Pablo Betancur:
What's happening here? Basically we have here intelligent distribution solution to predict the load inside the transformer district, for example, in your residential area, and also to manage solar power in distribution, like solar panels in your house, and also to redirect the current to be used in better ways.
These curves represent three things. The load in the transformer when it's here is negative, it means the person have a solar power in the house is producing more energy than being consumed. But this inverse current will cause inverse current in the transformer and can damage the infrastructure. So what we do is we put intelligent solutions to predict and also to shape this negative peak. So how we can do? Basically we can have a battery system in that site and for example, now the current is being controlled. When we have over power, we charge the battery. And actually here the peak shaving is not negative anymore because we switch to batteries and basically you can reduce this energy at night.
Steve Saunders:
Okay. How did you do this before you had AI?
Pablo Betancur:
Traditional metering is not able to manage this kind of inverters or battery systems. Then they have a smart meter system they call AMI, this kind of system can only do the management of metering but no inverters. So they will have three different systems. With this one basically it's one platform for all. More efficient, more intelligent, and also unified management, unified platform.
Here we have the architecture of this kind of system. In the cloud we run all the intelligent systems like the distribution transformer management system, message flow, another unified system that help to manage the whole equipment and devices.
But what is important, we took one ECU is in the transformer, but in fact this ECU needs to have a way to communicate back to the command center because we are talking with the cloud. So in this way we have the pipe. In the pipe, Huawei is expert in this area, we have the private wireless network or we have the optical fiber network as well. [inaudible 00:02:21] here we have several technologies. Now we have the edge, which we already talk and explain how it works, is the ECU, edge computing unit. But below this we have a protocol we call HPLC, high-speed power line communication. So in this way you don't need to install new cables. You use the same power line to communicate with meters, solar inverters, battery system, and electric vehicle chargers.
Steve Saunders:
This is actually the classic architecture of next generation industrialization, isn't it? Because you've got your IoT going into with all of the information, the data being brought up and collated at the edge of the network, that's really where the action is, and then sent into the cloud using private 5G or optical. And then the cloud is taking care of business, all centralized. It could be anywhere, the cloud. I mean the cloud could be in Shenzhen and the electrical system could be here in Shanghai, right?
Pablo Betancur:
You're right. In fact, in so many other maybe applications, no electricity, they can call industrial IOT, similar architecture. But here's dedicated to power.
Steve Saunders:
And it seems to me that the energy industry, the grid, is really moving ahead very quickly with this, isn't it?
Pablo Betancur:
When we talk about smart meters, sometimes these meters will help you to know how much is your power consumption, but this is mainly sometimes just to digitalize the information. What happens is because the electric vehicles are increasing the load, it will have additional load in the system and also so many people is generating power. So you need intelligent distribution solution or intelligent distribution systems to be able to control all these variables to protect infrastructure.
When we have the power grid, we have many assets in it. We have generation plants, we have transformation and transmission, actually you can see intelligence substations, you can see intelligent power distribution and you can see generation plants, hydro, maybe you have solar. So this kind of system, will help you to manage all the assets of the power grid. So, in the power distribution, you can see in this map, the energy generated in solar panels in the distribution grid at a city level, imagine you have 1 million, 2 million, 3 million solar panels, how you manage these independently? It's impossible. You need AI and a smart systems like this one.
And in fact, if we click one of these transformer or even meter districts, we can in fact see the device level, the meter, and also all the infrastructure and manage. When there is some failure, they will detect and send to the operator to go to the field. Now we have intelligent systems to predict maybe some problems and also to analyze the data. But the final decision need to be taken by the human because still have some operations need to be human. So we have a combination of both. It's more like AI assisted than fully AI operated.
Steve Saunders:
That makes good sense. Thank you, Pablo.
Pablo Betancur:
Thank you.
Steve Saunders:
Nice to meet you.
Pablo Betancur:
Nice to meet you. Thank you.
The editorial staff had no role in this post's creation.