Previous Chinese population studies had only involved thousands of people, mainly Han Chinese from the eastern coast.

The largest of its kind for the Chinese population to date, the study was conducted by Shenzhen-based Beijing Genomics Institute and published in the international academic journal Cell on October 4.

“The study is significant to the development of China’s medicine and health management industry, which has relied on foreign technology and studies for a long time because we didn’t have enough genome data from the Chinese population to work on,” said Xu Xun, president of BGI Research and the study’s lead author.

It discovered the genetic structure of the various ethnic groups, found six genes that display significant differences of people across latitudes, and identified the gene flow patterns between Europeans, South Asians, East Asians and Chinese.

For instance, people from the southern part of China have developed stronger immunity against malaria than those in the north, but a gene related to fatty acid metabolism has a much higher frequency in most of the northern provinces.

In addition, people in Gansu and Qinghai provinces, a key area of the Silk Road, share more similar genes with Europeans.

It also discovered new associations among Chinese between genes and conditions such as birth defects, infectious diseases and cancers.

About 78 per cent of studies in the world about the relationship of gene and diseases are based on European individuals, Xu added.

The latest data cover all 31 provincial-level administrative units in the Chinese mainland and represent Han Chinese and 36 ethnic minorities.

“In the next three years, BGI plans to further expand the database to 1 million,” Xu said. BGI started its Million Chinese Genome Project in 2016 and Xu’s study represents its first phase.

Genomic study on a national level has been widely regarded as a significant factor for a nation’s core competence in the biomedical field.

In the United Kingdom, the leader in the field, one study of 500,000 samples has finished and another project based on a population of five million was announced earlier this month.

BGI researchers admitted there is still a gap between China’s genetic research and that in Europe and the United States, but they said it is narrowing.

After the world’s first experimental cashier-free grocery store Amazon Go opened in the US earlier this year, it did not take long for the trend to land in Korea.

An increasing number of customers are going cashless or even credit card-free, as payment options through palm scanning or via app have become easier at offline shops. But there is also a pronounced trend among Korean customers who enjoy a traditional yet convenient way of purchasing products in comfort: home shopping.

South Korean home shopping companies, which have mostly depended on shopping promotions through TV air waves, are testing the waters with advanced technologies.

More South Korean retailers are starting to leverage data in the use of artificial intelligence. Moving on from allowing customers to simply look for information on goods, place orders, make payments and track their delivery status using their television remote controls, home shopping companies are now striving to offer advanced shopping experience with the use of information and communication technology.

According to data from the TV Home Shopping Association, the size of the TV shopping market grew to 15.8 trillion won ($13.8 billion, Bt451 billion) in 2017 from 10.6 trillion won in 2013.

Currently, home shopping companies GS Home Shopping, CJ O Shopping, Hyundai Home Shopping and T-commerce companies K Shopping and Shinsegae TV Shopping are the major players in the market.

Shop as if in the store

One of the overarching trends that home shopping companies are starting to adopt is augmented reality technology.

Home shopping and T-commerce companies are in the lead, making use of such technology to strengthen its online shopping experience and make it more realistic.

NS Home Shopping has teamed up with telecom KT to launch a service called AR Market, where a customer can take a 360-degree look at the product on TV as well as through a mobile app.

“We believe that IT development is one of the most important aspects of competitiveness for the shopping service and tried to overcome the limit of online shopping by launching AR Market by using KT’s advanced AR technology,” said Chung Won-deok, head of NS Home Shopping’s mobile operation team.

KT has also said it will add more home shopping companies on the list for collaboration and work on 3D contents by combining VR and AR technologies to set up more realistic T-commerce market.

VR fitting service

Hyundai Home Shopping is working with the China-based company of AR makeup app Meitu, and rolled out a service here in May in the first attempt in the industry.

Online operator TheHyundai.com is applying its AR makeup fitting service to eight cosmetics brand, including Estee Lauder and Shuemura, so online shoppers can test colors of eye shadow, lipsticks and blush with the front-facing camera, the firm said.

