Dan Rothrock, President of Zenitel Americas, explores why audio is vital for security and how AI will shape its future.
“As humans, communication is essential to the way we interact,” Rothrock explained.
“What we hear and the way we say those things are all shape how relate to the world.”
For example, the very first day an infant is placed in its mother’s arms, it will start to recognize the difference between the sounds that are around it – setting the stage for how important communication is throughout our entire lifetime.
He emphasized that, within the context of security, the proliferation of visual systems makes sense, allowing teams to have a remote set of eyes out in the field.
However, Rothrock pointed out that this information is mainly reactive – not interactive.
When an incident is happening, the video will pick it up and alert personnel in the control room who will then decide the next course of action.
On their radio, they can tell other personnel to check on the situation.
“Now, resources are tied up. However, if I had audio, then I can automate the whole process,” Rothrock said.
“If a video analytic says that there is a person trespassing which is linked to an audio system, then that person in the control room can speak to the potential bad actor and let them know they are being watched in real time and that law enforcement officials are on their way.”
Therefore, not only can an incident be mitigated, but it can possibly be prevented from happening in the first place when both video and audio are used together.
Once you have audio, it’s important to ensure that it is unobstructed and clear to the person listening.
“One of the great things about the IP world is that it has changed the things that we can do,” he said.
“Years ago, there was a government agency that wanted the development of a microphone which they could deploy on one side of a water fountain to hear the conversations.
“At the time, we told them that it is impossible, because the microphone picks up what the human ear picks up so there would always be interference from the sound of the water.”
However, Rothrock explained that highly sophisticated algorithms can remove repetitive noise from recordings.
He added that these algorithms are being improved all the time, so that a person in a control room can hear voices clearly, even near a water fountain, which Zenitel’s solutions were able to deliver.
Another example he gave was with different modes of transport – if a speaker is targeted towards someone that might be in a large truck, there will be a lot of engine noise.
So, there needs to be a high-powered amplifier and automatic gain control to make the station louder if it needs to be.
If the truck then moves away and an electric car then pulls up, the volume will adjust as needed.
“Proper communication is important,” said Rothrock.
“Making noise louder or softer as needed is key to customer satisfaction.”
Another way to improve audio quality that Rothrock highlighted is through echo cancellation.
When walking through a tunnel, audio can become distorted and create a high level of feedback.
Therefore, ensuring that technology has echo cancellation is critical.
“Additionally, when we take an analog signal and digitally pack it, we send packets down the network. If there’s a delay, it can make the audio unintelligible,” he said.
To combat this, he explained that an adaptive jitter buffer circuit can be used.
“At the end of the day, though, the customer is not focused on how the high audio quality is achieved,” Rothrock noted.
“All they want is clear results that sound like natural communication.”
Rothrock also spoke about the different components that make up an audio system, breaking it down into hardware, software and mechanics.
He explained that when deployed, these solutions need to be durable as often they are placed outside.
“Water ingress is a major concern. So, we need to carefully consider how we design that mechanically,” Rothrock said.
“For example, our Turbine station has a die-cast aluminum front.”
This means that the intercom can be placed in very harsh conditions and still function.
Even more, Zenitel’s MEMS microphone is a fully digital microphone that eliminates a lot of the impacts that other microphones experience due to temperature and humidity.
Looking ahead to the future of the intercom and the audio solutions industry, Rothrock highlighted that in security, we’re just getting started.
“We’re beginning to get a seat at the table where video surveillance and access control sit,” he said.
“Many people thought that voice communication was going away – particularly when some people prefer to text versus call – but think about Alexa and Siri. These days, we talk to machines.”
Rothrock noted that many of these technologies are not yet quite as successful as understanding instructions as a human adult does.
However, with advances in AI and machine learning happening every day, this aspect is likely to improve exponentially.
Another key evolution that Rothrock expects to see is vocal interfacing.
Rather than using a computer monitor and mouse to give commands to other systems such as access control, it might be possible to give instructions audibly.
Additionally, when AI is producing speech to communicate with humans, eventually it will sound more natural.
“When speaking to a machine in the future, it will sound more authentic,” Rothrock predicted.
“For example, more accents will be used.”
He noted that AI will move on from simply having the ability to give information and handle requests but will partake in conversations.
In the future, it will also be able to detect more from a person’s voice, including authenticating their identity, but also analyzing their emotional tone.
That will improve security and mitigate risks even more.
Overall, Rothrock said, with the advent of AI, voice communication will become the center of security; it will become the center control point.
“There is no better place to gather data than in a conversation,” Rothrock concluded.
“If we can use AI to capture data while we’re having conversations, it can be a huge multiplier to every other way that we collect data today.”
This article was originally published in the June edition of Security Journal Americas. To read your FREE digital edition, click here.