Speakers:

• Carlos Pascual, Head of Connected Energy at Iberdrola
• Guillermo Menendez, from Iberdrola Digital Energy
• Juan Inaki, from Deloitte

Today's agenda:

1. Brief introduction

2. AWS for Energy overview

3. How AWS helps clients innovate and improve their systems

4. Carlos will provide a brief introduction about Iberdrola's energy platform

5. Juan Inaki from Deloitte will introduce and explain how they built the platform for Iberdrola

Guillermo: The idea today is to explain how we started from the beginning when Carlos called AWS and told us about this idea and project they wanted to build, and how we were able to move from there until having the platform in production now.

I will start at the beginning, talking about what I work for, which is AWS for Energy. I will also try to explain the different tools and mechanisms AWS has to help customers innovate. Then I will pass it to Juan Inaki.

You all know the energy industry is everywhere - from when you wake up at home, commute, go to work, travel, industries, commerce, etc. Everything is powered by energy. I want to provide a high level view of the industry today.

It starts on the right with the energy value chain - traditional oil and gas industry. It starts downstream where we extract raw materials, explore, produce. Then refining - all the transportation and storage of oil and gas materials. Finally, the downstream part is refinement and processing of these materials.

Together with oil and gas, we have other means of generation. First, baseload generation like thermal generation with coal, which is being reduced. Also carbon-free or less carbon intensive like hydro, nuclear. And finally, new renewable energy sources like solar, wind, geothermal.

Besides generation and transportation, there is distribution of energy to final customers - gas and oil pipelines, power grids, distribution companies. In parallel, there are water and wastewater management companies.

Finally, there are the customers consuming energy - retail, commercial, industrial. And a new kind of customer is the prosumer - not only consuming but also producing energy through solar panels, storing it in batteries or EVs. This is a main target of the Iberdrola platform Carlos will explain.

From the AWS for Energy team, we've seen some trends in the industry:

1. Electrification - around 50% from renewables like solar, wind, geothermal, storage, new business models.

2. Growing focus on energy infrastructure security - events like wars, wildfires, storms put energy supply at risk. Companies and governments want more robust, distributed energy resources. Increased cybersecurity focus.

3. Decarbonization of end customers - consumers producing and storing energy, EVs, monitoring and reducing carbon footprint. This is a focus of the platform we'll discuss today.

AWS for Energy is a global team, part of AWS for Industries. We combine cloud/IT with industry knowledge. Our goal is to support energy customers globally to:

• Optimize efficiency
• Drive agility
• Increase intelligence

We build solutions - assets like code, architectures, partnerships - to solve business challenges. Some examples:

• Carbon Data Lake - ingest/store carbon emissions data, calculate/report emissions
• Seismic workflow - accelerate interpretation, reduce costs

There are more examples in the AWS re:Invent Expo energy section.

Before moving on, I'll explain some AWS mechanisms and tools to help customers innovate faster:

Jeff Bezos said to invent, you must experiment. If you know it will work, it's not an experiment. Failure and invention are inseparable twins. Amazon is a place to fail, a way to move forward.

But intentions alone don't get you anywhere. At Amazon, we organize for innovation around four pillars:

1. Culture

2. Mechanisms

3. Organization/Architecture

I'll focus on Mechanisms and Architecture.

Mechanisms are behaviors, tools and programs supporting innovation. One is prototyping - in the middle between a demo and production system. We prove use cases and KPIs.

AWS prototyping teams help build prototypes - short, time-bound engagements, 6-8 weeks. We demonstrate possibilities, agree on focused use cases and KPIs. AWS and customer work together, so knowledge stays when AWS leaves.

We deliver enablement tools and a plan to get to production. This is the path we followed here - prototype, demonstrate KPIs, then production system.

On Architecture, there are practices enabling faster, cheaper innovation. I'll pass to Juan Inaki to explain.

Juan Inaki: We needed an architecture supporting rapid growth and quick change. During design, select technologies fitting technical requirements - performance, scalability, cost.

AWS and Deloitte recommend microservices - each handles a specific function, accessible only via APIs. This increases speed, agility for new services, and innovation rate.

Microservices make experimentation easy and low cost - you build small pieces vs. huge applications. Lower cost of failure.

So it is very easy just to to throw away the the proof of concept or the microservice and start from scratch again after some failures. And that means ok, this approach enable to uh make a rapid scale in case of success. So if you think that your proof of concept, it's fine, it's very easy to evolve and uh the services and iterate through different different versions.

