AI head

CCW 2019: Motorola Solutions on 5G, AI and edge computing

This year’s Critical Communications World (CCW) in Kuala Lumpur, showed a glimpse into the future of public protection and disaster relief (PPDR), thanks in part to briefings by Motorola Solutions’ Senior VP of Technology, Paul Steinberg.

Paul, who serves on the technical advisory board for multiple companies that supply solutions to the telco industry, is also a member of the Federal Communications Commission Technical Advisory Council.

During CCW, he shared his observations that 5G has increased capacity (compared to 4G) because of new radio capability that can take advantage of high swaths of spectrum; 5G can utilise spectrum as high as the 5GHz frequency, although this comes with a trade-off – there is a lot of bandwidth capacity for high throughput applications, but the distance of transmission becomes shorter.

Paul Steinberg

Paul said, “The way I think this will deploy (in the critical communications space) is 5G deployed in highly dense areas, and 4G deployed for wider, open area networks. So, it’s probably a hybrid network.”

He also noted that when 4G was introduced it was a huge leap ahead of previous generations of connectivity technology. “With 5G, it had to be about using new spectrum, or a new network topology. Turns out, that’s what 5G is. It is using existing spectrum more densely and more frequently.”

Motorola and 5G

5G was a major highlight during the 2019 global critical communications conference, because of all the other technologies it will be able to enable like virtual reality, autonomous drones, Internet of Things and so on.

But these technologies, are a few years away at least.

Paul shared that all of Motorola’s communications platforms, whether deployable or fixed infrastructure and devices, will take advantage of 5G. “But that’s just riding the ecosystem, and we are not innovating there.”

Where it begins to get interesting is when many different types of media and data can be fused into the workflow operations of a command centre.

We are talking about data that comes from various sources like law enforcement, public safety-generated data, private enterprise data, smart city data, data from a video camera that is wireless in an enterprise, or data from a smart building that has fire sensors, location sensors and so on.

All these disparate data can be accessed in a timely manner, and presented in a way that helps command centre operators kickstart the right action via first responders at the scene of the event.

What’s happening here is Motorola potentially being able to take advantage of the scale of remote IoT and 5G connectivity.

Also, thanks to low latency properties of 5G technology, there is potential to enrich the user experience and interaction of first responders with virtual reality and augmented reality.

5G fast forward – autonomous drones

So, 5G’s reduced latency opens many doors for mission-critical services.

The use of drones as a ‘sensor’ that feeds data back to the command centre, for example.

Paul explained, “Feeding a drone’s video into the operational fabric of an incident management, or even monitoring in and around an event like a celebration… that’s step number one.

“Step number two is making the drone autonomous, and have it do some things for you like fly a search pattern along a border, or when an emergency happens the call-taking centre can send it as the first responder, on the scene.”

According to Paul, all of this is essentially possible technically, and Motorola is already working with companies that can interface with components on the drone.

What has to happen however, is flight regulations have to catch up. “Right now, this notion of beyond-line-of-sight autonomy for drones is not allowed in most places in the world,” Paul said. He also added that in the US at least, regulations are trying reduce restrictions so as to allow drones be used in public safety scenarios.

5G fast forward – leveraging AI on the edge

5G is taking advantage of the smarter edge, and vice versa.

‘We made investment in a company called Syntiant, and what they are doing amounts to an analogue computer.

What’s involved is a low-cost, ultra-low powered AI chip that can achieve about 100x efficiency improvement over stored program architectures like CPUs found in desktops. The outcome of this is an always-on voice interface to any battery-powered device, even one as small as a hearing aid.

Paul explained that this chip is capable enough to run just enough AI to process lexicons of speech; not natural language processing but a full dialogue; as well as recognise and process images.

“So, think about a very simple example like our push-to-talk (PTT) radios. In theory, we can use that chip for a couple of dollars and very low energy, to listen, speak or identify, and simply eliminate the need to ever press a button,” he said.

This aligns completely with the PPDR community of first responders needing to always be ‘eyes up’ and ready.

“This is something we are tinkering with. But, when you see things like this in the world, you start to see the notion that everything can be smart,” Paul added.

With more smart edge devices like the voice-activated radio, some of the processing may be offloaded from the network/cloud side. Without needing to send data back and forth, from the device to data centre, response times can greatly increase.

And this is a very important requirement in mission critical events.