sense that when they get to a certain size, they must be pruned in order to continue to grow. The hardest part stems from these companies knowing they need to prune, but not understanding which areas to focus on or what new areas to explore.Q: You strongly emphasize the need for companies to be sensitive to the "perception" of others. How do you go about strengthening that sensitivity?A: Perception is actually a skill. If you want to grow your muscle, you need to do certain trainings. I spend almost half of my book talking about how to develop these skills.Don't we spend our entire lives communicating with others in hopes that they will understand us, agree with us, and ultimately do what we want them to do? It is important to understand why, when, and under what circumstances the other person will say "no" to us. So in class, I have my students do an exercise: for a week, they need to have 10 people in 10 different situations say no to them. And for each situation, they have to write a short paragraph describing the situation and present in class.Interestingly, my students found that some people have a fixed communication pattern: when communication styles like tone of voice, words, and facial expressions changed in a certain way, the other person would say "no." They began to observe phenomena that they had previously taken for granted. It's important to train yourself to be aware of how you communicate in order to get your desired outcome. If you go directly to a person and say, "You are not giving me a promotion because I'm a woman," it is natural for them to immediately deny that accusation, even if it's true. When you understand others' perceptions of you, then you can flip any misconceptions to highlight your true talents and strengths, and have deeper and richer conversations and interactions with people.This ability is especially important for startups. Startups need to understand how their different customers, partners and investors perceive them so that they can understand what these different target audiences expect from them. This understanding enables startups to more effectively communicate their message and product in different ways to appeal to their various stakeholders. For example, you don't need to tell an investor a bunch of information about product specifications because they are more interested in learning about the business model. Contrastingly, consumers are quite interested in these product specifications, such as how the product works and pricing. Using your perception skills, you can better understand your different target audiences and communicate more effectively to get more done with less effort.(Editor's note: The Women in Venture series is a collaboration with Digitimes' strategic partner Anchor Taiwan, a platform to connect the world with Taiwan focusing on corporate innovation and cross-border expansion. The Women in Venture Roundtable is a network of 80+ female investors and bi-monthly sessions featuring world-class guest speakers. More info: Anchor Taiwan. For the podcast, see Spotify, Apple Podcast or Soundcloud: https://open.spotify.com/episode/0Dy6TKKLIazz8oaCW0oiiE.)Laura Huang, associate professor at Harvard Business SchoolPhoto: Shihmin Fu, Digitimes, November 2020

COVID-19 has been hammering the world, with many companies facing tremendous challenges amidst risks and uncertainties beyond their control. But many startup teams demonstrating their prowess at the recent Demo Day held by SparkLabs Taipei have seen double-digit growths.Edgar Chiu, founder and managing partner of SparkLabs, explained what it takes to overcome the adversities of the pandemic and how mature companies can remain open-minded and create the best ways to work with new businesses.Q: At SparkLabs Taipei Demo Day, you mentioned that 2020 is a year when many startups will be severely impacted by the pandemic. But the teams at the event stage have had brilliant results and rapid growths. How did they make it?A: For entrepreneurs, the unknown is an everyday occurrence, but during an epidemic, it is double uncertainty, and some entrepreneurs may be blaming bad luck, as they were ready to soar, only to be hit by an epidemic. I tell them the same thing: Be patient, keep your feet on the ground, and reexamine what your users need. Has it become a different form? Has it translated into more demand? For example, our investment in teams working on cloud kitchen, online education and Internet of Things (IoT) devices around year-end 2019 and early 2020 have all seen rapid growths because of the pandemic.The best example is the cloud kitchen team Just Kitchen. In early 2020 it was a team of only only 10 people but has now expanded to 70, opened eight satellite kitchens, and expanded to Taichung (central Taiwan). It will enter Hong Kong in the first quarter of 2021. Few would understand what cloud kitchen was in later 2019, but the pandemic has sparked strong needs for food delivery.Innovation teams should not only see whether there is demand in the market, but also whether their services and products deserve the money that people pay for them. Innovation is not just about satisfying demand, but also about establishing loyalty among customers would want to use them again and again, as if a habit.In addition, FunNow, which announced in late October its acquisition of Malaysian restaurant reservation startup TableApp, was supposed to be a company that would solve the problem of consumers making last-minute decisions to book hotels and local leisure trips for overseas travels. The pandemic has completely shut the door on cross-border travels, so FunNow has turned its business focus to domestic travel, and moved all of its overseas team members to Taiwan. In second-quarter 2020, FunNow's sales grew 