Good afternoon.
My name is Rik Barillot.
I've been a core member of Open Remote for a bit more than… louder?
Okay, sure.
A core member of Open Remote for a bit more than 12 years now.
I'm not the person who was supposed to give this talk, so I'll do my best to work it
through.
Don't hesitate to come back afterwards, and I can point you to some of my colleagues
that worked on those projects.
A bit away?
Yeah.
Okay.
Okay, I'll do my best.
And speak louder?
Okay.
Okay, so Open Remote, it's a 100% open source IoT platform, so it would do whatever you
expect from an IoT platform.
Back to the devices, have some logic, and user interfaces.
We'll come back to that a bit later.
So open source, fully free, available on GitHub, and a community throughout the world that's
pretty active.
But also some projects that we work on with some companies.
That's mainly what the core team does when I said professional.
It's working on those projects.
Also they have projects that are in home security or smart cities, typical IoT projects in more
exotic things like smart clothing, architecture, and of course a lot of projects in the energy
domain, energy management, but also some link to other aspects of energy.
And we'll go into a bit more detail in the Nottingham city project a bit later.
So looking at Open Remote, what is it?
It's mainly a middleware developed in Java.
It has a database that is both for the configuration of the system and for the state of the system.
So the current values of your sensors, but also all the historical data.
It has quite a few connections using standard protocols, so you can connect to gateways
or to data feed.
We'll see that later.
Awesome property hardware.
It has a set of user interfaces.
You have standard more management user interfaces where you can configure the system or see
the values or trigger some actuators.
You get Insight, which is a dashboarding kind of application.
But we also have a set of web components, freely available that you can use to build
your own custom application for a given project.
And so you have an application that you can access through a browser, or you can embed
it into a mobile app, what we call the consoles.
And you can also connect to other systems like Grafana, Power BI, if you want to have
extra features.
Then you have, of course, a mechanism for the logic.
We support different type of rules engines, simple through the UIs like IFTTF.
So if then that or more advanced features like Groovy scripting.
So if you want to go really deep.
There is a set of default services, so building blocks that you can use, for instance, to
push notification to the mobile phones or to place devices on a map or to implement
optimization services, what we'll talk about in a minute.
And this is, of course, built with security in mind.
So there is a strong identification, authentication and authorization layer in the system.
So coming to energy optimization.
We'll talk about two things.
As we say, what we call smart home, but it can very well be a smart office or even an
office complex.
Basically, it's the concept of an island behind a meter.
And you have kind of a sole proprietor of the island.
And then when you move to the smart district, it's a composition of many islands behind
one transformer.
The problems are a bit different, but the system is the same.
So if you look at the system, yeah, whatever, I'll do this.
You have your renewable energy, so solar and wind.
You have the grid, both import and export.
You have a battery with charge discharge, and you have your load, your consumers, but
can also sometimes feed in energy back into the system.
Some electric vehicles can do that.
So the goal for the smart home is to optimize either based on the cost, so you want to pay
the least amount, or on the environmental footprint, so you want to be green as much
as possible.
The data that we have to do that is for the renewable energy, we are going to estimate
the consumption based on the peak characteristics of the installation, so how much your solar
power can produce, solar panel can produce, and on weather data, so we can take the estimate
of that.
For the grid, we have dynamic tariffs, so people can, for instance, have contracts where
they pay a different tariff by the hour or by the quarter even, and so we have the data
to know those costs, but there is also a carbon cost associated with the type of energy that
is produced.
The battery, it's a charge discharge, but there is also a cost, so a levelized cost
of storage, so for instance, if your battery costs 1,000 euro, and it can do 1,000 charge
discharge, every charge is charged cycle is 1 euro, so you need to take that into account
when optimizing, and so for the loads, we have the path consumption, and we do a weighted
exponential average to predict the future consumption on that.
So now what we are trying to optimize, as I said, is minimizing the cost of the carbon
exhaust based on all this data.
And so the system will control what we call the flexible load, so depending on this data,
it can decide when to charge or discharge the battery, it can decide when to charge or discharge
potentially the electric vehicles, or it can decide to control heavy loads, like heat pumps
where you have a bit of freedom and when you can power them up or the temperature set point,
things like that.
And this can be automatic of course, but it could also be simply manual by pushing information
to end user through the UI.
