Samsung opened an artificial-intelligence research centre in New York in September 2018, adding another location to a network spanning Korea, the United Kingdom, France, Russia, Canada and Silicon Valley. The New York operation was led by Sebastian Seung, an executive vice president at Samsung Research, and concentrated on AI for robotics.
The expansion formed part of a plan to employ about 1,000 AI specialists by 2020. It showed how large electronics manufacturers were building international research networks to bring machine learning into products, factories and services.
Robotics Connected Software to Physical Products
Robotics was a natural focus for Samsung because the company operated across consumer electronics, appliances and manufacturing. Intelligence that could perceive, plan and act had potential inside factories as well as future household products.
Physical systems are more demanding than a purely digital recommendation. A robot must work with uncertain sensor data, respect safety limits and respond in real time.
Research therefore needed expertise in computer vision, control, learning and human interaction. A dedicated centre could assemble those disciplines around long-term problems that individual product teams might not have time to investigate.
New York Added a Distinct Talent Market
Global companies place research laboratories near universities and specialist communities. New York offered academic strength, a diverse technology ecosystem and access to industries that could create new applications.
Recruiting 1,000 specialists was ambitious in a market where technology companies, startups and universities competed for the same people. Multiple centres widened Samsung’s hiring pool rather than requiring every researcher to relocate to Korea.
Distributed hiring can retain local expertise, but a network only creates advantage when teams share methods and results effectively.
Each Centre Had a Different Emphasis
Samsung’s AI locations were designed around complementary research areas. New York focused on robotics, while the Cambridge centre in the UK explored human behaviour, emotion recognition and user-centred communication.
Specialisation can prevent an international network from duplicating the same work. It also gives each laboratory a clearer identity when recruiting senior researchers.
Coordination remains difficult across time zones and organisational boundaries. Shared infrastructure, regular exchanges and product partnerships are necessary if discoveries are to move from one centre into Samsung’s broader business.
Manufacturing Offered Early Value
Samsung’s factories contained expensive equipment, complex logistics and large quality-control operations. Robotics and machine learning could improve inspection, material handling and production flexibility.
Industrial deployment demands reliability. A model that works in a laboratory may fail when lighting changes, components vary or equipment ages. Engineers need monitoring and fallback procedures that keep production safe.
AI also changes jobs. Workers should participate in redesigning processes, receive training for new responsibilities and understand how automated decisions affect performance measurement.
Consumer Robots Required Trust
Robotic systems inside homes would operate near people and collect information about private environments. Convenience alone would not justify unrestricted sensing.
Privacy-by-design could keep sensitive processing on the device, minimise retention and give users visible controls. Physical safety required conservative behaviour and clear limits on autonomous action.
Samsung’s scale could help integrate hardware, processors and software into complete products. The same integration increased responsibility because one company could control much of the system experienced by the user.
Global Research Was a Competitive Strategy
Samsung described its centres as pivotal to creating new value across products and services. The network was intended to compete with technology companies investing aggressively in their own AI laboratories.
Corporate research can pursue ideas before a commercial market is certain. It also risks becoming disconnected from deployment if incentives favour publications without product transfer, or short-term features without foundational work.
Successful organisations create routes in both directions: researchers learn about real product constraints, while product teams gain access to tested advances.
The Centre Signalled a Broader Shift
The New York opening came when AI was moving from a specialised software field into the strategy of hardware manufacturers. Companies expected intelligence to become a layer across appliances, vehicles, mobile devices and industrial systems.
Samsung’s decision to build a global network acknowledged that no single location held all the necessary talent. Robotics in particular required collaboration between machine learning and established engineering disciplines.
The lasting question was not how many centres opened or researchers were hired, but whether the network could convert knowledge into dependable systems. A global footprint creates opportunity; shared execution, safety and useful products determine whether it creates value.