60% year-over-year.Q: We saw a very mature team, Kneron, on Demo Day. Why would such a successful team need the help of an accelerator? When training such a mature team, what are the goals you want to help them achieve?A: Most of the SparkLabs teams are in fundraising rounds of Series A or Pre-A. It is worth noting that the definition of fundraising rounds in the US is that Series A is when the team has readied the app, wants to expand or has already entered one to three countries, and wants to become the market leader. Series B is for teams that are market leaders in more countries and want to raise funds to widen the gap with competitors those in second place. However, many startup teams in Taiwan are not adhering to the definition, and they move up one notch in the fundraising rounds every time they raise funds.Back to the essentials, these seemingly mature teams are actually at a stage where they are either expanding rapidly into international markets, or want to expand even more so. SparkLabs can help at that stage, and some of the partners have had experience in expanding internationally from the time when they started their own businesses. The biggest challenge in expanding internationally is how to modularize and replicate your products and services in other markets. The challenges of international recruiting, business development, and fundraising are new to the Taiwanese teams, but they are critical to the success of the expansion.How can we help new startups? Once we decide to invest in a new venture, we go back to the basics and do a 2-3 hour interview to find out what the biggest bottlenecks are for the new venture. Then it's a matter of prioritizing these challenges, looking at what needs to be accomplished in the next 3 months, and then setting for the schedule weekly progress accordingly. You need to do weekly reviews and tailor support and assistance to individual teams.But the hardest part is corporate culture. People are the heart and soul of a company, but many companies seldom think about how to build a sustainable culture and spirit. How do you design an incentive program that motivates your employees to be proactive and take action to achieve these goals? And while business founders take this for granted, they should also examine whether their actions are consistent with their words. So we have a lesson in corporate culture for these new entrepreneurs to figure out where they want to take their companies and how to implement the corporate culture.As an accelerator, we help companies get back to basics - setting goals and implementing user feedback surveys. Next, we continually tweak the products, release updates in between, and then work on growth in the third month.Mature entrepreneurs come here to reset, or unlearn, to go back to the basics, review their status and improve their products, and make a transformation before a breakthrough growth. Even though a startup may have already raised a lot of money, they often say that they are most afraid of sitting down with us to reformulate their business strategy, re-examine the needs of their customers, and push forward plans to expand their business to overseas markets. Because if you don't meet your goals and get the traffic you want, you won't be able to attend Demo Day.Our role is not only as investors, but also as coaches, mentors, and friends, following the process. What we often tell new entrepreneurs is that even if they encounter a major setback, it's better to encounter it now than later. The psychological quality of the founder can be seen in the two weeks before Demo Day, when he or she has to prepare the presentation in English and promote business growth.Q: There are more and more companies willing to work with startups. There have been many unsuccessful cases in the past. What advice would you give to CVCs?A: There are a variety of ways for companies to connect with new ventures and gain access to innovative energy. One of them is to create their own acceleratos, either by setting up a corporate venture capital to invest in external startups, or by setting up a special department to support them.However, there is a practice in the US that is worthy of reference for Taiwan, and not many Taiwanese companies have done so yet: investing in several venture capital funds at the same time to see if the solutions of the portfolio teams are suitable for their own company's problems, or even to combine the technologies of different teams to form new applications. They are constantly scouting for innovations and investing a portion of their R&D funds in venture funds.Many Taiwanese companies fall into the strategic trap of trying to find external partners when they want to innovate, that is, to build on existing products and technologies to improve product specifications and performance, or to drive down costs.These companies often start from a profit/loss (P/L) perspective, concerned about how much room for growth they can bring to their existing business, but rarely look for new applications in the "blue ocean" market to bring in new revenue streams; nor do they pay attention to whether there is new demand in the market, or even the next big trend.It is usually too late to invest in new ventures because there are more and more competitors, and then it is too late to chase after them or enter the new ecosystem. At present, Taiwanese companies are still thinking about this kind of cooperation, but they are still very single-minded, only looking for potential partners in Taiwan, when they should be also looking outside to see if there are suitable opportunities overseas.Before co-founding Whoscall with my partner, I worked as a consultant at IBM to help the manufacturing industry establish new business mechanisms. Why do so many