AUTOMATION DEVELOPMENT SUMMIT 201805.03.2018

China is looking into a bright AI future

On March 5 2018, the 2ndedition of the Automation Development Summit organized by Messe Frankfurt and STM took place during SPS China. Here the summary of the presentations.

Georg Stieler, Managing Director China of STM 

Georg Stieler, who also moderated the one-day-event, provided an introduction to the topic. His keynote was structured in three parts:

1. Machine, platform and crowd – the major paradigms behind the rapid change our economy is going through. Mr. Stieler said that sometimes it might be valuable to take a step back to gain a holistic perspective on a certain subject. The emergence of the Internet has brought change at an unprecedented scale and speed to several traditional industries. 

  • With the increasing capabilities of Machine Learning, and Artificial intelligence, this trend will only accelerate and affect more business activities. During this process, platforms are becoming more important at the expense of products. Even the mechanical engineering industry has been relatively unaffected from this development due to highly specific products and an often limited number of vendors, this is about to change: With machine and production data becoming more valuable, platforms that run analytics software will gain in importance. Also, changes in sales markets, e.g. the move from owning prestigious individual cars to platforms of shared autonomous vehicles will have a big impact on the automotive industry and its suppliers. 
  • The increasing convergence of different disciplines and industries requires knowledge, which is often not within the organization. Therefore, the ability of companies to leverage the knowledge crowd, using decentralized structures of collaboration, is becoming more and more important. Technologies like blockchain can help to realize this, e.g. through smart contracts. 

2. Next, Mr. Stieler talked about existing digital business models for manufacturing companies in China. Even most companies are still somewhere between Industry 2.0 and 3.0, there are some companies who show a greater willingness to experiment. Chinese makers of construction machinery were early to introduce locking and unlocking features to penalize customers who didn’t pay their leasing rates. Some robotics and drone companies were offering robots as a service (RaaS). System integrators were entering joint ventures with industrial companies to get automation projects running and share the rewards of them afterwards. Industrial clouds were just emerging, foreign vendors need a domestic majority partner.

3. Mr. Stieler concluded with an overview of the special ecosystem of the Chinese Internet and its implications. Within the last three years, China had moved from a copycat to a global innovator. There are over 100 cities with 1m inhabitants, creating a huge pool of data. Inefficiencies in a lot of traditional industries. Together with relatively low concerns about data privacy, this created an ideal environment for new ideas to scale. Mobile payments, personal finance, mobility as well as food and retail had been disrupted this way. When it comes to Deep Learning, China was actually leading in facial recognition and voice recognition today. And this whole sector enjoys a high priority with the government in Beijing. 

Mr. Stieler went on to ask, which inefficiencies might be reduced next? Automation will certainly help to solve quality issues, workforce fluctuation and increasing labor costs in manufacturing. Other sectors offering plenty of space for disruption are logistics, healthcare, and transportation. Concerning the Industrial IoT, there is certainly a China gravity in many industries these days. Most of the machine data will be generated and processed in China. It remained to be seen, whether the comparatively low spending for enterprise and government software in China would prove as a long-standing disadvantage or as an opportunity for leapfrogging (as it has happened in the consumer field). 

Despite the difficult environment for international companies in China, major technology companies were still intensifying their efforts in this market. Google and Facebook were even coming back. Foreign companies should therefore try to benefit from China’s emerging strengths in digital business models, and machine learning in particular. Domestic companies should stay open for influences from abroad.

Jan-Michael Haehnel, Legal Counsel, Burkardt & Partner

Many of the Industry 4.0 solutions promoted by German companies in China have never left the status of a blueprint. Offering intelligent, interconnected smart manufacturing solutions is actually quite tricky in China’s regulatory environment.

Mr. Haehnel therefore illustrated the implications of the current legislation on two typical use cases: A predictive maintenance service and HR services offered by international vendors.

Despite the Internet being global, the legal perspective is still drawing national borders – with a Chinese business license as its entry ticket. While the jurisdiction of China technically ends at its rim, its factual reach may affect affiliated entities of an overseas business within China. Providing cloud services in China presupposes the establishment of Foreign Invested Telecommunication Enterprises (FITE) with a supposable requirement in registered capital of 10 mn RMB as well as proven expert knowledge in one’s field of activity, always in dependence on the favor of the MIIT. 

The recent Cybersecurity law asks all network operators, referring anyone with more than two devices connected, to store essential data as personal information within China. Any export of such critical data abroad may only be granted for specific business purposes, and after a security cleaning. Hence, both cases mentioned in the beginning require the set-up of both the corresponding legal and organisational structures in China.

In regards to VPNs, the old stance of a general prohibition sustains to be true. Further encryption may only be used if it is a registered version and a key is provided to the government, in case of an investigation into criminal activities.

Guan Zhen, Principal Technical Evangelist from Microsoft

Being a so-called “pure technology provider,” Microsoft possesses a unique position in the value chain of Industrial Cloud. Its Azure technology powers GE Predix, which enables smart manufacturing. 

In regards to predictive maintenance, Microsoft depends on tight-knit  cooperation with traditional machine manufacturers as IT companies have little knowledge about the inner workings of sophisticated machines and how to collect insightful data for industrial demands. 

According to Guan Zhen, Principal Technical Evangelist at Microsoft, AI has to be better defined and still has a long road ahead to reach a general state. On a practical level, Microsoft has helped Kohler to redefine their intelligent bathroom.

