NTU, Ericsson and KTH Team Up on AI-Driven 6G Positioning

Nanyang Technological University (NTU) in Singapore, Ericsson and Sweden's KTH Royal Institute of Technology have formed a research partnership to develop AI-powered positioning technology for future 6G networks, according to an announcement carried by Mirage News.

The collaboration links one of Asia's leading engineering universities with one of Europe's foremost telecom research institutions and the largest European network vendor by revenue. Its stated focus is positioning: using artificial intelligence to locate devices and objects through the radio network itself, rather than relying solely on satellite or sensor systems.

What is network-based positioning?

Mobile networks have carried location capability for years. Regulators in the United States and Europe require operators to deliver handset location to emergency services, and 4G and 5G standards include positioning features. But accuracy has remained coarse compared with GNSS satellite systems.

6G research programs, including the work driving early standardization discussions in 3GPP and ITU circles, aim to change that. The premise is that denser deployments, wider bandwidths in the upper mid-band and sub-terahertz ranges, and AI-native air interfaces will let the network itself sense its environment with centimetre-level precision.

That is the technical bet behind this partnership. AI models would process radio signals reflected from objects and devices to estimate position, in effect turning the cellular network into a distributed radar and sensing system. Positioning becomes a network function rather than an add-on service.

Who brings what?

The division of labor follows a familiar pattern for such alliances, though the parties have not published detailed work packages in the material available.

• NTU Singapore contributes its wireless and AI research base and a testbed environment in one of Asia's most advanced connectivity markets.
• KTH Royal Institute of Technology in Stockholm adds long-standing strength in signal processing and information theory, fields central to positioning algorithms.
• Ericsson supplies vendor-side system knowledge, equipment expertise and a route from laboratory research toward eventual productization.

The pairing of a Singaporean and a Swedish institution around a single European vendor mirrors earlier academic collaborations Ericsson has maintained across its home market and Asia. For operators, the relevance is direct: if positioning matures into a standardized 6G capability, it opens potential revenue in industrial automation, logistics, autonomous transport and asset tracking, segments where operators currently sell mostly connectivity.

Marketing claim versus commercial reality

Practitioners should hold this announcement in the research category, not the product category. No deployment timelines, spectrum commitments, operator customers or throughput figures appear in the announcement.

The gap matters. AI-driven positioning is an active field across the industry, with academic groups and vendors including Nokia and Huawei publishing comparable research. Turning such work into standards requires 3GPP study items, consensus among operators and vendors, and chipset support. That process has not started for the specific techniques this partnership will explore, because 6G standardization itself remains at a preparatory stage, with first 3GPP study work expected later this decade and commercial networks projected around 2030.

What the partnership does signal is vendor intent. Ericsson is positioning itself to have AI-based sensing and positioning intellectual property ready before those standards discussions crystallize. Universities gain funding, industrial relevance and access to live network research infrastructure.

Why positioning matters for 6G economics

For operators, sensing is one of the few 6G use cases with a plausible path to new revenue rather than faster existing services. Location data underpins use cases that connectivity alone cannot price:

• Autonomous vehicles and drones needing redundancy when satellite signals fail indoors or in dense urban canyons.
• Smart factories tracking tools, components and people in real time.
• Public safety and emergency response, where regulators may eventually mandate network-derived location accuracy beyond current requirements.

Each of these markets depends on positioning accuracy that today's networks cannot deliver. Research alliances like this one are how the vendor ecosystem tests whether AI can close that gap before anyone commits standardization effort to it.

What comes next

The partners have announced the collaboration but have not disclosed funding amounts, project duration or milestone dates in the available material. Watch for follow-on signals: joint publications at conferences such as IEEE Globecom, patent filings in machine-learning-based localization, and any eventual Ericsson 6G testbed demonstrations in Singapore or Stockholm that put measured accuracy numbers against these ambitions.