• Member
• October 13, 2026

Communication infrastructure as sensing and perception system

SRIJAN SANCHAR · TELECOM INSIGHT SERIES

When Networks Learn to See

Spatial intelligence and who will own it

The signal

In February 2026, a California company called ZaiNar emerged from nine years in stealth with more than $100 million in funding at a valuation above $1 billion. Its claim was striking: sub-nanosecond time synchronisation across existing 5G, Wi-Fi and private cellular networks, precise enough to locate devices to within a metre, indoors and out, using only the radio signals those networks already carry.

The claims are the company's own and await independent verification. But the direction they point to is now visible across the industry. Eight months later, at India Mobile Congress 2026, Tejas Networks and Qualcomm demonstrated live 5G infrastructure detecting and tracking drones and vehicles. In China, operators are already selling network sensing commercially.

The thesis

For four decades, mobile networks have done one job: move information between people and machines. A second job is now being added. The same towers, spectrum and signals are beginning to sense the physical world: where things are, how they move, and what is present in a space.

We call this network-native spatial intelligence. It could turn telecom networks into shared spatial utilities, much as electricity grids and roads serve many sectors at once. It also raises two questions this article takes seriously: who will capture the value, and who will govern a network that can see?

Three capabilities, often confused

"Networks that see" covers three different things. Keeping them apart makes the opportunity, and the risks, much clearer.

Figure 1. From networks that connect to networks that see

Capability	What it does	What it needs	Example
Positioning	Locates devices that transmit radio signals	Precise timing and measurement of signals from phones, tags or sensors	Tracking a robot or phone indoors to within a metre
Sensing	Detects objects that transmit nothing, from the radio waves they reflect	Base stations that work like radar while communicating	Detecting an unregistered drone or a vehicle
Perception	Interprets what has been located and sensed	AI that turns positions and echoes into meaning	Recognising a crowd forming or a drone off its route

 

Positioning is the most mature: networks have located phones for emergency calls for decades, and 5G added precise positioning features from 3GPP Release 16. Sensing, known in the industry as integrated sensing and communication (ISAC), is moving from research into standards and early commercial use now. Perception is where AI meets the network, and where most of the long-term value and most of the governance questions sit.

Technology update: where the field stands in late 2026

Standards are moving from study to specification

Figure 2. Sensing in 3GPP standards

Release	What 3GPP did on sensing
Release 16	Precise positioning features for 5G devices
Release 19	First sensing study: deployment scenarios and channel models for sensing, including how targets reflect signals and move
Release 20 (5G-Advanced)	Normative work on base-station sensing, with drones as the first target use case; study report TR 38.765 completed in June 2026; core-network architecture study (TR 23.700-14) complete
Release 20 (6G studies)	23 sensing use cases accepted for 6G; architecture, security, charging and management groups all working on sensing
Release 21 and later	First normative 6G specifications, with sensing expected as a native feature

 

Two details matter. First, the 6G architecture work explicitly plans to fuse network sensing with data from Wi-Fi sensing, cameras, sonar and lidar, so the network becomes a hub for many senses, not just radio. Second, the standards define how sensing results are produced and exposed, but not how applications use them. As with movement information in earlier generations, the use is left open.

Advances in practice

•      Timing as the key to precision. Approaches like ZaiNar's derive position from extremely precise time synchronisation of signals phones already send, such as the 5G sounding reference signal, rather than dedicated positioning signals. If independently validated, this makes precise positioning a software upgrade to existing networks.

•      Commercial sensing in China. In November 2024, ZTE, China Unicom Henan and a drone operator launched what they described as the first commercial deployment of 5G-Advanced sensing base stations for drone blood deliveries in Anyang, tracking over 20 drones with no detection failures. In February 2026, China's government ordered its telecom industry to upgrade networks to monitor airspace below 300 metres, starting in ten pilot cities.

•      Sensing on live networks in India. At IMC 2026, Tejas Networks and Qualcomm demonstrated sensing on live 5G infrastructure, using a Tejas base station with Qualcomm chipsets and AI inference to track drones and vehicles. TSDSI, India's standards body, is finalising a technical study on sensing use cases and architectures.

•      The cost reality. Reports put the cost of each 5G-Advanced base station for low-altitude coverage in China above $143,000, with dense deployment required. Sensing is real, but not yet cheap.

A sobering assessment

Industry analysts at ABI Research observe that the main obstacle is no longer sensing performance, which was largely demonstrated in Release 19, but business outcomes. They argue that early visions of a universal sensing platform are narrowing towards specific, premium uses built around operators' existing infrastructure. That is a useful corrective to the more expansive claims, including those in the first version of this note.

Why it matters now

•      Machines need to know where they are. Robots, drones, autonomous vehicles and warehouse systems need reliable positioning where satellite navigation fails: indoors, underground, in dense cities and under jamming.

•      Operators need new value from existing assets. Turning networks into sensing platforms adds capability without proportionally adding sites, which matters in an industry where connectivity alone earns less each year.

•      Societies need resilience. Satellite navigation is vulnerable to outages, jamming and spoofing. A ground-based layer of positioning and timing adds a second line of defence for critical infrastructure.