So there uh also there is another point that uh according with a study done by the lot in 2019 80% of the time is dedicated to operations and maintenance. So there's a huge percentage of their time. So that means a very high cost, very low productivity. So now the developer teams is there is quite difficult to keep the pace with the with the requirements from the business point of view.

At the end, everything means that is a loss of revenue because of loss of business opportunities because you are not delivering fast note for for keeping the the pace. So we need to find out an architecture that just uh allow the builders to be focused on the business loic and provides a very easy way to develop to deploy microservices.

So this is where uh the serverless architectures come to help us with the serverless architect tools. It is very easy developer just to be in focuses on on the business and provide another additional advantages such agility. So that's uh you don't need to to deal or or manage the the platform just to focus on what you need to deliver.

So the performance is much, much better, especially from the stability point of view, you may scale whatever you want without any hardware limits. So this is a very a big advantage. The cost is lower than the traditional approaches. So that means especially in a pay as you go model. So when you only pay for you are consuming and finally you get a better security that is not uh an advantage per se from development point of view, but from the business point of view in general, it's a, it's an advantage we will see later. What does it mean as well.

When we think about serverless, the thing, the first service a w service that came out into our mind is land that is quite uh quite common. But we should take into account that there are other serverless services in other areas, such uh data stores like a o like the a o db three like serverless databases, integration, serverless tools that could be a pap a gateway steps even bridge. So there are other families uh serverless families in, in e double s world.

We talked before about the security of the advantages of the increase of security the serverless uh services provides. So in order to understand this statement, we need to understand what uh what the aws opression responsibility model is.

So in this slide, you see on the left side, the a w services that are closer to the on premise world that could be ec2 or databases installed inside dc. two. On the right side of the slide, you may see uh one has several less services like alanda or the a o db.

So the the operation uh operational responsibility model means that the farther you are on the left side, the more responsibility you have in terms of management and security of the solutions and the further you are on the right side. So more uh a w is taking more responsibility both in the management point of view and there is the security point of view.

So yes, uh our advice is to be the more to the, the better and just uh talking about just uh so this how this uh technologies just help uh eber line, the transformation of the, of the business.

Ok. So um before jumping into the consequences, um of course, everything you see here are running on electricity. Uh but i would like to ask you, do you know how much energy is needed uh to run all data centers for why not only aws uh data centers? I mean, the the figure is difficult is around, i mean, from a study coming from the international energy agency, it is around 300 terabyte hours.

That again may be a difficult number to understand how big it is, but it is around 11.5% of total consumption of electricity uh in the in the world, it's huge.

Ok. So uh let me introduce a little bit about ila because idola is um it's a european company. Well, it's worldwide, it's also in the united states. We are second um in the world first utility in, in, in, in europe uh with a vision on renewables.

Ok, we have more than 40,000 megawatts of renewable capacity uh around the world and we are saving more than 30 million customers and one of them are aws. And in that sense, uh we are helping uh aws in not only having a cloud reliable, that is uh their responsibility, of course, but we help them in try to have a cloud that is sustainable.

And in, in the other way, we have the help from aws to use the cloud to build those amazing uh products for our customers. Uh like the one we wanna talk about today. Um and to, to grow uh in this way, we sign a partnership between the companies uh recently this, this year.

Ok. Uh so as i said, let's jump into the consequences on all the introduction coming from, from guillermo la la la. It connects multiple smart devices at home under one umbrella. La. It is designed to satisfy users needs in terms of comfort while saving on their energy el incremento de la la la.

Ok. So let me go back to the introduction. Uh guillermo did uh about the trends i wanna focus and, and, and, and deep dive a little bit on two of them. First one is decarbonization and this is uh figures coming from europe. Uh the uh change has been and is been um amazing. And now wind and solar are producing more energy than oil and gas. And we are just at the beginning, we have uh target for 2030 very, very, very challenging. And this is something that is happening uh worldwide. It may be a matter of time, but everyone is going uh into renewables uh because of the sustain inability, but also because of the economics.

Ok? And this is putting a lot of pressure into the grid and into the power markets. Uh i don't know if you are really familiar to this kind of terminology but uh the the amount of energy, the amount of renewables we are having every day in uh in uh in some countries uh is making the energy cost, the price of the electricity go very low.