large foreign companies come to Taiwan to work with SparkLabs? It is because Taiwan's technology capability is very strong. When more and more enterprises come seeking support for their needs, SparkLabs can help match them, and very often, Taiwan's technological innovation may be most suitable for foreign application.Q: How do you form partnerships with the foreign companies?A: There are several ways. The first is to come to our Demo Day, and then choose from our "menu," namely which projects or teams you're interested in. Companies in Northeast Asian countries have been unable to go to other countries to see new innovations or technologies because of the epidemic, so they are keen to communicate clearly what they want to achieve. These are overseas and foreign investors who have done their homework and have figured out how to work together before the conversation.The second way is to tell SparkLabs what their most important task for 2021 is, and see if SparkLabs can match a team to that. Most companies set their goals for the year in October or November each year.Some companies are very focused on their own field of business. For example, banks only look at FinTech, while manufacturing companies are much interested in AI-related solutions. Companies usually start with partnerships, investments, and acquisitions. Some companies will partner with SparkLabs directly, or their senior executives will become SparkLabs' mentors to foster rapport and synergy by coaching startups.Companies can also choose to invest in SparkLabs' startup fund. Because of this financial relationship, it is possible to see early on which teams SparkLabs has brought in, and to collaborate with them at an early stage of the new venture.SparkLabs Taipei founder Edgar ChiuPhoto: Judy Lin, November 2020

In terms of the maturity of the startup ecosystem and digital economy among Southeast Asian countries, Singapore is the leader, followed by Indonesia, Malaysia and Vietnam, according to AppWorks partner Jessica Liu.Southeast Asia is often associated with demographic dividend, especially Indonesia, Vietnam, and the Philippines. However, there are only 5.8 million people in Singapore, and when it comes to the digital economy, Singapore is more concerned with the development of core tech, such as AI, Industry 4.0 and blockchain, said Liu, who leads the accelerator's efforts in Southeast Asia and the region's burgeoning blockchain ecosystem.Singapore has reached a high level of digitization, while many of its neighboring countries still rely on paper, and their progress of digitization is relatively slow. As a result, most of Singapore's startup teams, unless they specifically target other Southeast Asian markets, are usually R&D and technology oriented, while others are usually very consumer oriented and focus on overseas markets, she said.On the surface, Singapore's small size seems to be an inherent disadvantage, with a total population of just over five million, while Indonesia has five million newborns each year. The small population makes it difficult for Singapore to find enough technology talent, which is why its companies are keen to recruit talent from around the world, with some companies running R&D centers, according to Liu.Singapore-based consumer-driven digital service companies target overseas markets with big populations, and even the entire Southeast Asia. AppWorks' portfolio of investments, such as Shopback and Carousell, which also operate in Taiwan, have been focusing on the Southeast Asian market since day one, she said.Smaller countries around the world with such an inherent disadvantage, such as Singapore and Israel, have turned a disadvantage into an advantage with a broader perspective. Take the two unicorns in Southeast Asia as an example. Grab is a Singaporean company and knew from day one that the market would not be limited to Singapore, while Go-Jek started out to serve the local Indonesian market and was forced to expand into other markets later on, Liu said.Go-Jek has a learning curve to overcome in Vietnam, where Grab has been operating for many years. They provide similar services, but brand building and consumer loyalty take time, she indicated.Indonesia's booming digital economyLiu said Indonesia is where China was 5-10 years ago, and Vietnam is probably where Indonesia was 5-8 years ago. Of the Indonesian startups that I saw in 2016 and 2017, 95% would be gone in about a year. But today, the survival rate has increased, and investors have a better idea of what the teams that have a better chance of success will look like. The Vietnamese startup ecosystem is now more like Indonesia's three or four years ago.AppWorks has invested in a number of startups in Indonesia, most recently in InfraDigital Nusandara, which provides a complete billing service; in HarukaEDU, which offers remote learning; and in Yummykitchen, which provides cloud-based kitchen service.HarukaEDU works with universities in Indonesia, and anyone who completes a distance learning course at one of these universities and passes an assessment test can receive a full university diploma. As Indonesia is made up of many islands, this service allows learners in remote areas to overcome the geographical barrier, said Liu. This is in line with the Indonesian government's policy to make higher education more affordable and to improve the quality of its workforce.The US-China trade war has been relatively beneficial to Southeast Asia's economy, reshuffling the supply chains, attracting capital to Vietnam and stimulating demand. And COVID-19 has accelerated the acceptance and popularity of digital economy, such as telemedicine and digital wallets in Southeast Asia.Indonesia's digital ecosystem stands out in Southeast Asia thanks to president