Is LoRa vying for the "right to speak" in the new development cycle of the Internet of Things?

Over the past decade or so, the Internet of Things (IoT) industry has gone through an infrastructure expansion period from "whether it can connect" to "how many connections there are". Connection scale, coverage, power consumption level, and cost structure constitute the core variables in the first - stage competition. However, when hundreds of millions of devices around the world are already online, the industry is facing a more fundamental question: where is the value after connection?

Therefore, AI has become an inevitable proposition for the IoT. In fact, the industry has been thinking about the topic of "how to combine AI and IoT" for more than three or five years. The concept of AIoT (Artificial Intelligence of Things) is not new. From cloud - based data analysis, to edge computing, and then to device - side inference, the path has been evolving, but there has always been a lack of a landmark node.

On March 2nd, at the 2026 Mobile World Congress (MWC2026) in Barcelona, the LoRa Alliance announced its vision: how LoRaWAN and artificial intelligence can be integrated to achieve transformative functions and greater value at the edge layer, core layer, and application layer of the IoT stack.

As we know, LoRaWAN is one of the most popular LPWAN (Low - Power Wide - Area Network) connection technologies. According to an officially released report, by the end of 2025, there were already more than 125 million LoRaWAN IoT devices globally, and the compound annual growth rate (CAGR) of its ecosystem was 25%. Now, LoRaWAN is taking on a more important role, providing a connection foundation for the IoT to become the "digital nervous system" of AI. LoRaWAN supports the collection of raw IoT data for AI analysis and transmits the insights and actions generated by the analysis to the most valuable scenarios through LoRaWAN, thereby improving operational efficiency, creating new revenue opportunities, and enabling users to realize more value through their connected devices and IoT applications.

The author noticed that Alper Yegin, the CEO of the LoRa Alliance, said an interesting thing in the official press release - "The collaboration between LoRaWAN and artificial intelligence paves the way for artificial intelligence to move from the pure digital world to the physical world." In other words, he no longer positions LoRaWAN merely as a low - power wide - area network connection standard but tries to shape it as an interface and nervous system for AI to enter the physical world.

What this reflects may be the re - competition of communication technology standards for the "underlying entrance" in the era of physical AI.

Three Ways of Collaboration between AI and LoRaWAN

Let's first see how the LoRa Alliance specifically elaborates on this vision - it states that AI and LoRaWAN technology collaborate mainly in three ways: AI at the edge, AI in the core, and AI in the application.

①AI at the Edge

More and more sensors and devices connected to LoRaWAN enable AI processing capabilities to run directly at the very front - end of the IoT network and inside the devices, not limited to the wireless access network layer. Device - side AI processing reduces the need to transmit large amounts of data to the cloud and lowers the latency between perception, insight, and action. These devices can take advantage of LoRaWAN's low - power, high - scalability, and low - cost connection to transmit only necessary results, such as notifications, alerts, and suggestions.

Currently, LoRaWAN - connected cameras have been deployed in various environments for event detection and people counting. By using device - side AI to process image data, alerts can be generated more quickly. In other scenarios, LoRaWAN - connected vibration and load sensors are used for device monitoring in large - scale industrial environments. In these scenarios, AI analyzes the device status and generates predictive maintenance suggestions when the wear reaches a specific threshold.

The official has provided some examples to prove that many member companies of the LoRa Alliance have promoted the implementation of the above applications. For example, Seeed Studio and Milesight have both launched camera products with device - side AI processing capabilities; Honeywell, Advantech, Watteco, and TE Connectivity provide vibration sensors that integrate LoRaWAN connection and AI processing capabilities.

②AI in the Core

AI processing not only brings value at the edge of the LoRaWAN network but can also be used by network operators in the LoRaWAN core network to analyze network patterns and detect anomalies, thus more proactively managing network performance, reliability, and security.

For example, the CanopyNOC product of Kudzu Technologies, a member of the LoRa Alliance, uses intelligent agent AI to autonomously monitor and identify network anomalies, providing operators with actionable intelligent analysis to assist in solving core network problems.

③AI in the Application

LoRaWAN technology supports various IoT applications with long - distance and wide - range coverage, such as low - power asset tracking, smart cities, smart agriculture, and large - scale environmental and industrial monitoring. Integrating AI into these application scenarios can improve operational efficiency and provide more accurate asset location and status information.

Currently, several members of the LoRa Alliance have carried out practices in this field:

Browan and Combain provide AI - enhanced indoor positioning products;

Akenza's IoT platform is equipped with an AI chatbot that can answer questions based on real - time IoT data;

Creative5 has deployed a hybrid LoRaWAN + Non - Terrestrial Network (NTN) satellite connection solution in Taiwan, China, for real - time environmental monitoring (such as temperature, humidity, water level, etc.) in remote mountainous forests where the terrestrial network cannot cover. The data is transmitted through its Hestia LoRaWAN gateway integrated with NTN satellite connection. AI analysis on the cloud platform supports anomaly detection, early warning of wildfires and floods, and predictive environmental insights;

Emergent Connext's Rip platform combines LoRaWAN connection with an AI intelligence layer to provide automation capabilities for agricultural producers;

inBiot's ANNE AI assistant is directly connected to its LoRaWAN sensor network to standardize and interpret real - time indoor air quality data;

MachineQ, a company under Comcast and a member of the LoRa Alliance, has developed its own AI application, demonstrating the integration trend of IoT and AI. This function uses AI to transform millions of IoT data points into clear and actionable insights, shortening the analysis that originally took days to seconds. By repeatedly integrating massive amounts of data, this application identifies patterns and trends in key areas, including asset location, utilization, alerts, status updates, and sensor readings of monitoring devices, generating concise and understandable summaries to help teams make quick decisions and optimize work processes.

