Microsoft keeps investing in Taiwan, having set up an IoT innovation center in October 2016, an AI R&D center in January 2018. It alsto introduced an AI 100 project in April 2019.
For AI technology development, Taiwan's hardware manufacturing capability and strong education in science and engineering are the best nutrients, according to Microsoft Taiwan chief operating officer Hedy Ho.
Q: For AI 100, the Taiwan government hopes to expand it to AI 200. How is the project progressing?
A: AI 100 aims to develop 100 AI-based solutions for use in various industries by the end of 2019. Expanding the target number from 100 to 200 is very challenging, but Microsoft Taiwan is confident it can make it by the end of the year because we see strong willingness to adopt AI solutions from industries, such as KGI Bank in the financial industry, Wistron in the manufacturing industry and Water Resources Agency in the government sector.
According to a joint survey of enterprises in Asia by Microsoft and IDC, 60% of Taiwan's firms consider AI to be important but only 30% plan to adopt AI solutions, with the gap higher than the average for the Asian region. There are three reasons for the firms that are interested in AI but have yet to take actions: Scattered internal data; lack of relevant experts; no appropriate IT partners. The AI 100 project seeks to break through the three obstacles.
Whether the target number is 100, 200 or even 500, the key point is that enterprises have to adopt AI solutions to create successful cases, then create influences, and eventually lead to forming of an ecosystem. In order to attain the goal, Microsoft Taiwan not only offers AI solutions but also actively promotes their adoption among industries and talent training.
Q: What effects are expected from enterprises adopting AI solutions?
A: The effects are in four aspects. The first is better customer relationship management (CRM). For example, KGI Bank has used AI-based assessment of clients' risks and O-Bank hopes to use AI to better understand clients and enhance connection with them.
The second is to optimize enterprises' internal operating processes. For example, Microsoft allows employees to use an app to apply for leaves and pay taxi fares.
The third is to shorten time taken in developing products or services. For example, Wistron has adopted AI-based AOI (automated optical inspection) for its production lines, enabling faster and more acurate quality inspection. And inspection data have been used to quicken development of new products.
Strengthening employees' productivity is the fourth. For example, customers' responses and downloads are classified for use by marketing staffs to beef up marketing.
Q: How can Microsoft's success in transforming its own business operation help Taiwan-based enterprises transform business operation using AI-based solutions?
A: Of the 10-20 enteprise clients I've come into contact since I became Microsoft Taiwan's COO six months ago, the questions they have asked most frequently are about Miscrosoft's transformation. What I have gathered during my 10-year work at Microsoft is that transformation is more about corporate culture and employees than about technology. Therefore, enterprises' adoption of AI-based solutions has to begin with corporate culture and employees.
For example, most of Microsoft Taiwan's executives have scored high grades obtaining Azure 900 certification. Now the number of Miscrosoft Taiwan employees passing Azure 900 certification is the second highest among all company branches worldwide, next only to the US. The certification enables employees to understand Azure 900 and additionally results in internal cohesion.
Q: As AI focuses on application and Taiwan is strong in hardware, how can Taiwan cultivate AI experts?
A: Taiwan's hardware manufacturing capability accumulated over the years offers great potential for developing AI technology, especially for development of IoT or edge computing. For example, of the 17 enterprise partners for Azure IoT PnP platform, Microsoft-developed cloud computing-based IoT plug-and-play platform announced at Microsoft Build 2019, six are from Taiwan, namely Compal Electronics, Wistron, Delta Electronics, Askey Computer, Aaeon Technology and VIA Technologies.
Besides, Taiwan's strong education in mathematics and physics is also a plus, enhancing logical thinking and creating momentum for technological innovation. In addition, Taiwan's abundant good-quality data on medical care are helpful for developing medical AI technology, for example, accumulation of X-ray films which have been interpreted by competent medical professionals.
Q: What is Microsoft Taiwan's next step following AI 100?
A: The first is to boost adoption of AI-based solutions among industries such as transportation, aviation and retail which already have large volumes of data available for analysis. For example, shipping companies have reduced shipping costs sharply by using AI-based analysis. Second, training experts. The final step is to extend promotion of AI solutions from large-size enterprises to small- to medium-size ones. As small- to medium-size enterprises generally lack resources to develop and maintain customized AI-based solutions, Microsoft Taiwan plans to develop modular "AI as a Service" to facilitate their adoption at relatively low costs.
Microsoft Taiwan chief operating officer Hedy Ho
Photo: Shihmin Fu, Digitimes, May 2019
Article translated by Adam Hwang