When you move to the smart district or the collection of island behind the transformer,
you have a slightly different problem, which is the transformer that is between your district
and the grid, which has a peak capacity, and so what you want to make sure is that you
stay under the capacity of the transformer, both for import and for export.
So when there is a real high production of renewable, you don't want to surcharge the
grid.
So the data that we have is basically the same for the battery, for the renewable and
for the loads.
In addition, we have real time peak power, not peak net power of the transformer, so we
know how much the transformer is currently taking in and out.
And we also can then adjust the optimization algorithm with a fake kind of tariff.
So if we know that we need to change the consumption on the transformer, we can like fake how much
the electricity would cost so that the optimization algorithm would steer one way or another.
And so we keep doing the optimization at each individual island, but we want to push
for the global optimization so that the grid stays or the transformer, the grid stays under
control.
And so one additional problem comes now with the fact that you have many households, for
instance, in a district, which can have their own technology.
So it's quite complex to control them, to automate them at all.
So one way, and we're exploring that, is interfacing with more home automation systems,
like Open Hub or Home Assistant, for instance.
Another way is to manually impact.
And so what we can do is send personal challenges to every household where the people can earn
points, which basically earns them money if they play nice within the whole ecosystem.
And there is a lot we can do is we also have shared flexible loads.
So for instance, in a district, you can have the shared charging station for the electric
vehicles.
And then we can control and, for instance, diminish the available power so that we can
also keep the grid under control.
So that is the general idea.
That is what we are aiming for.
There are several pilot projects that are starting to implement that.
So this is the global idea.
One of them is the Nottingham City Council.
The idea, it's a smart home, but really it's more smart, well, we could say office complex.
The idea is to control the charging of all the vehicles, electric vehicles that are used
by the City Council at Nottingham.
And so what it means is you can control a global static battery plus the charging of
all the vehicles to save money.
You can also control or you want to have your vehicle charged at least to some level because
you want to use it in the end.
And you also want to prevent surpassing the limit, the power limit that you have for the
whole district.
Oh, sorry, Council.
And so what you see on the right is the dashboard interface that we have in Open Remote that
can show you the different location of the vehicles.
So we can track that anonymously, but we can track the different vehicles and the global
power that is currently used by charging of this vehicle.
If we now move to the smart district, this is a project that is currently starting in
Amsterdam, where we have a community of about 500 households that are part of this project.
One thing is each household can control their consumption by we interface with the meter
and they can see a real time information about the power they're consuming through the mobile
app so they can adapt their own consumption.
We have the challenges that I talked about so they see how the whole district is doing
and their proposed challenges so that they can play nice within the neighborhood and
by doing so earn money.
And we can also, as we said, limit the if there is really an emergency, we can control
the heavy loads that are shared for the district to make sure that we don't go above the limits
of the transformer.
So it looks a bit like that and these are design of slides so there are some inconsistency
in the wording, but globally every participant will see his own consumption with a bit of
a history on how the district is doing.
And the green dots around the indication are a global indication of how the district
is doing.
So it's really gamification there.
Now you see that at some point the neighborhood might be reaching the limit so we are reaching
the limit at the transformer level and so we will propose to the person in each household
a challenge saying well for the next hour you need to keep your consumption below this
level.
If the person accepts then for the duration of the challenge they will see their own consumption,
see the limit, how they are doing against it and how many points they will collect.
And so they also receive tips, say well potentially if you want to keep your consumption under
the limit maybe charge your car a bit later or set the temperature a bit lower, something
like that.
When the challenge is done they see how many points they have collected and then they of
course can see a summary of all the challenges they have completed, how many points they
have earned, etc.
This is the view from the manager so we can see different meters that are all connected
to the system.
At this stage as it's pilot project they have 50 meters connected, the project just started,
the target is to have 150 by the end of the month February and with 150 this should be
enough to already influence the whole behavior of the district.
So with 150 connected meters we should be able to have an impact really on how the district
and the impact on the transformer.
And so this is here the dashboard where you see a summary, the small diagram I showed
with the consumption and the load on the transformer, how we are doing compared to the peak performance
of the transformer, a historical graph and things like that.
So thank you, these were the two projects that are currently running on energy management
at this stage, there have been others.
You can find the open remote platform in the GitHub repo, there is also the forum where
the community is active and other information.
Thank you very much.