IoT and AI are becoming more deeply entwined, said Eric Zhao, vice president of Huawei, during his keynote speech at last week’s MWC in Shanghai.

Zhao discussed how three elements were key to achieving full AIoT.

“They are all-scenario IoT that expands IoT connections to all scenarios, ultra-broadband networks that link all data to the cloud and computing resources, and intelligent applications that are driven by industry-specific models,” said Zhao.

AI and cellular IoT are converging, which means more things that are helpful in everyday lives will be connected, such as intelligent vehicles and embodied AI robots. This also means, an increasing number of intelligent applications will leverage these connections across diverse scenarios, such as production lines and smart ports, to transform industries with intelligent technology.

Moreover, the previously unconnected spaces will become connected, and this will enable a plethora of innovative applications, including drone-based power grid inspection, to boost smart urban governance.

First, all-scenario IoT expands connections to all scenarios, enabling the collection of all production data. Secondly, 5G-A ultra-broadband networks transport these data from the physical world to models in the cloud in real time. Thirdly, industry-specific and scenario-specific models transform core production processes across industries by making applications intelligent.

All-scenario IoT provides a growing range of devices with diverse IoT connections. 5G RedCap and ambient IoT are joining existing technologies such as NB-IoT to make all-scenario IoT possible. This will enable a great number of innovative applications, including embodied AI robots requiring real-time connections, AI-based product quality inspection requiring super-fast connections, and other applications that operate with low latency and power consumption.

Ultra-broadband networks are possible with 5G-A technology, which offers a wide array of new capabilities, including Gbit/s uplink, ultra-low latency and extensive coverage. By adding these powerful functions to networks, 5G-A allows a single network to provide IoT connections for diverse services, such as those that need hyperscale data collection and on-the-fly data movement to cloud computing and industrial application platforms for AI training and inference.

Intelligent applications are playing an increasingly prominent role, as demonstrated by the soaring numbers of deployments of industry-specific and scenario-specific models. A massive amount of quality data is needed to make these models more effective. In one intelligent manufacturing factory in Guangdong, China, AI algorithms have been integrated into 5G HD cameras to enable intelligent product quality inspection. This has not only improved overall product quality, but reduced equipment repair rates by 20% and saved annual costs by more than CN¥1m.

Zhao concluded his speech by calling for industry-wide efforts to promote the development of cellular IoT.

“We will continue to work with industry partners to develop more converged applications of cellular IoT and AI,” he said. “We will develop a thriving ecosystem to usher in a new age of full intelligent IoT connectivity.”

The commercial adoption of 5G-A is accelerating in 2025. Huawei (www.huawei.com) is collaborating with global carriers, industry experts and opinion leaders to explore how innovations in AI can be used to reshape telecom services, infrastructure and operations to generate new revenue sources and accelerate the transition towards an intelligent world.