Joko Widodo's strong support in the past 4-6 years, which has helped Indonesia make rapid progress in terms of policies, human resources and resource allocation, said LiuIs Taiwan's digital advantage becoming a stumbling block?Taiwan is in an awkward situation, said Liu. The domestic market is actually quite strong, with sufficient suppot from the upstream supply chain and downstream logistics, which provides mid-size startup teams a comfort zone beyond which they are reluctant to expand.Southeast Asian teams usually come to Taiwan to participate in AppWorks for three purposes: first, to join a community of entrepreneurs and get help from like-minded people to lower the learning curve; second, to expand the market, because Taiwan's digital market is booming and its overall scale is still larger than six major Southeast Asian countries combined, said Liu.The third is access to talent, as Taiwan has a much better supply of software engineers and data analysts in Taiwan than in these countries, she said. In terms of IT engineering, the quality and quantity of talent in Taiwan is much better than in Southeast Asia, so both Shopback and Carousell have set up R&D centers in Taiwan to handle overall digital support and operations.AppWorks partner Jessica LiuPhoto: Shihmin Fu, Digitimes, November 2020

Connectivity has become an essential commodity that determines how people live, work and play, according to John Mathew, co-founder and CEO/CTA of Cavli Wireless, an Internet of Things (IoT) module company that works with telecom operators to provision local IoT data in LTE/LPWAN technologies like NB-IoT, LTE-M on eSIM technology.Mathew believes connecting every object in the physical world to the Internet will solve a variety of problems and improve quality of life and human productivity - which he describes as the "true power" of IoT. In an email interview conducted by Digitimes recently, he talks about the company's innovative solutions and how he sees future trends and opportunities in IoT.Q: Tell us about Cavli Wireless and your team. What is its current valuation and are there fundraising plans in the near future?A: Cavli Wireless is a technology enablement company that aims to democratize the Internet of Things sector with our software-hardware product suite that includes a proprietary connectivity and device management platform, smart IoT modules, and eSIM solutions.We directly work with telecom operators around the world to facilitate local data connectivity thereby ensuring that any IoT product maker or enterprise in any industry, be it logistics, smart agriculture, industrial IoT, smart city applications can easily launch, build and scale their IoT solutions to any geography in the world in the most seamless manner possible.We believe that connecting every object in the physical world to the Internet will most seamlessly and efficiently solve a wide variety of problems and massively improve quality of life and human productivity.At Cavli Wireless we are simplifying the IoT connectivity infrastructure landscape with our end-to-end suite of smart embedded modules, eSIM solutions, and connectivity cloud platform.We are one among the first players in the industry to launch connectivity modules integrated with eSIM that are preloaded with global IoT data, which can be managed by our own platform - Cavli Hubble. Cavli Hubble gives remote backdoor access to smart modules & eSIMs, thereby giving businesses the confidence to scale seamlessly. There is a layer of intelligence and analytics for connectivity management for IoT solutions that are currently not on offer in the global cellular IoT deployment market.Our journey began in 2017 when four industry professionals aligned on the seemingly simple straightforward idea that enabling IoT should be made seamless, scalable and accessible to any smart product maker so that the industry accelerates and human life is greatly improved and efficient. But only when the team started decluttering the current industry problems, did we realize that though the idea is straightforward, the path towards achieving this reality was a complex one. Over the years, the team at Cavli Wireless has grown to become a multicultural community of motivated and curious engineers and business professionals who are on a quest to declutter IoT and make it seamless, secure and scalable one day at a time. Our 50-member team is spread across four regions including the Americas (the US), Europe (Spain ), Middle East (Turkey and Oman), and South Asia (India). We have an R&D innovation center set up in India, with sales operations in all other regions. Together, we aim to be at the forefront of innovation and deliver the best IoT solutions for our customers.I am afraid that I won't be able to share details regarding fundraising at this point in time, but we'd like to share that we are moving towards our Series A raise in 2021.Q: Tell us about the growth Cavli has been witnessing and your insights into the market trends.A: The global adoption of Low Power Wide Area Network technologies like NB-IoT and CAT-M1 and the emergence of 5G technology is certainly resulting in a massive growth in the number of connected IoT devices in the next few years. Cavli is foreseeing this transition and is on a mission to ensure our solutions are relevant to the needs of the cellular IoT industry.Cavli's product suite is the result of the amalgamation of three different verticals - telecom technology, embedded electronics, and cloud engineering. Over the years we have worked with top industry majors spread across the globe at upstream and downstream ends to ensure our IoT CaaS solution always remains few steps ahead of the game. To emerge as the top-5 global solution providers for cellular IoT implementation, Cavli