Crossover Talks: AI & the Economic Cycle
Highlights from Taichung Session
As we enter the countdown to 2023, major international organizations have painted a bleak picture for the global economy in the coming year. The impending significant challenges are testing the resilience of businesses and the intelligence of future market strategies. Profet AI, a leading player in the field of AI applications for the manufacturing industry, held the Crossover Talks: AI & the Economic Cycle forum in Taichung on November 24. David Huang, former CEO of YAGEO, Peter J. Sher, Dean of School of Interdisciplinary Design of Feng Chia University, Jack Chao, Executive Vice President of AUO Digitech, Charlie Huang, Manager of Smart Factory Platform & Solutions at Gogoro, Justin Chen, Justin Chen, CIO of MAXXIS, along with Vice President for Industry-Academia Collaboration at Taiwan AI Academy, Benjamin Kuo. These experts shared their experiences and insights with attending enterprises, collectively brainstorming solutions for the challenges ahead in this era of economic cycles.
Profet AI collaborates with customer partners to pave the path to success.
Kicking off the event, Jerry Huang, co-founder and CEO of Profet AI, stated that Profet AI has long focused on the manufacturing industry, and has a deep understanding of the pain points of manufacturing customers while navigating through the current international situation and market challenges, as well as how to implement AI applications. AI technology is highly significant to Taiwan's manufacturing industry, so Profet AI has been committed to transforming AI into a tool which everyone can use. This enables production and sales units within enterprises to apply AI in practical scenarios within a week, following just two hours of education training. Jerry Huang pointed out that after customers have implemented AI in practice, there is a growing demand for industrial AI application topics. Therefore, Profet AI has also launched Ready To Go Application knowledge base that encompass various industrial application topics. Following the tools and knowledge base, services tailored to meet customer needs for methodology and core team workshops have also been introduced.
Continuous listening and meeting customer needs is one of Profet AI's core capabilities. Jerry Huang commented, "Since the beginning of this year, due to the uncertain economic prospects, we have been approached by second-generation and third-generation business owners or C-suite level executives of companies on advice for how AI technology can help in the deployment of business strategies and make key decisions. This led to the birth of the "Crossover Talks: AI & the Economic Cycle" series of forums, which we hope to brainstorm solutions with industry and academia experts and businesses." The inaugural Crossover Talks held in Taipei in August this year, invited Hsing-Chien Tuan, Honorary Chairman of Innolux Corporation, Professor Chia-Yen Lee, and a number of second-generation business owners to share their experience. The response was overwhelming, and the second session saw the invitation of more than 30 new-generation business managers to visit AUO Longtan Plant to observe how AI can be actually implemented based on their own manufacturing competitive edge.
Crossover Talks has reached the Taichung stop today, inviting David Huang, Founder of Jensen Capital and former CEO of YAGEO, and several veteran industry-academia leaders to share their suggestions under the economic cycle and practical experience in AI implementation.
▲ Jerry Huang, Co-founder and CEO of Profet AI, welcoming the guests
Believing in the economic cycle, pumping in investments during lull market and taking a novel approach to fight future battles
The Taichung session invited David Huang, former CEO of YAGEO and current founder of Jensen Capital, to share his experience in leading YAGEO through the previous financial tsunami and turning losses into profits. David Huang is a firm believer in economic cyclicality and believes that the economy will rise and fall. When others see a crisis, good managers see opportunities. David Huang took the procurement of sintering furnaces as an example where the procurement cost of a sintering furnace during the market boom period was NT$120 million, but it only cost N$60 million with free installation thrown in during the market slump period. The contrarian investment at that time also allowed YAGEO to quickly seize business opportunities when the economy rebounded and to achieve unprecedented success. It reflected that the market slump period is the best time for companies to improve the quality of people, machines, raw materials, methods and environmental conditions. The most trying times are the moment when enterprises should have the courage to invest, review, and reinvent their companies.
David Huang also shared that whilst the formula for the first-generation business owners to gain experience is time plus money, the second-generation can no longer follow this path. David Huang gave an example, "When we planned to invest in a mold company, we found that there were no mold craftsmen in Taiwan because all the mold craftsmen had moved to the Chinese market many years ago. What should I do if I don't have a mold craftsman? Therefore, companies should think about how to re-establish the operating mode instead of relying on the same people to deal with tomorrow's challenges as generals always fight the last war." To this end, David Huang stressed that the new generation business owners must first invest in talents, "The return in investment for upskilling talents is always more than 100 times larger than that of investment in equipment." Only the correct individual is willing to learn.