Um in the middle of the day when we have most of the solar production, it's going into €0 per megawatt hour. This is an example uh in april in, in the netherlands. Uh but again, it is happening uh in many different places also here in the states.

Ok. And um for the people that doesn't know this has a name is the dark core because of the shape uh of, of the core. Um but it's not staying just there. Uh the thing is that uh this is same month, same country and the netherlands uh you see it was not zero, it went down to minus €200 per megawatt. It is not, you are not paying for the electricity is that if you are consuming, you will be somehow being paid for that.

Uh so this is going to uh give um in our case, as a retailer, a lot of opportunities to build new products for the customer and the customer receives somehow uh benefits on, on energy management uh services. And by the way, i told you before, uh the previous curve uh has a name and there's not really a name for this one.

So uh um a few days ago, i asked my, my son if he could help me find out uh some animal or something that could uh seem similar to this. And he told me uh that iii i love uh dragons. I think that's the best uh possible name for this. And now ii i use always the uh the dragon curve when talking about this.

So the second trend guillermo introduced before was the home electrification. Ok. I don't know uh here, very interesting to know how many people are having solar panels ev or maybe a heat pan at home. Uh ok. Yes. Uh few leaders, yes, but uh things will change, you know, uh very quickly in idola we have and you saw in the video, we have a division that is responsible for developing this kind of services which is uh a smart solution uh is uh providing uh electric mobility services, uh solar panels and batteries, uh heat pumps and so on and for houses and, and, and buildings.

Ok. Um but going back in, how can we help our customers um to be more energy efficiency. I always like talking about um what we call the energy management value chain and it's trying to um to show like the different steps depending on the sophistication of the customer. Uh so the different ways we can help them in, as i said, being more energy efficient.

Ok? And, and the first uh steps are trying to give recommendations uh for them. They are uh looking on uh saying how much energy you are going to consume at the end of the month, how how big your bill or your energy bill is going to be or things like disaggregated your consumption into different appliances. So the customer can understand how much energy they are spending in the dryer, for example, or in the fridge or situations like identifying that the performance of one of those appliance uh is is not good. So the customer can know that there may, there may be a malfunction in, in some of them.

Ok? Um all the things that we are telling the customer is ok. These are the possibilities in this uh small solution um uh portfolio. So you can electrify your consumption at homes. But something very important here is we are giving uh recommendation, we are giving insights, but the one that has to do something is the customer?

Ok. So um that is the tomorrow. How can we do things so we can take some part of that responsibility uh to help them be uh more energy efficiency. Ok. Um um i i like always talking when, when describing a house or a building um two capabilities of the loads uh you have over there, one is the uh the control capacity which is how easy you can change a consumption from one time to another.

Ok? And the other is the the size of the load. So how big it is compared to the rest of the consumptions you have at home. So, so if you draw uh the typical appliances like the washing machine or the microwave or or or the fridge or whatever, uh you can see that it's very, very low on the left like not giving so much value.

But when we start drawing those smart solutions like the uh heat pump, the smart water heater or like the self consumption with batteries. And the ev as guillermo said before, it's like a i mean it's a battery with wheels. So you can see now they are higher, they are on the right. They are really giving uh giving value, ok.

Um now we are going to be able to draw what's coming next. Um but the additional characteristic, very important in those smart solutions is connectivity. Ok? We we we know now that they can be managed. They are very big. Uh but the key is that they can be uh connected and they can be uh monitored and controlled uh from outside the house.

So we can build services to help the customer uh do that control uh remotely. But we can also optimize that uh for the customer from outside the house and maybe uh create flexibility services uh for the network operator. So we can benefit uh the network as well.

Ok. We built a platform uh through this process. Uh we have been talking about uh with a product which is the advanced manage system, that is the product we are selling the customer, the platform, it is taking autonomous decisions through artificial intelligence to decide things like this is not the moment to charge the car, uh stop the charging session and wait till the moment.

We know the time of use tariff of that customer is going to be lower or maybe trying to move energy consumption to the hours when there is an excess of solar production uh in the house or maybe uh taking advantage of the thermal inertia. So we decide when is the best moment to heat the house or heat the water we use for uh for shower, achieving a more efficient use of energy and taking care of the planet.