The company added such service, which now has over 10,000 users, will become a competitive point amid market trends focused on fast delivery. The service will |be applied to 20 more beauty brands.

“What customers want these days is not just a price discount, but an unexpected and interesting shopping experience and service. (Our service) will become a new tool for us to integrate offline distribution and IT to define a new shopping experience,” said Lee Hee-joon, a manager at TheHyundai.com’s e-commerce team. Hyundai Home Shopping has also unveiled a VR fitting service that allows customers to try fashion items on with a 3D avatar. While the program is on air, a customer can simply click on the menu on the screen and try the items on the avatar.

“It has now become available to try on the product as if a customer is inside the shop while shopping and help them be immersed into the virtual fitting service,” said an official.

Purchase by voice

Voice shopping has made both product ordering and payment easier.

In March, CJ O Shopping introduced a voice payment service where a customer can make a payment by connecting his or device to SK Telecom’s Nugu mobile application.

To shop via voice, some services do not require a separate AI speaker.

Lotte Home Shopping, which also sells its products on its mobile application, has launched a service inside the mobile app that recognizes a customer’s voice to make the order and payment. For example, all one needs to do is to click on the microphone icon inside the app and say, “search air conditioner,” then “put it to the cart.”

Amid fierce competition among online retailers and mobile shopping apps, home shopping and T-commerce will need to make clever use of TV and mobile apps, industry insiders say.

“While mobile apps have become one of the most popular and convenient shopping platforms for customers, the traditional home shopping and T-commerce companies will need to find a breakthrough by actively making use of technologies that can differentiate their platforms,” said an official from CJ O Shopping.

Some of these technologies were unveiled last week at Naver’s annual developer’s conference Deview 2018 held in Seoul, organised to showcase the Korean portal site operator’s future tech development direction each year.

Stressing “location” as the most important cue for context in a physical space, Naver Chief Technology Officer Song Chang-hyun said Naver’s in-house research lab is developing and testing out an array of location and mobility-based technologies that will serve as the backbone of new, real-world services.

“If technology can fix mobility problems in the physical space, it will inevitably lead to the development of new services. And the underlying platform technology for any location and mobility problem is mapping,” Song said in a keynote speech during the conference.

In line with this vision, Naver plans to open source the Naver Maps Enterprise application programming interface to developers next month. The soon-to-be-free mapping API includes including point of interest data and navigation data, which most smaller ICT firms find difficult to build and maintain.

The Korean internet giant also introduced for the first time the extended Definition & Dimension Map, or xDM, platform which includes high-dimension mapping, localization and navigation technologies to guide people and self-driving vehicles in not just outdoor, but also indoor settings.

For instance, Naver utilised its xDM platform to build an AR navigation system that guides people through indoor spaces without GPS access, using just a smartphone camera. On a smartphone, walking direction signs and messages would be overlaid on top of real-life pathways as captured by the camera.

This service is slated to debut soon, via a partnership with Incheon Airport. This week, Naver signed a memorandum of understanding with the airport’s operator to launch an indoor AR navigation app that would guide airport passengers to their respective destinations inside the airport.

“We plan to open source our xDM platform in the form of API and SDK so that they can be used for various location-based service development and mobility research in the future. We plan to continue improving xDM so it can grow into Korea’s top location and mobility platform,” Song said.

In addition, Naver Labs is developing technologies for self-driving cars, including a high-definition map that combines high-definition aerial maps and mobile mapping devices to build intricate, real-time maps even in crowded urban centers where GPS access is partly blocked out.

The company also showcased the latest updates in its robotics technology development, which had taken center stage at last year’s Deview conference.

M1, an autonomous indoor mapping robot that creates intricate 3D maps while moving by laser-scanning its surroundings with a 360-degree camera, and Around, a self-driving robot that navigates itself by accessing M1’s mapping data via the cloud, are undergoing commercialisation procedures via a partnership with Hyundai Heavy Industries Holdings.

Aircart, an electric book cart equipped with a module that amplifies human power to make pushing the cart significantly easier, is preparing for commercial deployment at local libraries in partnership with Samsong Caster.

