Hi Elodie, what do you do and what brought you to Asia?
As a kid I lived abroad with my parents and loved the experience. I have always been eager to explore what is out there and get to know more countries and cultures. Since it would imply many changes at the same time, I have decided to focus first on finding a professional opportunity matching my expectations, and turns out, it was in Singapore!
I was looking for a diverse and stimulating job in a start-up in the energy sector, and ideally with a connection to the IoT (Internet of Things) sector. And that’s exactly what I found!
I am now the Managing Director Asia and Head of Sales Asia for BeeBryte. My mission is to develop the Singapore office and its outreach in Asia.
BeeBryte improves buildings’ energy efficiency and users’ comfort using AI prediction and patented technology. BeeBryte is providing an IoT Gateway along with a software-as-a-service to optimise energy consumption of commercial and industrial buildings. It uses weather forecast, building occupancy and activity, and energy prices as input, to automatically and dynamically control air-conditioning systems to make them more efficient. Artificial intelligence is allowing us to grasp all the complexity of the systems while optimising the comfort, and reducing the costs. For instance, anticipating the rain to avoid the all-to-common freezing within the building.
My combined engineering and business backgrounds are the right skillset for the job! After almost two years on my own in the Singapore office doing a bit of everything, we are now 6 people!
Could you please tell us more about innovating for energy efficiency and conservation?
Oh, I could talk about it for hours… For me, energy – production, consumption, carbon emissions… – is the number one challenge to be solved in this 21st century. Humanity today cannot live without electricity, but each kWh consumed has a direct carbon impact and threatens ever more the ecological balance. And it is true especially here in Singapore! 97% of electricity is produced by gas. This means carbon emissions here are almost 10 times higher than in France for 1 kWh consumed (419g/kwh versus 58g/kWh) where the energetic mix is quite different (nuclear, hydroelectric..)! (To compare the carbon emissions of electricity produced in each country you can check this cool website.)
It is hence urgent to change the way we produce and consume electricity.
In the past years, a lot of effort has been put into improving the carbon intensity of electricity production. Renewable energies are slowly getting more and more efficient, cost effective and deployed globally. Further innovations and improvements will for sure emerge in the coming years. In spite of all this, huge investments are still being made in coal and gas plants, producing carbon-intensive electricity…
What has not been addressed and is only appearing today on the market is how can we modify our energy consumption. Obviously, consuming less and being more energy efficient should be the first focus. But it is not the only way consumers can help.
To explain how we can help, I will go back to the basics: electricity is very difficult to store. Batteries are still very expensive and have an efficiency around 80%: for 1kWh stored, you only get 0.8kWh back. However, the electrical grid absolutely needs to be balanced, matching exactly the production with the consumption: otherwise it’s the blackout.
Up to now, production plants have been entrusted with adapting production with consumption. However, most renewable energies are intermittent: they do not produce all the time and are dependant on weather conditions mostly (e.g. solar panels only produce only during the day, when the sun is out). It is now up to us to adapt our consumption to what is being produced, maximise our consumption from renewables, and avoid the need for dirty peaker plants that are most of the time coal power plants.
But how can we know when we should reduce or increase our consumption? Well as most things in this world, through a price signal! Power plants don’t actually sell electricity directly to the consumers, they sell it on a market to electricity retailers who then resell it to you. This market is pretty much like a financial market, price varies every 30 min (in Singapore) and depends on supply and demand. In some countries it even goes down to 5 minutes! We just have to follow the market!
This is also one of the services BeeBryte can offer: we have access to the real-time market price, generate forecast, automate “price arbitrage” not compromising on the level of comfort, and obviously share monetary benefits with our customers.
For instance, if we predict that in half-an-hour the price is going to be more expensive (i.e. more carbon intensive and/or need for lower consumption to avoid blackout) we will pre=cool the building (to 23degC) – maintaining the comfort level as decided by the customer (i.e. between 23-25degC) – so that when the price actually goes up we can use the thermal inertia of the building, let the temperature rise (to 25degC) and lower the overall electricity bill. We won’t change the quantity of electricity consumed but only the timing.
What are people’s literacy level on energy efficiency in South East Asia?
As for many topics, it is difficult to talk about Asia as a whole.
First of all I will differentiate two points: Energy Efficiency, and Energy Conservation. Energy Efficiency is about using less energy to provide the same service (providing cold air for instance).
Energy Conservation is the effort made to reduce the consumption of energy by using less of an energy service. To illustrate, you can have the most energy efficient air conditioning system, but if you’re letting it ON at night when no one is the office tower, you are still wasting energy and would need an Energy Conservation strategy.
In Singapore, air-conditioning is definitely a sign of luxury, hence they’d always rather have too much than not enough: setpoints at 22 or below, shops’ doors wide open to the outside, etc. With the GreenMark initiative from the Building Construction Authority (BCA) most large buildings are looking at energy efficiency and making efforts in that direction. To be honest though, I have seen completely unreasonable and unnecessary consumption even in the highest ranked energy efficient buildings. Mindsets need to changed in favor of energy conservation. Even in my condo, lights are on 24/7…
With very large and international companies, energy efficiency and conservation initiatives are really about reducing their carbon emissions, for most of the others it’s mostly about the money. It’s a good way to start!
