Analytics & Information Management Lab
Minseok Song (Industrial & Management Engineering)
Nothing is complete until one puts it in its final shape. This old adage echoes the sentiment of many businesses that are eager to leverage the abundant data that continues to ever expand. While data, perceived as a key corporate asset, is exponentially increasing in the Big Data era, translating and analyzing it into innovative processes across industries proves to be a daunting task. As a way to address this challenge, Artificial intelligence (AI), blockchain and other latest information technologies are gaining attention as viable solutions.
The Analytics & Information Management Laboratory (AIM) headed by professor Minseok Song at the Department of Industrial & Management Engineering, POSTECH, teams up with Samsung Electronics, POSCO, Samsung C&T, Seoul National University Hospital, Korea Football Association and other diverse partners with the aim of pursuing data-based innovation. The Lab assists businesses facing difficulty in fully harnessing the massive quantity of data they possess by clearly defining what is holding them back and then delivering a broad array of actionable solutions to move them forward. In case the existing methodologies fail to present solutions, researchers at the Lab go the extra mile in helping to create new ones.
AIM is primarily focused on process mining that aims to harness data for process improvements, thereby connecting linkages between data and process. For instance, POSCO, Korea’s leading steelmaker, handles a host of processes in its steel manufacturing operation. In such case, even a temporary setback affecting a specific process could potentially generate sizable losses. Analyzing these processes to predict and even reduce potential stumbling blocks could greatly innovate the whole manufacturing process.
Another emerging area of research concerns the application of AI across diverse industries. Starting last November, Samsung C&T’s online shopping mall launched AI-powered fashion coordination services developed by the Lab. The AI system is able to recommend various fashionable outfits drawing on its learning of over six hundred thousand sets of data for actual apparel coordination. So, for example, when a potential consumer picks out a trendy top, this AI coordinator recommends trousers and shoes that complement it perfectly based on its learning outcomes, unlike the conventional approach of simply showing what others have chosen to match that specific top.
AIM is extending its research scope from industrial applications to other broader areas. In partnership with hospitals, the Lab analyzes and optimizes the traffic flow of patients and aids in the simulation of treatment processes. Normally, doctors gather prior to surgery to establish standard guidelines to identify necessary treatments. AIM, again, steps in to contribute to this guideline-setting process on the strength of its data. Recently, the Lab has been working with the Korean Football Association to harness data to explore ways to help players improve their competitive edge.
The ultimate goal set by AIM is to ceaselessly commit itself to a brand of research that delivers industrial innovation. One of its long-term goals is to translate a host of ideas created at the Lab into startup companies. ‘Puzzle Data’, Korea’s sole process mining solution provider established by leveraging AIM’s technology, has successfully attracted KRW 2.3 billion in Series-A funding and already employs more than 30 people.
Head of Lab
Science Building Ⅳ 408