This is what the advanced smart assistant does. The new smart solution by iberdrola. How does it work? It's really simple. First of all, you enter the appliances, your habits and your usage and comfort needs on the app. Second, the assistance artificial intelligence takes care of everything, adjusting and switching your electrical devices on or off. Third, the advanced smart assistant offers you increased efficiency and bigger savings on your bills, prioritizing the use of renewables and reducing co2 emissions

In short, more digitization, bigger savings on your bills and more environmentally friendly ela for you for the planet. So uh le let me focus on those advantages for the customer.

The first one is the digitalization. Um we are building this in a easy way through the app so they can monitor uh continuously. what is happening, something that is difficult is that we are doing things um uh outside of the, of the, of the view of the customer in the at the back of the platform. So we need to let them know uh the real value we are we are making here. Um so again, is uh money is, is is providing them with monitoring uh information, but it's also uh giving again recommendations so we can keep this customer to grow in the in the energy efficient.

At the same time, we are doing that uh that energy management. The second will be savings. I told you it is taking advantage of the time of use tariff. Uh they have and moving energy to the cheapest to uh but at the same time and it really depends on the country. Uh we can also do peak saving which is avoiding simultaneous loaded at home. So we don't have to pay or the customer doesn't have to pay uh for a lot of uh power um at that home. And of course, uh that is um local benefits for them. But there are also benefits we have on the retailer side, one will be moving energy to the hours and you saw before the hours uh sorry, the curves uh in the power markets. So we can move energy consumption to that times where the price in the market is the lowest. So the energy supply for the retailer is cheaper and there's a saving we can share with them.

And at the same time, i told you also before we can provide network services uh to the grid. So we can have an additional income that can be shared with uh with the customer again. And the third one is really obvious sustainability. It is trying to maximize self consumption in the houses. It is facilitating electrification because we reduce that power needs at at home. And of course, reducing co2 in the same way, we are moving energy into the cheapest hours. Those cheapest hours in the power markets are gonna be the ones with the more amount of of renewables.

Ok. This is the kind of um screens, it is more complex but this is the kind of screens we are building that are giving the customer those information that are uh facilitating them the possibility to give us their con sorry, their comfort limits. One of the things that is very important in this kind of platform is that uh we don't want to affect the daily life of the customer. So they asking things like um uh how many kilometers here or here is going to do next day and at what time they are gonna need the car or maybe what is the minimum temperature or the maximum temperature we can play at home. So we do not affect, as i said, their comfort.

Um and with that again, we start playing with the algorithms and give instructions to all the devices we have at their homes. Uh something um um kind of an ending um on what is coming next is i mean, the platform was built with uh one use case purpose that the residential customer i uh i told you but um the idea is that the platform at the end is a connectivity platform with different devices uh with all the information that is happening in, in all of them. So we plan to use that to build b to b services like uh micro grids or uh ev fleet uh energy management or using that to create uh esquire systems uh to understand what is happening in all those devices. So we can have a better after sales um service for our customers or maybe creating a virtual power plan so we can provide better uh network services uh to the grid.

Ok. So in conclusion, 33 ideas. Uh so the first one is we have talked about the smart solutions electrification at home. This is a way to integrate uh those smart solutions in an energy management perspective. Second is that we are being uh or we are anticipating the needs of the customer that is coming, the one that has huge amount of consumption coming from this new um devices at home. And the third one is that a new platform that is going to allow us to build new services and applications and maybe sell these kind of services to other companies that plan to do the same um with the customers.

So how we build this platform? So the platform was built in three different phases. First of all, idola has the vision, had the vision in mind. So what what they want the platform to to be. So with this vision, as uh we both uh iws and deloitte work for a proof of concept. So both uh the a as prototype team and deloitte were working around two months just to build a prototype. And then the prototype, we validate three main topics. First is uh if we were able to control devices at the home of the clients. Second, if we were, if we were able to connect to iberdrola systems. And the third one, it is possible, we need to validate the scalability of the solution. So the the proof of the concept was a success.

So we were able to fulfill all these requirements, especially the letters, one in terms of scalability, we validate in that uh in doing the proof of concept that we were able to create up to 1 million users in mostly. So after this proof of concept as applying the microservices, microservices um architectures, we start to iterate and to evolve this proof of concepts just to get the final solution.

So in the final solution, we uh provide and build all the business uh requirements i beto had in in their vision in terms the full functionality of the platform, we uh started to integrate all the devices we would like to start controlling. So we are talking about uh air conditioner, uh heat pumps, uh electrical vehicles, solar panels. So all those devices uh are controlled by the platform so we can control the devices of the user with those technologies.