Most of other Southeast Asian countries are developing countries. The first question is very often access to energy for a non negligible portion of the population. It is easy to understand that energy efficiency measures have quite a low priority in that context… But those people still consume way less than us in our air-conditioned homes, with our two computers, smart-phones, washing machines…
What is the added-value AI brings to energy management?
I come from an IoT background, and it’s no surprise to anyone today that the number of sensors and data produced is increasing every second. And with this amount of data captured comes an amazing opportunity for optimisation! It is obviously also the case in the energy sector.
If I go back to the example of an air-conditioning system and the environment in which it is installed, it is an extraordinarily complex system. In order to control it, optimize it to reduce consumption, but also improve the comfort, you need a comprehensive approach of the system to understand how it behaves and what impacts it.
One option is to build a physical model or BIM (Building Information Modeling) but that requires 1. a lot of time, 2. a very good knowledge of EVERY parameter of the building: the material in which the walls were built, the number and orientation of the windows, type of glazing, roof insulation, etc.
The other option is to use a data approach! Build a model and understanding of the building purely by looking at the temperature, activity and consumption data. Singapore has no distinctive seasons. This means that one to two months of data is enough to build an efficient building model that could have taken longer in temperate regions! In any case, during the first year we always check regularly and retrain the models to ensure we capture the business seasonality as well. And bonus, thanks to Machine Learning, this model is going to keep learning and evolving with the building itself: you change the use of one floor or redo the isolation of the building, just feed the algorithm the new data and we’re good to go!
The other super interesting part of using AI compared to a physical model is speed. Once the AI has been trained with the model, it requires a lot less computing power and time to achieve the same results – hence more efficient!
What do you think of the availability of AI talents in Singapore compared to a country like France?
When the company was created 4 years back by two French founders, the lack of AI talents on the Singapore market was actually one of the main reasons the headquarter moved to France: very soon the Singapore office was abandoned. Indeed, at that time there was a lot more talents on the market in France, since it had been previously implemented in university and engineering school programs.
From a start-up point of view, it was also much more fashionable to work in a start-up in France than it was here. Things are now changing, but it was definitely striking a few years back when large companies where setting up their global AI R&D center here, and we couldn’t find anyone!
Thankfully, and in line with Singapore’s objective of being an AI and more generally deeptech hub, things have started to change. Among the efforts that I feel have not really emerged in France, are transition programs. University programs have evolved but it’s going take a few more years for students to go to market. In the meantime short training programs (3-9 months) have been put in place with a project coming directly from the industry, for instance, thanks to the collaboration of AI Singapore and iMDA: AI Apprenticeship. The aim is to rapidly shift experienced people to this market.
To be honest it has been difficult for us – for whom it is our core business – to participate in those programs. Most of the times, they were more developed for companies who don’t have the skills and needed them for a specific and well defined project.
The brighter side is more people with AI skills are now on the market, and we’re really looking forward to seeing more! And this allows us to build a real hub here in Singapore to address Asia, not only for sales and operations but also for R&D.
How “Dare you” to be a women in tech?
To be honest I find that question a bit strange, it seems so obvious to me that everyone and anyone should dare to be what they desire! People’s first impression on who you are, what you do, shouldn’t dictate your life. Only your own desire, passion and experience can define what you are good at and what you want/can do.
That is my strong belief, and I can get into quite heated conversations about it. But that doesn’t mean that it is easy. I can’t count the number of times where people have asked me to bring one of my “technical colleagues” to the meeting, or to turn to my male colleagues when looking for seniority, because they are not expecting a young woman to have such a deep knowledge on technical details. But to be fair, I have definitely made the same mistake!
But I can tell you, there’s nothing more satisfying than spending the first 5-10min of the meeting showing them that you really know what you’re talking about and that you should be taken seriously. And in most cases, it actually earns you a lot of respect from whoever made the mistake in the first place.
How do you intend to bring more diversity in AI?
I think the first step is to recognise that we all have biases. ALL OF US, no exception. And because it is easier to find common ground with people that are similar to you, well it is easy to build a company, a group of friends, with people that are all thinking the same way!
Oh how easy it is, on a personal level, when you arrive in a new country to contact the French living there, and recreating a little France around you in this new country. But there is so much to learn from other cultures, other mindsets.
Don’t get me wrong, I love the French community here, and I’m proud to be a part of it today. My point is: I didn’t start there, because it would have been way too hard to get out of it in a second time.
The same applies in our professional lives. It is important to realise that innovative solutions, game changer solutions, come by confronting different point of views, experiences, and expertise.
As a French engineer, I give a lot of value to that status, and trust this path of education and the input a French engineer can give to a company. But I also have learnt so much from my Singaporean and international colleagues that have followed a different path of life. For me, diversity is essential to innovation and progress.