Priyanka Bagade demonstrates an intelligent infrastructure and shows how to gather data from a traffic camera and a road sensor. This demonstration uses an Intel® NUC to process real-time data and send it to the General Electric Predix Cloud*.
Hi. I'm Martin Kronberg, and welcome to our first IoT Developer Show. If you didn't get a chance to see our show announcement last week, this is going to be a monthly show all about IoT where we're going to be showcasing some awesome projects and interviewing the creators behind them.
With me today is Priyanka Bagade, who is an engineer at Intel. And she has a really cool intelligent infrastructure demo that we're going to show off. Priyanka, thank you so much for being here. Could you tell us a little more about what you do at Intel?
Sure, Martin. I'm an Internet of Things Developer Evangelist at Intel. After receiving PhD in computer science from Arizona State University, I immediately joined Intel. Currently I work with cloud partners to integrate their cloud technology into Intel IoT platforms.
That's really cool. And you know, I'm really excited to see these kind of collaborations happen. Well, let's take a look this demo. I'm really excited to see it.
Sure. Here is the intelligent infrastructure demo which utilizes existing infrastructure and IoT technology to make our cities smarter and safer. So in this demo, we are using Intel NUC to do each compute and we are using GE Predix cloud to do historical analysis of the data.
Very cool. So what are the infrastructure pieces that you're modeling here?
We already have the traffic ports deployed on our streets which have traffic cameras. So we are using that. We have emulated that using logical camera here. Also we are using the inductive sensors on the road which are being used to detect the car on the street. We have emulated the inductive sensor using Hall effect sensor.
So how are you analyzing all of this data?
So what we are doing, we are taking video feed from this camera. We are sending that feed on Intel NUC. Intel NUC is using computer vision and machine learning algorithms to process that data to compute speed of the car. So when the car is moving on the street, we are calculating ROI. That is the region of interest for that particular street. And then we are converting that ROI into [? counter. ?]
And when the car is moving, we can see the blue vectors. We are computing magnitude of those vectors with respect to length of the dotted lines on the street. For each country, this length of the dotted lines is fixed. For example, for US, it's 10 feet.
And after obtaining magnitude of the vectors, we are calculating average of those magnitudes, which gives us speed of the car. And then we have also developed a web server application which shows what kind of computation is going on on the NUC. That is, the H [INAUDIBLE].
So whenever I'm moving this car, you can see the actual speed of the car on this graph. And if I move the car faster, we can see that it says "Danger." That means the car is orbiting through.
After that, we are taking the Hall effect sensor data to see how many cars are actually passing through the street. You can see a nice blue light here which shows that car is actually passing through that street. We are taking the speed of the data as well as how many number of cars pass through that street, and sending that data to the cloud to do historical analysis using the data visualization and analytics service.
So are systems like this in place right now in the real world?
That's a really good question, Martin. So GE Current has deployed lampposts on streets of San Diego. The project is called CityIQ. They are collecting data from the infrastructure and processing that data on Intel [INAUDIBLE] Gateway, which is in-built in that lamppost. And they are using this collected data in applications such as smart parking, pedestrian safety, and environmental planning.
Really cool. Well, it sounds like we can start leveraging existing infrastructure and making it more intelligent and make our cities, like you said, safer and smarter, more efficient.
It's really great. Priyanka, thank you so much for coming here and showing us this great demo. If you guys want more information about it, we're going to provide some links. And Priyanka, once again, thank you so much for being here.
Thanks, Martin, for inviting me on the show. I really enjoyed it.
So before signing off, I want to talk about a couple exciting events that are coming up. First, we're sponsoring the fourth IoT Open Challenge hosted by the Eclipse Foundation. Submit your proposal for an IoT solution around smart cities, industry, or remote help by November 13 for a chance to receive hardware development kits and really cool prizes.
Second, we're also going be having an online global IoT DevFest on November 7 and 8 to explore all things IoT through keynotes, demos, and mentoring opportunities. Thanks for watching this episode of the IoT Developer Show. Don't forget to like this video and subscribe to the Intel Software YouTube channel. And remember to check out the resources and the links provided as well.