AI, IoT and 5G: a winning combination

The Internet of Things (IoT) consists of smart objects equipped with sensors that collect data, which is then shared via a network. By 2025, it is estimated that 22 billion IoT devices will be connected to the internet. Artificial intelligence (AI) has also made great strides in recent years, with applications in a variety of sectors. And the advent of 5G has heralded the dawn of a new age: one in which we are connected faster, smarter and more efficiently. We talked to Stefaan Van Aken, 5G Consumer Solution Architect at Orange, about the combination of these three technologies.

AI, IoT and 5G are three emerging technologies. What makes their interplay so important?

“These are three technologies that not only feed off each other but also reinforce each other. The synergy between AI, IoT and 5G leads to intelligent connectivity. This is not some far-off pipe dream: we are already implementing this combination in the real world.”

Which applications benefit the most from the grouping of these technologies?

“A key area in which the combination of these technologies is proving most useful is the inspection of industrial buildings using drones. At the moment, these inspections for damage to silos, for instance, are still mostly done manually. A drone with a camera flies around the building and captures images. When the drone has completed its inspection round, an inspector retrieves the SD card and analyses the stored images on a computer. This is a time-consuming process. And if the images are inconclusive, the drone has to be deployed again.

“We are working on a project with a customer in which a drone is equipped with a SIM card for our 5G network. This network has sufficient bandwidth in the upload direction to send live camera images to the cloud. Our customer has an AI model there that analyses the incoming footage in real time. A well-trained AI system can spot damage not only faster but also more accurately than the human eye. An important advantage of 5G compared to 4G is the ability to reserve bandwidth for the application. This means the inspection will still function properly if, for example, a major event is taking place nearby.”

And which applications will benefit from AI, IoT and 5G in the long term?

“Autonomous vehicles will undoubtedly benefit greatly from these technologies. They are fitted with an array of sensors, cameras and lidars (LIght Detection And Ranging) to detect objects around the vehicle. At the moment the vehicle itself processes this sensor data to be able to make instant adjustments. For even better performance, however, we need AI models in the cloud that harness more computing power, as in our example with the drones. This requires the vehicle to send all that sensor data and camera images to the cloud for analysis.

“Autonomous vehicles can’t do without 5G, for a number of reasons. On the one hand, the camera images mean they need a lot of bandwidth. On the other hand, you also need low latency: the vehicle must be able to correct its course instantly when an object is detected. The decision made by the AI model in the cloud must therefore reach the vehicle in time. We are working on a pilot project with a customer in the port of Antwerp, where self-driving trucks ferry containers between terminals. This application will be possible in the near future, albeit under controlled circumstances. One day, most vehicles in the public space will drive autonomously and then 5G will be the ideal network to make this possible in a safe and stable way.”

What can Orange do to support companies that want to commit to AI, IoT and 5G?

“Orange has been selected by the Federal Public Service Economy to implement 11 innovative 5G pilot projects. We are taking their execution very seriously: we truly want to grasp the requirements to successfully develop such innovative applications. In these projects we are the point of contact for connectivity but we are also involved in the process as a whole, from design to implementation. We increasingly see ourselves as an integrator, a role in which we work with an extensive ecosystem of partners, for instance in the area of AI. Whereas until recently we primarily established network connections, we have evolved over the last few years into an end-to-end partner who works closely with diverse partners.”