Conf42 Machine Learning 2025 - Online

- premiere 5PM GMT

5G Infrastructure Automation: Building the Foundation for Data-Driven Smart Cities

Video size:

Abstract

Revolutionize urban landscapes with our 5G automation framework. We slashed network deployment time while enabling real-time applications that improved traffic flow, energy usage, and citizen satisfaction. Discover how DevOps created the backbone for connected cities.

Summary

Transcript

This transcript was autogenerated. To make changes, submit a PR.
Hi. Hello everyone. This is Anil Kumar. I'm here to talk about 5G Infrastructure Automation, and let me give you a little bit intro about myself. So I studied at Harrisburg University, worked at Verizon, and I'm currently working at United Airlines as a DevOps engineer. Now when you hear of 5G most people immediately think a faster Netflix or smoother Zoom calls and maybe one or two dropped calls less, right? So it's all about 5G. So I will let explain something more about 5G today. But what if we had told you we used 5G? Not to binge watch more tv, but to help cities think we automated the traffic. Include emergency response, even made energy usage, more sustainable. We gave cities a digital nervous system, and today I'm walking you through, not in theory, but real world deployed solution. You'll see how edge computing small cells, automation, and gold, good old DevOps made cities faster, smarter, cleaner, right? So let's dive in The first section of it it's a smart city problem. Let's go back to the straight point. The mass cities are now well complicated, but their systems are way behind. Like when you compare, traffic lights are like set on 2005 timers with the grid locks and all. All right, so like power grids always on whether you need it or not. Emergency services like often routing through traffic with no dynamic data, like when an ambulance is trying to go. So they just follow the GPS and go and follow through the routes. Which they gave one funny example, in one city the traffic sys light system couldn't be updated remotely. We had to send someone out. But with the USB stick there, and that's not outdated, that's prehistoric. It wasn't just inefficient, it was dangerous, wasteful and expensive. Need a manpower. We need a real time data driven cities that where 5G and DevOps, that's where DevOps came in and we come up with some strategy. Here. The DevOps revolution, right? So when I talk about the DevOps revolution, so where DevOps meets infrastructure, usually when you hear right DevOps, you think deploying microservices, managing CICD, et cetera. But we use the DevOps to deploy networks. Every router switch and edge was defined as a code, like an infrastructure as a code. We had terraform plants that spun up 5G clusters. We did it in AWS hand charts configured edge services, pipelines, auto tested latency, bandwidth, even failures. Instead of saying, let's send a guy with the LA ladder, we set, let's push to pro, we reduced deployment time by 85%. You know what I'm trying to say is that, so instead of sending a person with a USD stick with, along with a ladder. So what we just did was let's push it to the prod. Whatever the automation we did, we reduced the deployment time by 85% and created a rollback options that didn't exist in the real traditional networking. So here comes our next slide, which is like a building the low latency network. So it's about a latency breakdown. I'm talking here to make smart cities smart. They need to respond instantly. We need a sub 10 milliseconds latency. That meant small cells on the light poles building business buses, basically anywhere with power and wifi and fiber back haul, connecting everything, edge nodes, doing local processing. No more round trips to a cloud 2000 miles away. This enabled right live video analytics, real time alerts. And autonomous drone patterns. We are talking Miss milliseconds here. That's less than a blink. Imagine a camera, spots a fire and instantly rewrote the traffic with a dispatching alerts. No waiting, no bottlenecks instantly rero the ambulances to a different route so that you can save lives. So it's like interesting thing we built here. So next go, I'll go to the section four, which is a IOT mess like iot ecosystem. What I taught, what we presented, so it's like we have connected like network. We layered in devices, lots of them. So like 10,000 connected endpoints per square mile. Smart meters, pollution sensors, cameras, emergency buttons, EV chargers, so all talking, adapting, learning autonomously. We didn't just install gadgets, but we created a nervous system for the city. One time our sensors picked up an air quality spike and rerouted a school buses away from it. So that isn't a mesmerizing. So no humans involved in there. Like just the systems have been taken care and they, the system just knew it. So it has picked an air quality in the air and spike