is building an aggressive global sales operation that would make strong inroads into newer markets. With a passionate team that believes in the vision and an agile R&D process built within a very scalable business model, we are very sure that we not just coexist but capture a good mindshare of the market, where product makers will consider "Cavli tech" as one among the most trusted options to get solutions "smart connected."In the next five years, 790 million cellular IoT connections are set to surge to over four billion connections solely on cellular IoT. The market will be worth US$125 billion for IoT connectivity management alone and we are very confident that we will create a dent in the industry with our innovation in technology and business model.Q: In view of the LPWAN cellular-IOT global market distribution, the top-10 telcos take up 85% of the LPWAN IoT connection (in which China enjoys 60%). This is a winner-take-all market, and it may take a long time for the LPWAN to get widely adopted without policy advancement. That's why it depends on government subsidies so much for the time being. Since telecom operators collect limited revenues from IoT connection services, the demand from operators is still very weak for now. How do your products and services address the pain point of the LPWAN IoT market?A: As far as the customer is concerned there are several hurdles or pit stops which makes the journey to figure out seamless IoT connectivity an arduous one. Some of the challenges would be: Connectivity hardware - which involves evaluating multiple embedded hardware product lines based on network technology type); application software/firmware - whether to flash the software on a separate processor Vs. programming the software into the connectivity hardware; network access - eUICC (eSIM) vs single profile SIM technology. Data contract negotiations with telcos or connectivity resellers; R&D effort for the integration of network access technology and connectivity hardware; figuring out telco certifications and other device certifications; and device-cloud connectivity management integration for data subscription managementWith our integrated solution and launch of Hubble99, deep technology players like us are now getting to work closely with telecom operators to bring about a holistic offering that could also push the telecom's enterprise sales efforts as well. The higher purpose of Hubble99 is to democratize IoT - the very problem which you have hinted in the question - that right now industry is dependent on telecom operators to drive large scale IoT adoption, which is not a burden that the telecoms need to carry on their own. By working hand-in-hand with preferred telecom operators in each region, we are making life easy for multitudes of IoT SMBs who would otherwise take even up to a year to hit the market with their solution. The telecoms we are partnering with for Hubble99 are already seeing the results where relatively smaller enterprises with not so huge volumes are moving to market faster. The more the number of such players, it adds up to significant revenue for our telecom partners giving them all the more reasons to ensure continued service quality for LPWAN in the deployed markets.Q: As for the future growth potential for the unlicensed spectrum IoT market, you seem to have focused on cellular-IoT, but has not developed products such as LoRa and SigFox which applies to the unlicensed spectrum technology. However, this unlicensed spectrum market share is bigger than NB-IoT/LTE-M and may have a bigger potential to grow in the future. What is your view about it and plans to meet the demand of that market?A: As of now, our focus is on cellular IoT which we are convinced that it would be a sizable market to prove our worth given the four billion devices that are forecast to go live on cellular networks over the next five years. We have an agile R&D team that is working towards introducing innovations in the cellular-based IoT space in the near future. In fact, we have already started working with complimentary network technology players like Sigfox to define hybrid solutions for certain smart city use cases.The alternative unlicensed LPWAN market will surely grow, but we are confident of being able to cement our position as an IoT enabler for licensed LPWAN especially in application and use cases where security layers are going to be very high.After all, we are in the space to make the "technology for the connected tomorrow," and I believe with the advent of 5G in the near future and with its accelerated proliferation, many unlicensed network technology-based applications could also sway its way towards cellular.Q: What is your plan to transition to the 5G IoT module? It is probably unavoidable for a transition to a 5G NR SA network. But currently, most of the IoT modules that are based on 4G LTE (namely, NB-IoT/LTE-M) may be phased out and make all investments obsolete. What is your opinion about that trend and how do you address it?A: We were set to launch a test network at our Innovation Center in India by Q3 2020, but have slightly delayed it to mid of 2021 owing to the shift in priorities due to the pandemic.When realized, our 5G test network platform will enable startups and companies in the IoT segment in and outside India, prepare their products and prototypes to be ready for the latest technologies.This test network will be available with 5G SA and NSA modes (supporting sub-6Ghz bands), alongside LTE NB-IoT and CAT-M for the LPWAN side of the IoT. Access to this test network will be totally free of cost, fast-tracking different use cases, allowing you to take your smart products globally anywhere completely pre-tested and compliant.As 5G will bring many new test requirements and