MOEEBIUS Mafra Jardim de Infancia Venda Pinheiro (Picture: ISQ Group)
In Mafra, machines are learning to care for humans and for the environment. As part of the MOEEBIUS project, engineers are developing a software that let’s machines discover human preferences, and how to save energy while meeting these needs. People in city hall, a school complex and a kindergarten, where the experiment is taking place, are shown the energy impact of different levels of light, heating and cooling and so on, and can then set the levels that they are most comfortable with.
They can change these levels as they see fit, depending on the weather, time of year and other variables. The software comes to understand general preferences by cross-referencing all the choices that people made over the course of its use, becoming more and more apt at saving energy by predicting minimum appropriate levels.
Interview by Anthony Colclough, EUROCITIES
Combining the variables
The Energy Plus software can combine these variables with input from motion sensors, so that it knows how many people are in the building and at what locations, and minimise energy expenditure accordingly.
The software has one further capacity: by looking at the energy use of different appliances, such as the air conditioners, and comparing that to the air quality and temperature, it can determine if these machines are working properly, or whether they need to be repaired and replaced, and even compute the most efficient time to intervene.
Does not compute
Once it got going, this project demonstration site encountered three major challenges. The first of these was to ensure that all of the platforms, software and sensors were able to communicate with each other.
A common pitfall when cities begin down the smart route is that they discover too late that the systems they have invested in are not interoperable – now they are stuck with a smart parking system which cannot receive congestion data, or a district heat network that cannot read the thermostat of the smart energy meters in people’s homes.
Now you’re speaking my language
To avoid this bungle, the MOEEBIUS team made sure that all the programmes and systems they used were ‘open source’, in this case with RabbitMQ, meaning that they can be used and modified by anyone, without payment or permission. This ensures that they can be integrated not only with each other, but with any future technology the city decides to use.
MOEEBIUS Mafra Edifício CMM (Picture: ISQ Group)
Human versus machine
The second issue is one that can be equally catastrophic, but which often receives too little attention until it’s too late – working together with the people who use the buildings in their day to day lives.
Given the project’s mandate, to create environmental benefits at the same time as increasing people’s comfort, the team was surprised to encounter some resistance among those working in the chosen buildings.
The team used two methods to get past this resistance. The first was talk, talk, talk. By building relations with participants, ensuring that their concerns are heard, and that the details of the project are clear to them, engineers can foster an environment of trust, where participants are more willing to accept change and disruption. The second was simpler: choose people who are already enthusiastic about energy efficiency – the buzz they create may even spread.
Toil and trouble
The final challenge was the only one that was truly insurmountable – tapping into the huge old boilers used by the public swimming pool. The city was very nervous about damaging the enormous, expensive and ancient boilers, and did not want to conduct an intervention that might lead to costly problems. The company hired to install the sensors and other equipment agreed – it was just not worth the risk.
To get past this, the decision came to do a new tender for non-invasive sensors, but the tendering and procurement processes took such a long time that in the end it was pointless to install the sensors. The monitoring phase would be so far along that the data would be useless.
“In the next project” said Filipe Neves da Silva of Instituto de Soldadura e Qualidade, “we will deploy the technologies at the beginning, so that if we get a problem, we have time within the project to deal with it.”
Start with the basics
So, with all this sophisticated analysis, which interventions had the greatest positive effect relative to the level of investment necessitated? According to Mr Neves da Silva, it’s just what you might expect: replacing normal lights with LEDs represents an enormous saving requiring a relatively low investment. This is an intervention that makes good sense for absolutely everyone to do everywhere, and yet there are still so many old an inefficient bulbs in use.
The same is true for additional insulation, both of walls and pipes. The payback period is short and the benefits are well known. “Everybody knows this, but not everybody does this.” Mr Neves da Silva lamented.
A more novel insight from this project is related to air conditioning. If a room is set to 20°C, when it goes below that temperature, warm air will kick in (and vice versa). However, by looking at the tendency of the room (towards hot or cold) as well as its occupancy, the team was able to introduce another energy saving option: just turn off and let the room correct itself. This is to say nothing of famously wasteful habits like having the air conditioning and/or heating on and the window open at the same time. People must be conditioned to avoid such blunders.
MOEEBIUS Mafra Escola Básica da Venda do Pinheiro (Picture: ISQ Group)
Recommendations for future policy
So, given the viability of these simple solutions, is there any need for the complex sensor and simulation technology being developed by the MOEEBIUS project? Mr Neves da Silva says yes, absolutely, but not immediately.
These technologies have a real and considerable positive effect on energy use, and can be employed to achieve massive reductions, especially if they are taken up on a massive scale. However, for the moment they are also very expensive, and looked at from a purely financial point of view, it is difficult to make a case for investment as things stand. Once the demand increases for technologies like this, the price can be expected to come down. When that day comes, these sensors and systems will become as much of a no-brainer as LEDs and insulation.