and rerouted the traffic along the distances. So here comes our next slide. It's a smart traffic management which we call it as a smart traffic and power. So I will talk a little bit about the smart traffic here. With AI powered video analysis, our system adjusted traffic lights in a real time. So I told you right, so it, the traffic got rerouted in the previous lights, so it's like a lights adapted to the condition, not timers. Like what we talk about previous lives, green waves reduced stop and start driving. Emergency corridors were dynamically open for the ambulances and fire trucks. We reduced the commute times by 25% and cut idle time in half. That's gas saved time saved and air cleaned up, right? So now on to the power. We used a smart grids IOT meters that automatically balanced energy, a solar output high charged. The EV stations demand low at night power street lights using stored batteries, neighborhood surge shift load instantly, and this led to a 35% reduction in the city using it and 28% drop in the carbon output. And it was just efficient and it was sustainable. So that's where my smart traffic and power management came in. And let's do the next slide, which is how we rolled it out into the real time traffic here. The, this wasn't just a one day magic or something like that. It was structured and in a phases like, assessment phase and automation work pilot zone. Parallel rollout and all. So assessment phase is every duck pole and antione was mapped in the assessment phase in the initial phase. Then afterwards it's an automation framework. There, there comes our picture like, building A-C-I-C-D pipelines with the A bill for infrastructure and all this comes into the automation framework. And next phase we went to the pilot zone. We started in dense downtown zones, like where we have a a heavy traffic and also to test it out. So we did that in a, as a pilot zone phase. And after that parallel roll rollout using IAC and kops, we scaled citywide. So we treated infrastructure. Like software, like a clean, modular, testable, each deployment validated itself. And yeah, as I mentioned, this all conducted in a phases, so that helped a lot for us to move forward. So here comes the, a real impact, like a measurable impact, what we have achieved using this one, right? I'll tell you the numbers, right? So it's all about 23% lower commute times 42%. Of a faster emergency response and 35% of our energy savings have been showed up and 64% bump in a resident satisfaction. These are like an app approximate values. What we collected so far from the c, from the automation. This wasn't just a dashboards and data I was talking here, it was a people living better lives. Like when it comes to the commuter, whatever the automation we did using the 5G one CT manager said, we used to respond to the problems, now we prevent them. That there isn't that amazing. Yep. Yep. So that's what we achieved with the automation. And next to the scaling to the global cities. What we talk here about is a scalable architecture and open standards and knowledge transfer. So we didn't want to be a onetime success year, so we opened or we, so we open source the blueprint model which is a modular playbooks. Vendor, neutral tech and a training program for cities. So now as I spoke, we are working with the 15 plus cities globally, scaling up with the AI driven automation and adding privacy fast designs. So Citizen Trust is built in, this is the future and that, and it's, and repeatable, so scalable architecture, what. We, when we talk about is that our frameworks seamlessly adapts to metropolitan areas of any scale implementation can be faced. Yeah. So this is, as I told, this is the future and it's unrepeatable when it comes to the the path forward of what it is, expanding the partnerships and, enhance the privacy framework and advanced AI integration. So this all comes under the forward thinking, like a path forward, what I'm talking and he, it's just, it is a wrap up. To wrap up everything. So 5G isn't just a speed, it's about a smart infrastructure, and DevOps isn't just for a code, it's also for automating cities. Just as an example, as I mentioned, we are not just solving traffic pro jams. We are building a future where cities think, adopt, and care for their people in the real time. And thanks so much for listening my speech. I'm happy to take your questions and yep, that's all. And you guys have a good day.
...

Anil Thimmapuram

@ The Harrisburg University of Science and Technology, USA



Join the community!

Learn for free, join the best tech learning community for a price of a pumpkin latte.

Annual
Monthly
Newsletter
$ 0 /mo

Event notifications, weekly newsletter

Delayed access to all content

Immediate access to Keynotes & Panels

Community
$ 8.34 /mo

Immediate access to all content

Courses, quizes & certificates

Community chats

Join the community (7 day free trial)