Skip to content
  • KOSPI 2784.90 -49.39 -1.74%
  • KOSDAQ 912.50 -30.35 -3.22%
  • KOSPI200 370.04 -6.04 -1.61%
  • USD/KRW 1193.6 3.20 0.27%
  • JPY100/KRW 1,049.36 5.20 0.50%
  • EUR/KRW 1,353.54 6.78 0.50%
  • CNH/KRW 188.04 0.39 0.21%
View Market Snapshot
Artificial intelligence

Korean AI startups blaze trails in overseas markets

From fighting wildfire to creating virtual news anchors, domestic players explore diverse markets

By Jan 13, 2022 (Gmt+09:00)

Alchera's SaaS (Software as a Service) allows speedy smoke and fire detection
Alchera's SaaS (Software as a Service) allows speedy smoke and fire detection

Artificial intelligence-powered image analysis company Alchera Inc. is working with PG&E (The Pacific Gas and Electric Company (PG&E) in the United States to build an early stage detection system for wildfires.

Prior to teaming up with the Seoul-based startup, PG&E had already stationed 140 cameras throughout the state of California but still had difficulty detecting smoke and fire. Alchera came in and incorporated its AIIR (Artificial Intelligence Image Recognition) Firescout solution to 46 of those cameras. 

According to the company, the technology is similar to teaching a phone's camera to recognize one’s face. Instead of faces, however, it is teaching cameras to tell the difference between a cloud, fog or smoke through a continuous learning system. 

The AIIR Firescout software studies images around the clock and can differentiate between city lights and car headlights. Utilizing data from the broadband camera network across the US, Alchera developed a program that can estimate and track smoke and dispatch firefighters accordingly. 

The US Fire Administration estimates the direct financial loss from wildfires in 2019 to have reached a whopping $14.8 billion. In the record-setting 2018 wildfire season, Yale Climate Connections estimates direct and indirect damages to have amounted to $150 billion. 

Alchera CEO Hwang Young-kyoo said he plans to expand the business across the US. 


AIMING FOR 55% OF MEDICAL DEVICE MARKET SHARE 


Lunit Inc., an AI software company for cancer screening and treatment, received MDSAP (Medical Device Single Audit Program) certificate late last month. The MDSAP, operated by International Medical Device Regulators Forum (IMDRF,) is a global audit program that leverages regulatory resources to manage an efficient, effective and sustainable single audit program. 

A certified manufacturer is allowed for a single quality management system audit in place of multiple audits for five countries, including the United States (FDA,) Canada (Health Canada,) Japan (MHLW,) Australia (TGA,) and Brazil (ANVISA.) The certification is widely considered evidence of a medical device manufacturer’s commitment to product quality and regulatory compliance.

According to the Korea Health Industry Development Institute, the medical device market size in the above five countries is estimated at $223.2 billion, which makes up about 55% of the global medical device market.

“MDSAP certificate signals global recognition for an excellent quality management of our products,” said Brandon Suh, CEO of Lunit. “We are looking forward to speeding up global market expansion with our AI solutions both currently commercialized and in the pipeline.”

Other homegrown AI medical solutions providers include Vuno, which received a manufacturing and sales permit from the Taiwanese government; and JLK, which received emergency approval from The Pharmaceuticals and Medical Devices Agency under Japan’s Ministry of Health, Labour and Welfare.

DeepBrain AI creates virtual humans in a diverse number of occupations
DeepBrain AI creates virtual humans in a diverse number of occupations

AI ANCHOR

Korean companies are also leading in producing virtual personalities such as news anchors and influencers. 

Virtual influencer Rozy, developed by SIDUS Studio X, is probably the most famous example in Korea. There are such examples in nearly all countries -- for example Lil Miquela in the United States and Imma in Japan.

At CES 2022, DeepBrain AI introduced a virtual anchor named Jennifer. 

Modeled after a news anchor at Arirang TV, it can speak both English and Korean fluently and copies the human’s intonation and gestures.

When a user inputs a written script, the AI human delivers it on camera as a human reporter or anchor would. Several media outlets and US universities showed interest in the product.

In June 2020, DeepBrain AI established its Shanghai office and has been providing AI anchor technologies to local broadcasters since then. 

CEO of DeepBrain AI Eric Jang said he would “strive to create diverse opportunities by exploring new business options beyond Korea."

Write to See-eun Lee at see@hankyung.com

Jee Abbey Lee edited this article.
Comment 0
0/300