Naver’s robotics lab is also continuing research and development efforts with Seoul-based Korea Tech Co to develop robotic cable-driven arms with humanlike sensitivities called Ambidex.

Weighing just 2.6 kilograms, the robot’s core motors are placed near the torque and transfers power down to the lower arms and fingers via flexible, lightweight cable wires.

Looking ahead, Naver plans to showcase its fleet of robots at the Consumer Electronics Show in Las Vegas in January next year, marking its first-ever attendance at the global electronics convention.

The AI platform was built by studying the vast amount of research on central nervous system disorders and developing optimized algorithms in partnership with SK Holdings C&C and its AI platform Aibril.

AI-supported drug development is being pursued by many drug developers worldwide, as it can significantly reduce the required development time and lower the risk of a failure.

Among the many local pharma firms in such pursuits, SK Biopharm claims it is the first to launch a viable AI drug development platform and apply it to real-life new drug research and development procedures.

SK Biopharm’s AI drug discovery platform consists of three technologies – AI modeling including drug substance behavior prediction and drug design proposition; chemical property data storage; and AI modeling data storage.

The AI modeling portion utilises machine learning and deep learning to check a drug’s properties, including absorption, distribution, metabolism, excretion and toxicity, as well as to predict its action mechanism.

Based on the results, the AI discovers hidden patterns and properties in the drug substance to propose a new combinations. This new drug combination technology is unique to SK Biopharm’s AI drug discovery platform, it said.

The chemical property data storage portion collects various lab testing data for a chemical substance, and translates it into a format that can be used to train the AI model, so they can be accessed by researchers via a data portal.

SK Biopharm and SK Holdings C&C said they are currently reviewing the possibility of opening their AI drug discovery platform for broader use by Korea’s pharmaceuticals industry.

Industry interest in AI-powered drug development remains high here. The Korea Pharmaceutical and Bio-Pharma Manufacturers Association and the Korea Health Industry Development Institute co-hosted the “AI Pharma Korea Conference 2018” in Seoul last week.

The conference brought together experts from local and global companies leading AI-based drug development technologies, including IBM Watson, Numerate, NuMedii, Innoplexus, twoXAR, Standigm, Syntekabio and 3BIGS.

China’s largest telecom equipment maker and smartphone vendor Huawei Technologies Co Ltd has unveiled a range of artificial intelligence chips, in what is arguably its biggest-ever push to expand its presence in the semiconductor sector and to compete with US companies such as Nvidia Corp.

The move is also part of Huawei’s broader effort in AI. The company is working to offer a full-stack technological portfolio to support all AI application scenarios.

Huawei said one of the new processors, known as the Ascend 910, boasts the greatest computing density in a single chip in the world, whose performance can beat its rival product Nvidia V100. Ascend 910 is scheduled to hit the market in the second quarter of 2019. And it will be deployed on Huawei’s cloud solutions to drive data centres.

The new processor is part of broader chip series Huawei released on Wednesday. Another chip, known as the Ascend 310, is designed to drive low-power computing with high efficiency.

Xu Zhijun, rotating chairman of Huawei, said these chips will not be sold as standalone products. “Instead, we will develop products such as AI servers, AI accelerators and autonomous driving solutions based on these processors and sell them to clients.”

These chips are based on an architecture called DaVinci, developed by Huawei. The tech heavyweight has already used intellectual properties from Chinese startup Cambricon Technology to design AI chips for its smartphones.

“Cambricon’s IP is quite good, but it is not enough to meet our needs to have chips that can support all AI application scenarios. We want to have an architecture that can deliver both extremely big computing capacity and extremely low-power consumption,” Xu said.

“We will give our processors to partners for trial use,” he added.

Wang Yanhui, secretary-general of the Mobile China Alliance, said it will not be easy for Huawei to take on Nvidia just by winning in terms of technology parameters.

“Nvidia has developed a well-established ecosystem with large pool of clients,” Wang said. “As a newcomer, Huawei’s approach of selling chips and services as a package may be more appealing to clients than selling them standalone.”