challenges by the use of SDN/NFV and cloud services, the technology can also be used for creating new test solutions. We've already partnered with two global chipset players to build 5G based modules.OEMs and ODMs can make use of this platform to test their prototypes in real-world scenarios and to leverage their product development. We will be supporting companies with our modules which are in testing - C200-Series 5G IoT modules/modems and the C1X, C3X, C4X series modules/modems for the LPWAN networks, which are currently in production.Q: It seems that the prices of cellular IoT modules have been falling in recent quarters. That could imply wider adoption in the future, but cost-effectiveness will be a challenge to you. What's your strategy to overcome the problem?A: It is true that the price of cellular IoT modules has been dropping over the years. From being a specialized component, cellular modules in its standalone form are now becoming a commodity. However at Cavli, what we bring to the market is not simply hardware. Cavli modules are "intelligent modules."In a span of three years, Cavli Wireless has managed to bring to the market a product suite that is at par or surpasses the capabilities of the incumbents in the IoT space. We recently launched Hubble99, the world's first truly subscription-based IoT offering with zero hardware cost. For an irresistible package price of US$0.99/month/device, IoT solution players get the eSIM integrated connectivity hardware of their choice of technology (on LPWAN/LTE technologies - NB IoT, LTE-M, CAT 1, CAT 4, 2G,3G), pre-loaded IoT data subscription (ranging from 5MB to 30MB), and a robust connectivity-device management cloud platform - Cavli Hubble all bundled into one single package. In 2016, Ericsson predicted a near future where the LPWAN module will be priced below US$5 in the market. With Hubble99, Cavli has brought it down to US$0.Q: Your slogan attracted my attention: "Cellular IoT will never be the same." Tell us how you re-imagine IoT in the next 10 years.A: IoT connectivity enablement is a multi-stakeholder ecosystem that includes connectivity providers/telecom operators, application providers & system integrators. For SMBs, OEMs and enterprises in IoT, figuring out and activating IoT connectivity has not just been the most important cog in the wheel, it has also been a very distorted one. To top it, for most sensory application products, the cost for connectivity hardware at the edge is a significant component of the total product cost, thereby bringing up the initial capital investment involved. This has in fact affected decisions for scale and has been a major reason why many IoT SMBs and benefactors of smart city projects have been cautious about scaling up. There have been projections galore, but not enough on the ground deployments to match it up. Out of the 50 billion devices that were expected to get connected on cellular by 2020, as per predictions by Ericsson, so far only nine billion seems to have found its way through.Thus to unlock the real potential of IoT and to get the next 10 billion devices connected, the currently fragmented ecosystem needs to be re-engineered, consolidated and thus re-imagined from start to end.The last few years have seen a tussle for prominence and dominance amongst the major connectivity technologies that have been grazing the IoT deployment landscape. While LoRaWAN has seen steady growth thanks to the low latency feature and resultant low power consumption, it has stumbled when it comes to quality of service and data packet delivery. Sigfox though it establishes security layers which are commendable with strong firewalls, when it comes to scalability, it has its limitations. WiFi over time has found its permanent seat for indoor home, personal automation and wireless access use cases, with Zigbee taking its fair share of indoor applications from WiFi. Wading through this ensemble, cellular IoT began its ascent up the curve with the definition and advent of cellular LPWAN - NB-IoT and LTE-M which have sealed whatever service gap there was for LTE 4G in terms of enabling low power, low latency, wide coverage applications.Despite the evident advantages that cellular IoT has over its unlicensed spectrum peers like LoRaWAN, Sigfox, Zigbee, and the likes, in terms of security and quality of service, to be the preferred choice for a wide range of industrial and outdoor IoT applications, it has had its share of jinxes. Cellular IoT has had its critics go back and forth on questions around its actual scalability. Theoretically, it can be a non-steered connectivity option but the real challenge for the IoT solution players has been how they can manage the hassle of liaising with multiple telecom operators for the regions they wish to deploy their projects in. So far, from how the market has evolved, it is safe to assume that this has not been an easy ride for the countless IoT product and solution companies trying to expand beyond home turf. The other jinx that has pulled back cellular IoT has been the burden of the initial cost of associated connectivity hardware including smart modules and SIM solutions. With Cavli Wireless, cellular IoT is shedding the shackles that have held it back as it enters a new age.Just like food and fuel, connectivity has also become an essential commodity, one that will determine how we live, work and play. At Cavli Wireless, we believe that connecting every object in the physical world to the Internet will most seamlessly and efficiently solve a wide variety of problems and massively improve quality of life and human productivity. This is the true power of the Internet of Things.(Editor's note: Digitimes Research analyst Benson Wu contributed ideas to the making of this interview.)

