A recent report from Juniper Research anticipates a significant surge in cellular roaming for low-power Internet of Things (IoT) devices over the next five years. The report projects that roaming connections from narrowband IoT (NB-IoT) and LTE-M devices will experience a remarkable fivefold increase, rising from 90 million in 2023 to an estimated 490 million by 2028.
This substantial growth is expected to be driven by an increase in bilateral roaming agreements among mobile operators, specifically tailored to optimize the monetization of roaming traffic generated by resource-constrained IoT devices. These agreements are poised to facilitate seamless connectivity for assets traversing international borders, unlocking novel applications in areas such as supply chain tracking, connected cars, and smart city infrastructure.
However, Juniper emphasizes that contemporary carriers often face challenges in accurately detecting and accounting for IoT traffic on their networks. The sporadic connectivity patterns of low-power devices add complexity to collecting roaming revenue. To fully capitalize on the potential revenue stream presented by the growing volumes of roaming, operators will need to invest in AI and machine learning capabilities. These technologies will play a crucial role in identifying IoT device connections and usage patterns.
According to Alex Webb, a research author at Juniper, operators must leverage insights gained from AI-based detection tools to introduce premium billing for roaming connections, maximizing revenue. This involves implementing roaming agreements that price connectivity based on network resources used and time connected to the network.
This evolving landscape presents both opportunities and imperatives for global carriers to adapt their networks and business practices to accommodate the burgeoning category of cellular IoT roamers. Capturing value from the increasing number of mobile assets promises to emerge as a central narrative in the telecom industry’s growth trajectory in the coming years.