Secondly, David Huang suggested that business managers must respect the power of data, invest in the right AI tools, and combine the right data with domain knowledge to generate intelligence to make decisions. In addition, business owners should have the right attitude when investing in AI tools and adopt an attitude of AI assisting companies to be sustainable during onset investment, rather than simply applying AI to fragmented projects.
▲ David Huang, former CEO of YAGEO and current founder of Founder of Jensen Capital, recommended that business owners should believe in the economic cycle and introduce AI applications as soon as possible.
The formula for successful AI implementation: domain knowledge plus algorithms and computing power while having the attitude of "making a mountain out of a molehill"
Peter J. Sher, Dean of Feng Chia University School of Interdisciplinary Design, has long assisted Taiwan's industries in digital transformation, including the machine tool industry that Taichung is most proud of. Peter Sher said that it was a battle of human wits for the first-generation business owners, and the second generation must rely on computers, because we are facing many unknowns in today's complexity of the times.
Peter Sher cited AUO as an example to illustrate the successful implementation of AI. When enterprises adopt AI, domain knowledge in various industries is vital on top of computing power and algorithms.
Peter Sher stated that mentoring team would not possess specific vertical domain knowledge, and it must integrate the expertise of corporate customers to be successful. The reason why AUO can become an exemplary showcase of AI application is the ability to open up internal processes, so that different teams with domain knowledge can collaborate with the AI mentoring team for cross-departmental collaboration to achieve good results. Within 18 months, AUO successfully applied nearly 30% of the AI projects to the production line, which is a commendable effort.
▲ Dean Peter Sher of Feng Chia University School of Interdisciplinary Design reminded enterprise domain knowledge is key.
Over and above integrating domain knowledge and cross-departmental collaboration, companies should not think about overnight success when introducing AI. Peter Sher reminded, "Don't romanticize introduction of AI and start with small steps. Only by scaling up can success rates increase. Don't think that introducing an AI can turn the company around as no such thing exists." Peter Sher also recommended that small and medium-sized enterprises in the central part of Taiwan can regard this economic downturn as a turning point, start to refine their systems through the Profet AI solutions, so that they have a chance to emerge as winners when the global industry is restructured.
Gogoro uses AI to extract the essence of massive data for decision optimization
Charlie Huang, Manufacturing Director of Gogoro, said that the company generated more than 400 million pieces of data every day, including test data and process data. The main purpose of collecting such massive data for analysis is to improve daily operation and management needs such as production line yield, as well as to obtain process and user feedback to optimize design and improve product reliability. This information is crucial to the lifeblood of Gogoro today. In addition, after analyzing these data, clear presentation of these insights to business managers for decision making is also important. The Profet AI tool, which condenses the six standard deviation process, is very helpful in assisting in providing information for business decisions.
▲ Gogoro Manufacturing Director Charlie Huang uses Profet AI tools to help optimize business decisions
Charlie Huang pointed out that Gogoro is currently accelerating the expansion of overseas production bases, and huge amounts of data will be generated from Indonesia, Vietnam, the Philippines and the entire Southeast Asia manufacturing bases. The maintenance cost would be astronomical if they were to rely solely on headquarters for IT support. As a result, Gogoro is now adjusting its IT structure to meet the needs of different regions or countries.
Charlie Huang, Director of Manufacturing of Gogoro, who has professional manufacturing experience across mechanical and aerospace engineering, commented that at present, the company is also recruiting fresh blood from different fields and different expertise. He echoed the aforementioned suggestions of David Huang and Peter Sher, recommending participants strengthen their domain knowledge and leave the IT tools to experts.
Cheng Shin Rubber optimizes manufacturing parameters through AI to reduce testing costs and improve yield rates
Cheng Shin Rubber had long been an enterprise Taiwan can boast of and has actively driven digital transformation in recent years. Justin Chen, Associate Director of Cheng Shin Rubber, pointed out that Cheng Shin began its transformation as early as 2015, transforming into a "lights off plant" through smart systems. The coronavirus pandemic, which began in 2019, further spurred companies to integrate legacy systems and adopt public cloud applications. When Cheng Shin was about to start adopting AI, more than 95% of the people actually did not know what AI is, had no idea on how much server room resources would be needed, nor how to implement it. In the past, the production management center can obtain information from the MES system regarding temperature, pressure and other information in manufacturing processes. However, it is very costly to repeat relevant tests and calibrate to reach formal production. For Cheng Shin, a single test may result in several hundred kilograms of rubber scrap. Through the collaboration with Profet AI, Cheng Shin continuously optimized parameters, successfully reducing the number of experiments, effectively improved the yield rate, reduced the test time and the scrap quantity.