What it could mean for India

Area	Application	Why the network helps
Railways, metros and ports	Positioning of assets, wagons and containers; backup timing	Works in tunnels, yards and covered areas
Drones and low-altitude airspace	Detecting and tracking drones near airports, borders and sensitive sites	Base stations already cover the country densely
Crowd safety	Sensing density and flow at melas, stations and stadiums	Warning before crowds become dangerous
Mines, warehouses and factories	Indoor positioning for workers, vehicles and robots	Uses private 5G or Wi-Fi already being deployed
Disaster response	Locating people and assets after floods and earthquakes	Works when other systems are damaged
Navigation resilience	A terrestrial complement to NavIC and GPS	Reduces dependence on any single satellite system

 

Some of these touch safety and security. They would need strict separation from safety-critical systems, such as train control, and clear legal authority, especially where sensing is used for policing or defence.

Who will own it?

The first parts of this series argued that value is created where a frontier is still open, scaled by shared capability, and kept only by those who hold control. Spatial intelligence passes the first two tests easily: the frontier is open, and one network can serve railways, ports, drones, factories, cities and emergency services at once. The open question is control.

Figure 3. Three possible owners of spatial intelligence

1.    Operators own it. The network itself is the natural control point: operators hold the spectrum, the sites and the licences. Spatial intelligence could be one of the few ways for operators to climb out of the "dumb pipe" valley, selling sensing and positioning as services. Standards now include charging and exposure for sensing, which supports this path.

2.    Specialists own it. Positioning companies and equipment vendors supply the intelligence as software running on operators' networks. Operators earn a fee; the specialists hold the value.

3.    Platforms own it. Cloud and AI companies fuse network sensing with cameras, maps and other data. The network becomes one input among many, and value moves to whoever builds the perception layer.

Which of these happens will depend less on technology than on who builds trusted, reusable services first, and whether operators act together. A drone-detection service that works on only one operator's network is far less valuable than one that works everywhere.

A network that can see must be governed

The same capability that tracks a delivery drone can track a person. A network that renders the physical world continuously knowable is, seen from another angle, a surveillance infrastructure. If spatial intelligence is to earn public trust, governance has to be designed in from the start.

Figure 4. Principles for governing networks that see

•      Predict the crowd, not the person. Default to aggregate sensing: density, flow and presence, not identity. Individual tracking should require explicit consent or clear legal authority.

•      Purpose limitation. Sensing authorised for drone safety should not quietly become sensing for advertising or profiling.

•      Consent and transparency. Under India's Digital Personal Data Protection Act, personal data needs a lawful basis. People should know when and where sensing is active, much as signs announce CCTV today.

•      Clear rules for state access. Use for policing, security or defence needs explicit authority, oversight and records.

•      Security by design. A sensing network can be spoofed, jammed or used to mislead. Results that feed safety decisions need authentication and independent checks.

•      Separation from safety-critical systems. Sensing can support railways, airports and grids, but should never replace their certified safety systems.

Standards bodies are already working on security, charging and management for sensing. Privacy and public accountability, however, are questions for regulators and society, not engineers alone. India has an opportunity to set those rules early, as it did with its digital public infrastructure.

What could slow it down

•      Accuracy in the real world. Dense cities, crowded indoor spaces and interference can degrade both positioning and sensing.

•      Cost. Sensing-capable base stations and dense deployment remain expensive.

•      Business models. Who pays, and for what, is still unsettled.

•      Spectrum and interference. Sensing and communication compete for the same resources.

•      Trust. A single misuse scandal could set public acceptance back years.

•      Verification. Several headline claims, including precise positioning from existing signals, still come mainly from the companies making them.

Signals to watch, 2026–2029

By	Signal	What it would show
End 2027	Release 20 sensing specifications completed; Release 21 6G work including sensing natively	Sensing is becoming a standard network feature
End 2027	Independent validation of timing-based precise positioning on commercial networks	The software-upgrade path is real
End 2028	At least one Indian operator offering a commercial positioning or sensing service	The opportunity is reaching India
End 2028	India issuing guidance on privacy and permitted uses of network sensing	Governance is keeping pace
End 2029	Sensing revenue reported separately by operators anywhere	Operators, not only specialists, are capturing value

 

Closing

The first version of this note, written in March 2026, saw the signal early: communication networks are acquiring the ability to perceive the physical world. Eight months on, the picture is clearer and more grounded. Standards are being written, commercial sensing exists, and India has shown sensing on live networks at IMC 2026. The technology is arriving. The open questions are who will own it and who will govern it.