Li Baomin, Vice GM from CASICloud

Li Baomin, Vice GM from CASICloud, illustrates the fundamental change that has upended the global corporate power game in just ten years by showing the ten highest valued companies, in 2007 as well as 2017. By 2017, seven out of ten of the world’s largest are in fact high-tech companies, with data fueling their business models. Apple, Alphabet, Microsoft, Facebook, Amazon, Alibaba, and Tencent are spearheading this surge of tech.

Li Baomin believes that within the next ten years the value of data will exceed that of oil, as comparable exploitations and extractions of value should be feasible, whereby the potential of oil is deemed to drain off soon.

On a more worrying note, it is critical to realize that any collected data may give insights, not only on the current status of equipment, yet also on details of the actual product being produced as well as the processing techniques. Mr. Li advises governments and companies to be very careful about their data collection.

He suggests concentrating the efforts on in-house data generation across various corporate processes including R&D, purchasing, manufacturing, logistics, auditing, and sales. 

Huakang Li: Sales Director at Sany Rootcloud

According to Huakang Li, sales diurector at Rootcloud, the quality of datain China does not yet satisfy the existing demand and lacks behind the quality available in the USA and Europe. Changing such quality issues is key for China’s progress in industrial internet.

Rootcloud tries to incorporate digital technologies into traditional industrial manufacturing to turn production plants into smart factories. It already has provided digital solutions to players in the CNC machine tools, casting, insurance, textile machinery, agricultural machinery, as well as dryer machine industries.

For example, its cooperation with CNC machine toolcompanies creates an industry cloud that can monitor the efficiency and quality of productionin real-time. 

The solution installed for the casting industry purportedly builds a casting ecosystem that connects companies and enablescollaborative manufacturing and supply chain services.

Vincent Y. Wang: Chief Innovation Officer at Wanxiang

In general terms, Vincent Wang, Chief Innovation Officer at Wangxiang, believes that the sharing of big data among companies has promoted the digital transformation of industry, but also led to data abuses and illegal access of competitors. 

Wangxiang has developed various blockchain solutions to protect the security of data. Combining blockchain with their Edge Xtechnology may potentially re-define the way people and companies share as well as access data. 

According Mr. Wang, a reform in the industrial cloud service sector is yet to come, as the digitalization of traditional manufacturing is much more difficult compared to other industries. In contrast to the widespread internet of social networks, where billions of individuals automatically generate data patterns every day, industrial manufacturer firstly have to install sensors at all their machines to enable data generation. Such investment has to come from informed managers, who are aware of the practical gains of IoT, yet also know the imminent pitfalls scattered along the path towards smart factories.

Mr Wang believes that an external nudge, for example via mandatory government regulations, may propel the state of industrial internet as companies are forced to invest.

John Gao, CIO Schindler Asia Pacific, Senior VP Schindler China

Schindler’s digital strategy can be subsumed via the Internet of Elevators and Escalators (IOEE). It is envisioning a global outreach in the elevation segment and has therefore signed an anchor partnership with General Electric as well as a cooperation agreement with Huawei. 

In most of its projects, the smart sensors are being supplied externally, yet Schindler provides its extensive, highly specialized knowledge of elevator 

technology, drawing from decades of experience around the world. While elevators are always uniquely adapted to different environments, whether in its speed or resilience towards adversity as aggressive levels of humidity. On a global scale, however, the use cases are quite similar, creating an ideal breeding ground for groupwide learnings via big data. Further integration of Tencent’s WeChat may turn the one to two minutes elevator ride into a fun and shareable social experience. 

Schindler’s plan ahead aims to provide elevators, equipped with a multitude of sensors, transmitting performance values to its cloud. A centrally located engineering control rooms shall be able to remotely deploy local service personnel over-the-air and inform them through its app about imminent technical errors and problem-solving advice for a seamless elevation experience for customers. 

Dr. Mohamed El-Refai, IBM Greater China CTO

What has started in 2007 within the in-house R&D department of IBM, has led tot he creation of an AI machine, called Watson. 

IBM distinguishes artificial intelligence from mere mathematically refined calculations, with the former simulating the human process of decision-making. For example, in the process of deciding between taking the elevator or the stairs, we go through the mental steps of „observing-interpreting-evaluating-deciding.“ Mohammed el-Refei himself has written his Ph.D. on such intelligent systems, yet in the 80s, enthusiasm about AI was not as commonly widespread as nowadays, although the critical paradigms have been the same.

Watson’s application field ranges from healthcare, automotive, and aviation, to engineering, and manufacturing. While the IoT of IBM may not have been able to track down the missing Malaysia Airlines Flight 370, it purportedly is being applied in more than 90% of all flights globally, connecting and thereby warning other planes. The return can be seen in more efficient and comfortable flight experiences, leading to happier customers and higher revenue. 

In the engineering segment, IBM’s Watson can alleviate the loss of knowledge of retiring engineers by aggregating their experience, analyzing the critical components via Natural-Language-Processing and providing it to inexperienced staff at the site of deployment. 

While other AI providers may cross-share the generated knowledge, IBM leaves the clients’ data in their possession and thus is only benefiting from use cases as Watson's algorithm gets more refined as it processes more data.

Contact us for more information:

Stieler Enterprise Management Consulting (Shanghai) Co., Ltd.

Phone:  +86 21 6078 0385

E-mail: info@stm-stieler.de