▲ Justin Chen, Associate Director of Cheng Shin Rubber, shared his digital transformation experience to significantly reduce testing costs through AI.
Justin Chen, who is also a second-generation successor, said that the most common problem encountered by the new generation successors is convince the previous generation of business owners to adopt new tools or methods to improve the efficiency of the companies. Taking cloud applications as an example, it is a major challenge for traditional industries. Only by using new technologies and techniques to provide quick and meaningful data as evidence and yield short-term achievements can the second-generation successors win the trust of first-generation leaders.
AUO's secret to AI DNA: AI talent development and process improvement
AUO, which can be regarded as an AI model student, teamed up with Taiwan AI Academy to nurture more than 1,000 AI engineers in the past five years. Chao Shin Min, Executive Vice President of AUO Digitech, said that the high awareness of AI and process improvement of enterprise personnel are critical to AI introduction. "It is also because of such investment and cumulation that AI has become part of AUO's DNA, successfully reducing labor costs such as handling and testing, increasing output, and increasing productivity by 30%.
Chao Shin Min supplemented the increase in output through AI by increasing yield and preventing equipment failures. In the past, complex production equipment parameters often had to be adjusted manually or based on experience of skilled technicians, but such data may not necessarily be the true optimization, but only the best numbers at the point in time. However, through AI, AUO was able to increase the yield of many high-end products from 92% to 97%. While yield rate increased by a mere 5%, this optimization from 92% to 97% needed to rely on AI parameter optimization to make it happen. Chao Shin Min said that if there is equipment failure in the production process, it will cause production disruption, and the cost of these failures can be calculated. AUO uses data science methods in conjunction with AI models to conduct Prognostic and Health Management (PHM, real-time monitoring mechanism and device health pre-diagnosis and management system), which is also a service currently provided by AUO Digitech.
▲ Chao Shin Min, Executive Vice President of AUO Digitech, highlighted the key points to the introduction of AI by companies.
During the forum, Chao Shin Min also introduced SPIIDER, the new tool launched by AUO Digitech, which can assist companies to capture data without equipment shutdown and plug and play data retrieval. After obtaining the data, supplemented by the Profet AI platform, PHM can be effectively performed to optimize formula among other tasks.
AIA suggested companies treat data as a daily routine, embed AI into the corporate DNA to facilitate change to take place.
The annual conference of Taiwan AI Academy (AIA) just concluded last month, and AIA Director of Industry Development Benjamin Kuo also shared observations during the event. Kuo echoed AUO's proactive approach to nurturing AI talent. He remarked, "When it comes to talent development, it's really about changing the mindset of business owners. This change only occurs when business owners change their mindset towards AI, embed AI into genes, and treat data as a daily routine. The company can move forward only when the business owner changes." This is one of the reasons why the theme of this year's AIA Annual Conference is "AI in Action - Supply Chain Restructuring and Talent Development under Geopolitics Tension".
Another phenomenon observed at this year's annual conference was the restructuring of the entire supply chain. Benjamin Kuo pointed out that in the current geopolitical situation, everyone's role is interchangeable and should think about how to adapt. The best way to adapt to the crisis is to constantly improve one's strength and enter the market in crisis. It includes strengthening one's domain knowledge, and then complementing it with computing power and algorithms of AI.
▲ Director of Industry Development of AIA Benjamin Kuo suggested that business owners change their mindset towards AI and view data as a daily routine.
Benjamin Kuo stated that the requirements for introducing AI are different between large enterprises and small and medium-sized enterprises. Large enterprises need more application cases and also need a change in the mindset of business owners while the requirements for small and medium-sized enterprises are simpler, which means that everyone can use AI tools. Through tools such as Profet AI AutoML, 80% of the problems of companies can be solved, and the remaining 15% to 20% is to be addressed by in-depth collaboration with the academia or by the agility of the companies to invest in their own development.
Benjamin Kuo reiterated that in this era, regardless it is supply chain restructuring or other relative issues, the most important thing is that companies must first prepare talents. All business owners can first nurture themselves into data and AI savvy, so that when the talents are available, they can discuss the transformation blueprint and what tools to use to move forward, in order to bring the business operations to greater heights.