Solar Battery » Ask the Expert: Luke Peck, Data Science Team Lead

Ask the Expert: Luke Peck, Data Science Team Lead

By Mara Jun 29, 2022

Ask the Expert: Luke Peck, Data Science Team Lead

For this week’s “Ask the Expert Series,” we caught up with Luke Peck, Data Science Team Lead at Moixa. Luke has been working with us for over a year, and his role focuses on managing the team that builds data-driven applications to optimise users’ batteries and EVs.

Before joining Moixa, Luke did a PhD in Astrophysics, where he began to hone his problem-solving and analytical skills. He then worked for another start-up company in the energy industry, helping them create their energy supply business from scratch and break into the UK flexibility market with several National Grid services utilising grid-scale batteries.

Luke sat down with us to discuss his team’s work on our GridShare software, how AI is critical in the fight against climate change, and so much more. Read on for the full interview.

Hi Luke, thanks so much for taking the time to talk to us. First of all, we wanted to ask you, what made you want to take the job at Moixa? Do you think there’s something unique about Moixa’s vision?

There were a number of reasons I took the opportunity to work at Moixa. Throughout the entire company, everyone is genuinely interested and motivated to do their bit in pushing toward a more sustainable future and tackling the climate crisis. For us, this means reducing electricity bills for homes and pressure on the network during peak times by utilising batteries for behind-the-meter optimisation and for delivering grid services.

I was also impressed by the sheer scale in terms of the number of households on Moixa’s GridShare platform, optimising tens of thousands of sites worldwide! Our aim is to optimise the world’s batteries and what is unique about Moixa is that this is not just rhetoric; we are rapidly growing the platform all the time.

Can you describe your role? What is your main focus area?

The main focus for me and the data science team is optimising our GridShare software, Moixa’s customisable cloud-based platform. This includes behind-the-meter optimisation of residential batteries to reduce our customer’s electricity costs and carbon footprint.

We do this by generating predictions for household consumption as well as solar generation if the house has PV systems installed. We use local weather forecast data, data collected from each home and machine learning to produce site-specific predictions, which are then fed into our AI algorithm along with the customer’s electricity tariff to minimise the cost for the household.

GridShare then generates a personalised daily plan for the battery based on these predictions, enabling customers to get maximum benefits from their self-generated power from both an environmental and a financial standpoint.

We also optimise electric vehicles (EVs), scheduling the best time to charge from either solar PV if available or during cheaper periods if the customer is on a time-of-use tariff.

We are also responsible for delivering extra value to our customers and partners by participating in grid services to help balance the grid. External market signals from energy suppliers and network operators are issued against our fleet of participants, and we optimise the cheapest way to fulfil the request.

When GridShare receives a command to dispatch, it uses AI to provide flexibility dispatches most efficiently in terms of monetary savings or carbon emissions for the fleet as a whole while still taking into account the behind-the-meter value delivery to the end customer.

Read about our journey to building and scaling GridShare and its architecture.

How do you think artificial intelligence is helping fight the climate crisis?

Artificial Intelligence (AI) is vital in the fight against this climate emergency. From predictive analytics (what might happen?), such as forecasting renewable/ intermittent generation so network operators can balance the grid, to prescriptive analytics (what should we do?): optimisations of assets such as batteries to decide what to do based on forecasts etc., to minimise costs.

The stronger the financial case that can be made for renewables and assets that enable their use, the more accessible they will be to everyone, and the more likely people will adopt them.

I believe that in order to be free of fossil fuels, we have to go to a more decentralised system where generation is more localised, e.g. solar PV on rooftops or creating microgrids in communities. In such a decentralised system, thousands of controllable devices could help balance the grid in real time. We can’t do this manually, and thus AI is essential in making all the constituent components of this system work together in concert in the most optimal way.

What skills do you think your role requires the most?

Good problem-solving skills are essential. More often than not, there are a lot of unknowns or unclear requirements for a new project, nor is there usually a clear template or example on how to tackle it. This is because it is the first time for all of us to do certain tasks or projects and, in some cases, the first time for the wider industry.

Another key skill is communication. Data touches all parts of our business, and different teams will have different knowledge of data analysis and software development. You have to adapt your approach in discussions accordingly.

How do you and your team tackle complex challenges?

One of the most effective ways to overcome challenges is to break them down into smaller subproblems, solve the smaller problem and then use that to tackle the next/ bigger problem. There’s an algorithmic technique called dynamic programming which does precisely this, but we can apply this to higher level/ abstract concepts we may face.

Another way to conquer challenges is to use data. Everyone will have an idea or an initial viewpoint to drive an investigation. Still, after analysing the data, the results will often challenge the original assumptions that we held, and new avenues for investigation can arise. The rabbit holes can be very deep, so it’s best to use evidence to guide decision-making.

What are your top tips for someone considering a career in data science & in the renewable energy industry?

Many businesses require pragmatism, which for a data scientist sometimes means that producing a simpler solution on time is more valuable than a “cool” solution late. The best data scientists are those who put the end-user/customer’s needs first.

I mentioned earlier that one essential skill in my role is problem-solving. This can be honed over time, but it does take practice. I first learned to programme by solving online challenges such as There are multiple ways to solve a problem, and once you’ve solved one, you can compare it with other solutions posted by the community. I learned a lot from this in my early days on how to make my solutions faster, but also the practice on toy problems taught me how to approach a problem, break it down and then solve it.

Lastly, I’ve seen many people take a sideways step into data science from another team in the same company, particularly from operational and commercial roles, which require similar skills. Two of my current team members come from operational/commercial analyst roles in Moixa, and the same was true in my previous company.

Here at Moixa, we invest in people who want to learn and further their careers. For example, we offer Python training sessions to anyone in the company who is interested in learning this programming language. We have people from Business Development, Product Management, and Operations teams attend and grow their skill sets, learning to use Python to help with their daily tasks.

Interested in how other Moixans have used their training budget? Check out our latest blog.

What is the one thing we should know about you that isn’t on your CV or LinkedIn profile?

I like to do DIY and I’m in the middle of renovating my house.

A huge thank you to Luke for discussing his role at Moixa and sharing many helpful tips on starting a career in data science. You can connect with Luke on LinkedIn here.

Do you want to join our team and help us build the energy system of the future? Visit our careers page and see if there’s a role that suits you, or get in touch directly with our Talent Acquisition Manager Gail Solomon.