Taiwan will focus its next wave of technology development on materializing smart healthcare, precision medicine, smart city and digital data governance by leveraging its robust semiconductor prowess to incorporate AI, IoT and blockchain technologies into diverse vertical applications, eventually embracing a smart new future featuring high-security Internet of Everything and thriving digital economy, according to Tsung-Tsong Wu, the country's minister of science and technology.Wu told Digitimes in a recent interview that the 5G traits of high bandwidth, low latency and wide connectivity, as well as 4k and 8K display screens and camera lenses with increasingly high resolution can be combined to support versatile applications in the near future, particularly allowing cost-effective remote surgery and high-performance security surveillance systems with clear facial images and multi-screen displays.Wu said the gradual availability of privacy protection regulations, growingly mature tech applications and exchanges, the relaxation of regulations governing healthcare data uses, and the incorporation of medical technology sandboxes will further help usher in massive innovation-driven business opportunities in Taiwan.He stressed that humanity and scientific technology should be closely blended and his ministry hopes to join forces with other ministries in integrating resources and building ecosystems from diverse facets such as infrastructure construction, frontier technology, talent cultivation, digital economy and digital government.IaaS, PaaS, SaaS solutions requiredBig data collection, modeling, prediction, transmission and storage as well as information security protection for diverse vertical domains all require the establishment of IaaS (Infrastructure as a Service), PaaS (Platform as a Service) and SaaS (Software as a Service) systems in addition to joint efforts by members in software and hardware ecosystems, according to Wu.He noted that the National Center for High-Performance Computing (NCHC) at National Applied Research Laboratories (NARLabs) under his ministry has teamed up with academic and industrial segments for many cooperation projects seeking to strengthen IaaS, PaaS, and standard data specs and promote them among industries.With its tech sectors and innovation ecosystem developing vigorously under the tech island chain architecture, Taiwan has been striving to create a friendly environment for startups and innovations by connecting technology and humanity, aiming to facilitate exchanges of international talent and build a fast and safe Internet of Everything with strong innovation momentum. This has attracted major cloud service giants Microsoft, AWS and Google to set up large-size datacenters in the country seeking to better capitalize on Taiwan's key position in Asia, Wu indicated.In line with around-the-clock services by these web giants, Wu said, the NCHC must also build a trust mechanism encouraging tech startups to utilize national cloud network services to process and transmit data under tight information and domain knowledge security and full data privacy.Digital data governanceIn terms of future digital data governance, the focus must be on smart healthcare offerings including generic drugs, new pharmaceuticals, medical supplies, and health and well-being promotion, and the strategy for biomedical big data concerning disease prevention, diagnosis, medicine, nursing and long-term care services should be even more comprehensively formulated, Wu said.Citing human biobank as an example, Wu said after a biobank is integrated, biomedical startups should be allowed to join efforts to catalyze a new wave of digital transformation at small- and medium-size biomedical firms and medical service and retail firms.In cultivating startups, Wu stressed that it is imperative for innovative technologies and startup teams to grow up initially in innovation parks with easy exchanges of international talent. And after their C-round fundraising, the National Development Fund can step in to assist, and venture capitalists and field experiment partners can also join forces to help the startups build international connections, he added.Wu said tech innovations usually involve high risks, which in turn also imply great momentum for success. He urged Taiwan enterprises to nurture the culture of allowing errors, withstanding frustrations and fighting on despite repeated defeats, so that they can stand out in certain pivotal segments in the next wave of tech innovations.Taiwan science and technology minister Tsung-Tsong WuPhoto: Michael Lee, Digitimes, November 2020