Uh we uh we integrate with other idora systems. We integrate all the mobile applications, the the screenshot we we see we saw before. So we integrate the application, the mobile application with the platform. And we, and finally we integrate all the mature learning algorithm that is controlling the all the devices that the 100 of users from the technical point of view.

So this is a high view of the of the platform is 100% serverless. So all the, all the, all the products aws products are serverless that they give us uh an advantages that we see later. So we are using land a that is the core of the solution. But what we are using also a p a gateway are using said maker. For all the mature learning algorithm, we are using e bridge and we are using kinesis. All those services are uh organized in forms of a module.

So we have a model for the ingesting all the all the data we have mole for controlling the devices. We have model for everything about around mature learning. We have models for handling and storage, all the data in some sort of uh data lake. So everything is in modules that it is quite easy to add new functionality to the to the platform.

So all this solution has been built following the aws world architecture uh pillars uh with with operations in mind safety radi performance that we will see later cost optimization. And most importantly also thinking in in in sustainability as uh as i used before.

So thank you for, for using uh microservices. It's quite easy to extend this platform. That means that integrate new devices is quite easy. We have one model that is taking care of all connectivity with the devices. So in order to add a new device is just uh if we have a family similar to that device, for example, a new plug or a new uh a new solar panel. So we can either try to reduce the current phone extend that functionality with new micro services just in one model just to, to add a new device. And also it's quite easy to provide new, new functionality in terms of business. Because of the modular of modernity of the of the solution is a matter of new services, new micro services on some specific part of the of the solution.

Thank you for using serverless technologies. The platform is quite scalable. So uh in during this uh when we have built the platform, we also make some stress test. We, we prove that the platform is able to scale up 10 million currently concurrently users. So we didn't find out the breakpoint of the solution, but we stopped just simulating 10 million users because the injectors we we, we use, they, they couldn't have more workload to the platform.

So we are just what uh uh we think that the platform can escalate up uh whatever we want because we haven't found any bottleneck on the architecture. It's a matter of how many hardware we would like to put on on this uh the solution. And finally, from the cost point of view, the platform is 100 server less. So that means that the case is a pay as you go model.

So it depending on the number of users they they are consuming the platform. So it is the the amount of the amount of money you are going to to pay. So just uh as a rule of thumb, so the platform has been defined just to pay $1 a month per user. So it's a very low cost and is uh we have uh be be able to achieve this cost because of the use of uh of surveillance every time that the customer switch on the light when interacting with them and it's a great opportunity to do good, we face a great challenge as a society is to have a sustainable energy system and we want to help the customer, the citizen to have an active role in this transition.

Now, the question is how do we do that? Iberdrola is one of the largest energy companies in the world. We are a global company leader in renewable energy and with more than 35 million clients ev chargers, heat pumps, solar panels. All the smart solutions for idola are connected devices. So what is important here is how can we provide a tool that controls all these devices that is easily understandable and easy to use for the customers.

As i said before, the main goal is to help the customer, the citizen to have an active role in the energy organization. Now, from a business perspective, what we are achieving is to have a new way of interacting with our client. A more digital way, a more interactive way from a technical perspective, what we achieve is to connect via the cloud with different energy devices in a way that we minimize the use of energy, we reduce the bills. And also we help to minimize the impact on the environment.

Obviously, we cannot do this alone. I like it. Come on. Yeah, we have this very close relationship with the people in deloitte, victor and yi, as you said, savvy uh all working together every day, every week. Uh while we develop this platform, working with deloitte has been a great experience. They are very deep in the technical side. They are customer centric, they understand what our needs. And also uh frankly speaking has been very fun to work with them. A good partner, commits with your project, a good partner. It's with you in the good and in the bad moments, a good partner accomplished the vision you to share together a cloud revolution is gonna be huge in the energy sector.

I live in a house that is already fully electrified. Connected to this cloud. I am passionate in this development, being able to expand these capabilities to the rest of the customer in spain and in other countries. Idola has presence.

Ok. So yes, uh i would say said thank you for attending this uh this presentation. I have uh do you have any questions for any of us? Yeah. Get bus up. Yeah. Uh well, for the for the point on this um i wanted to know if this, for this product customers and also.