Sources

•      ZaiNar funding and technology, Pulse 2.0 (February 2026)

•      ZaiNar at MWC 2026, Fierce Network

•      Tejas Networks and Qualcomm ISAC demonstration at IMC 2026 (PR Newswire via UNI)

•      ZTE and China Unicom commercial ISAC deployment for blood delivery (November 2024)

•      China orders telecom upgrades to monitor low-altitude airspace, Caixin (February 2026)

•      3GPP TR 38.765, Study on ISAC for NR, Release 20 (June 2026)

•      ISAC-enabled non-terrestrial networks for 6G, arXiv 2604.11593

•      5G-Advanced and 6G in Release 20, arXiv 2506.11828

•      ISAC moves into formal 6G standardization, ABI Research

•      TSDSI chairman interview, tele.net

Srijan Sanchar Foresight  updated 

,

The technological transformation emerging around Zainar Company signals the beginning of a profound infrastructural transition in the relationship between communication networks and physical reality. For decades, communication infrastructure has primarily enabled connectivity—the transfer of voice, data, and digital information across distances. A new evolutionary phase is now unfolding in which the same infrastructure begins to function as a sensing and perception system. Networks are no longer limited to carrying information; they increasingly acquire the ability to understand spatial disposition, movement, and proximity within the physical world.

This transition represents the early formation of what may be described as network-native spatial intelligence. Instead of devices independently determining their location through satellite systems, positioning intelligence shifts into terrestrial communication networks themselves. Wireless infrastructure—5G systems, enterprise networks, and distributed radio environments—gradually transforms into a continuous spatial awareness layer capable of interpreting real-world dynamics in real time. Space becomes computationally observable, shared across systems, and available for machine reasoning.

The emergence of this technological direction is structurally driven rather than speculative. Autonomous systems across logistics, transportation, manufacturing, and defense increasingly require reliable positioning in environments where satellite navigation is inconsistent or unavailable. Dense urban environments, indoor industrial operations, underground systems, and contested electromagnetic conditions expose limitations in existing navigation models. At the same time, telecom operators seek new value architectures capable of leveraging already deployed infrastructure. Converting communication networks into sensing platforms introduces a new economic logic—expanding capability without proportional expansion of physical assets.

Over the next three years, this convergence is expected to catalyze the rise of physical AI ecosystems. Artificial intelligence systems, which currently operate largely within digital environments, begin interacting safely with the physical world through shared spatial awareness generated by communication networks. Robots, autonomous vehicles, inspection systems, and industrial machines gain the ability to coordinate movement collectively rather than operate as isolated intelligent units. Intelligence progressively shifts from device-level autonomy toward network-coordinated cognition.

A parallel evolution is likely in navigation resilience. Network-based positioning systems provide continuity where satellite signals fail, enabling critical infrastructure—railways, ports, airports, mining operations, emergency response systems, and defense applications—to maintain operational awareness under disruption. Navigation thereby transitions from dependence on orbital infrastructure toward ground-based intelligent networks, strengthening systemic resilience in an increasingly uncertain geopolitical and operational environment.

As sensing capability becomes embedded within communication systems, telecom networks themselves evolve into spatial utilities. Beyond connectivity services, networks begin supporting real-time asset visibility, mobility analytics, safety monitoring, infrastructure diagnostics, and coordinated automation. Communication infrastructure gradually assumes characteristics similar to electricity or transportation grids—shared foundational systems supporting multiple economic sectors simultaneously.

This evolution also enables continuous synchronization between physical environments and their digital representations. Cities, industrial facilities, and logistics corridors develop dynamically updated digital twins informed directly through wireless sensing rather than periodic data capture. Planning and management processes shift from episodic decision-making toward continuous optimization, allowing infrastructure systems to respond adaptively to changing conditions.

The technological trajectory points toward increasing decentralization of intelligence through edge-native spatial computing. Positioning and perception computation occurring within distributed network nodes reduces latency and enables cooperative machine behavior at scale. Emerging applications include swarm robotics, coordinated mobility ecosystems, autonomous fleet management, and adaptive urban systems capable of responding collectively to real-time conditions.

Across plausible futures, certain evolutionary characteristics remain stable. Precision timing, interoperability with existing communication standards, and passive compatibility with devices emerge as foundational traits. Deployment models, governance mechanisms, and commercial structures remain fluid as ecosystems mature. Legacy isolated sensing systems gradually lose relevance as shared spatial intelligence embedded within infrastructure becomes economically and operationally superior.

From a foresight perspective, the period between 2026 and 2029 represents an early but decisive transition phase. Initial adoption is expected within enterprise and industrial environments, followed by integration into national infrastructure and large-scale urban systems. By the end of this phase, communication networks may increasingly be understood not merely as channels of information exchange but as perceptual systems enabling real-time understanding of physical environments.

The deeper implication extends beyond technology adoption. The conversion of communication infrastructure into spatial sensing capability introduces a new civilizational layer positioned between connectivity and intelligence. If the Internet connected information and cloud computing enabled large-scale computation, spatially aware networks may enable societies themselves to perceive, coordinate, and evolve with greater adaptive intelligence.

In this emerging landscape, infrastructure begins to sense before it reacts, anticipate before disruption occurs, and coordinate across distributed actors without centralized control. The foresight emerging through Srijan Sanchar therefore indicates that the next stage of technological evolution will not merely digitize systems—it will render the physical world continuously knowable to intelligent networks, fundamentally reshaping how economies, cities, and institutions operate under conditions of growing complexity.