Competing for the Right to Speak in the New Round of Dividend Cycle?

What really deserves attention in this news is not the evolution of a certain LPWAN technology route, but the shift of the IoT value axis - IoT may be entering the second wave of the dividend cycle.

The First Wave of IoT Dividends (2014–2020): Looking back at the period from 2014 to 2020, it can be regarded as the first - stage dividend period of the IoT. The core task of that stage was very clear: to solve the connection problem, establish technical standards, and cultivate the industrial ecosystem. From NB - IoT, LTE - M to LoRaWAN, from operator networks to private networks, from the decline of module costs to the increase in terminal scale, the industry's focus was on "being able to connect", "connecting stably", and "connecting many". Connection itself was the value, and the number of online devices became an important indicator to measure the maturity of the industry. The establishment of standards and the expansion of the ecosystem laid the foundation for the connection of hundreds of millions of devices globally today.

The Second Wave of IoT Dividends (2025–2035): However, when connections are no longer scarce, after several years of a sluggish period, it may also indicate that the second wave of the dividend cycle is taking shape. From 2025 to 2035, the growth logic of the IoT is likely to no longer come from the linear increase in the number of devices but from AI - driven decision automation, large - scale deployment of edge intelligence, and in - depth construction of the digital twin system in the physical world. In other words, the value of IoT will shift from "data collection" to "intelligent decision - making" and from "visualization" to "actionability".

In this context, it is not accidental that the LPWAN camp represented by the LoRa Alliance begins to emphasize concepts such as "physical AI" and "closed - loop action" that are strongly promoted by AI companies. In the existing LPWAN competition pattern, whether it is NB - IoT, LTE - M, or satellite IoT, the technical narrative has long revolved around coverage ability, power consumption performance, and cost advantages. LoRaWAN also had clear labels in the past: low power, low cost, flexible private network, and strong deployment flexibility. But now, it is trying to re - define its role - not only as a connection protocol but also as the data entrance of AI, the action exit of AI, and even the communication nervous system of physical AI. This change in positioning is essentially an upgrade of the right to speak.

As AI is gradually embedded in industrial equipment, agricultural systems, energy facilities, and urban infrastructure, whoever controls the connection layer of physical devices is closer to controlling the starting point of intelligent decision - making. The competition of communication technology standards is shifting from the comparison of bandwidth and coverage to the competition of who can become the best physical interface in the AI era.

The future standard advantage may no longer be determined only by network performance parameters but by the degree of coupling with AI capabilities - whether it supports edge inference, whether it has low - latency closed - loop capabilities, and whether it can efficiently carry the transmission of intelligent results. The focus of LPWAN competition may also shift from simple "network capabilities" to "intelligent capability integration".

This trend poses new propositions for all links in the industrial chain:

For device manufacturers: Simply providing sensing and reporting functions will gradually become marginalized. Whether they have local inference capabilities and whether they can make preliminary decisions at the terminal side will become the key to product differentiation.

For network operators: The real value is not to provide a connection pipeline but whether they can achieve anomaly detection, autonomous optimization, and service - level guarantee through network - side intelligence, upgrading from "channel providers" to "intelligent network service providers".

For platform manufacturers: Data aggregation ability is no longer sufficient to build a barrier. The real core lies in whether they can provide a decision - making engine to transform massive amounts of data into actionable business actions, rather than staying at the level of data warehouses and reports.

In a sense, whoever can become the underlying channel for AI to enter the physical world has the opportunity to occupy a more central position in the next round of the industrial cycle.

Conclusion

In the past few years, IoT practitioners have inevitably felt the development bottleneck. Looking back from this time point from a more optimistic perspective, the IoT is not at the end of its maturity stage but may be at a new inflection point for capacity upgrading. When AI begins to be deeply embedded in the physical world, the meaning of connection technology is re - defined - it is no longer just a data channel but the neural network of an intelligent system. Whoever can carry intelligence on the connection layer can occupy a higher - value position in the new round of industrial division of labor.

Whether it is LoRaWAN or other IoT connection technologies, the real challenge does not lie in parameter indicators but in the ability to build a closed - loop ability of "perception - analysis - decision - execution". In the second half of the IoT era, it's not about scale expansion but about intelligent density. For the entire industry, this is not only a technological integration but also a re - evaluation of value and an early layout for the right to speak in the next decade.

References: LoRaWAN and Physical AI Unite to Boost Global IoT —— IoT Business NewsLoRa Alliance Outlines How LoRaWAN and Physical AI Are Teaming up to Maximize the Value of Both Technologies in the Global IoT Market —— Business WoreFour things to expect at MWC Barcelona 2026, according to Spectrum Effect —— RCR WirelessLoRa Alliance Reports 125 Million LoRaWAN End Devices Deployed Globally —— ARC

References:

LoRaWAN and Physical AI Unite to Boost Global IoT —— IoT Business News

LoRa Alliance Outlines How LoRaWAN and Physical AI Are Teaming up to Maximize the

Value of Both Technologies in the Global IoT Market —— Business Wore

Four things to expect at MWC Barcelona 2026, according to Spectrum Effect —— RCR Wireless

LoRa Alliance Reports 125 Million LoRaWAN End Devices Deployed Globally —— ARC

This article is from the WeChat official account “IoT Think Tank” (ID: iot101), author: Sophia, published by 36Kr with authorization.