What does data science mean in the built environment sector?
Quy Vu, a data scientist at Mott MacDonald, explains exactly what a career in data science and the built environment looks like.
The demand for data scientists has skyrocketed in the past decade and the potential for data science to make a powerful and meaningful difference to the world isn’t lost on Mott MacDonald and the built environment sector.
Quy Vu, a data scientist at Mott MacDonald, offers some exclusive insights into how working with data can lead to real-world changes and the types of skills that graduates need to build – in and out of university – to get ahead.
Data science and the built environment
Mott MacDonald defines data science as using statistical, programming and machine learning knowledge to follow a scientific and data-driven approach to solving a particular problem. In the built environment, your projects will be linked to infrastructure. For example, you could be considering how to get a train to run more efficiently or predicting pipe failures or flooding. ‘The water sector is quite a big focus for us at the moment,’ says Quy.
Sound like small work? Not exactly. Data scientists work across any and all of Mott MacDonald’s sectors, including education, environment, health and international development. They are creating solutions to some of the world’s biggest and most pressing challenges. ‘The projects we work on are all very high impact; they’re going to affect millions of lives,’ says Quy. This is just one of his favourite things about his job.
‘I wouldn’t have the opportunity to solve the problems I’m working on if I didn’t work at Mott MacDonald. These problems don’t have answers you can find on the internet!’ he adds. ‘That’s why data science is so crucial at Mott MacDonald. There’s no shortage of new and unconventional problems that are not well researched, alongside problems that have been studied for centuries. Traditional approaches to solving them take a lot of time and are heavily constrained by assumptions. A data science approach is helping to create quicker, more adaptable, more accurate and more intelligent solutions that have previously been out of reach.’
The job of a data scientist at Mott MacDonald
Mott MacDonald is described as a consultancy, but Quy would not call the data science team consultants: ‘I would say our role is more of a research and development role than a consulting role. If anything, we are more like a data science team in a software company. We spend a lot of time developing products that will solve problems across the built environment sector, not just for one client.’
The data science team works with multiple clients, both internal and external. ‘I spend roughly 20% of my time talking to clients,’ Quy says. ‘It’s important for us to understand what they are trying to do and what problems they’re up against so we can build products to address these concerns.’
Day-to-day tasks and projects
As a data scientist at Mott MacDonald you won’t be limited to just one type of task. Quy explains that there is a huge variety in day-to-day work, including:
- Cleaning and moulding the data – for example, if the data is in a database, you will need to clean and transform it to a format that can be easily analysed and store it safely, but in a way that can be easily accessed by other developers.
- Exploring and analysing the data – what are the trends and patterns that you need to capture?
- Building the statistical and machine learning model – so you can extract insights from the data, predict the future and inform decisions!
- Integrating or productionising the model – so it can be used by other people, such as teams at Mott MacDonald or clients, and even other applications as well.
Software and programming languages are vital parts of the job. Quy identifies Python and SQL as the two languages he uses the most. When it comes to productionising a machine learning model, he uses cloud computing services such as Azure and Amazon Web Services.
Do data scientists work with others?
Collaboration is at the core of a data scientist’s role, so they rarely work in isolation. ‘As a team, we are usually always communicating with each other and, if there are any problems, we reach out, whether we’re sat in the office together or we’re working from home through Microsoft Teams, says Quy. ‘We also attend meetings with clients or technical leads at Mott MacDonald to discuss solutions.’
And at Mott MacDonald the emphasis is on combining digital expertise with domain experience, which is why data scientists partner up with engineers to work together on a solution. ‘This way, we combine our knowledge and work together to create the best solution,’ says Quy.
‘I get to learn a lot about different fields, such as transport and water. I talk to engineers from different disciplines who can provide specialist knowledge about their area and the problem I’m tackling. For example, If I want to predict pipe failures, do they know of any factors I need to consider? This knowledge sharing works both ways, too: if I find something interesting, I’ll always go back and tell them.’
Training for data science graduates
You will start your career as a graduate data scientist. In your first few months, you’ll be developing your software engineering and data handling skills, as well as familiarising yourself with the platforms the team uses – all with the help of your senior colleagues. After this, you’ll start to work on developing and productionising statistical and machine learning products.
In addition to on-the-job training, Mott MacDonald offers further development opportunities. If somebody finds an education resource (course, conference, hackathon etc) that they think would be helpful, there is a budget for training, and graduates are encouraged to participate in events to brush up on their technical skills. ‘In my first year, I went to three hackathons and one conference, which were great learning experiences,’ says Quy.
You’ll also be enrolled onto Accelerate Your Future, a structured three-year development programme that develops the strengths graduates need to be successful at Mott MacDonald. It gives you the opportunity to network and meet other graduates in your cohort, too. The programme is a mix of residential events, classroom-based learning, virtual webinars and business challenges.
Mott MacDonald understands that each person’s career goals vary, so they tailor-make each individual’s development programme to suit them. With a vast library of e-learning courses available to you, you can choose which direction you want your career to go in.
How could your data science career progress?
After a few years, you could be promoted to senior data scientist. From there, you can choose to become a data science manager or data science technical lead. For Quy, the technical lead route appealed more: ‘I really enjoy the technical challenge and problem solving, whereas the manager route is more about supporting people, coordinating a team and building a community.’
And Quy is particularly pleased with his career progression: ‘One of the biggest highlights of my career is being promoted two times in two years, which is all thanks to how the progression at Mott MacDonald works, especially in our digital division. It doesn’t require you to put in X number of years before a promotion, so you can fast track your career if you can deliver value and demonstrate excellence.’
What skills and experience does a data scientist need?
According to Quy, the top three skills for data scientists are:
- problem solving
- knowledge in statistics and machine learning
You can develop each of these skills while you’re at university. Quy recommends: ‘For problem solving, try to think of anything in your life that you could improve. Start from there and work your way up. For example, if you are waiting for a train because it’s delayed, can you estimate how long it will take based on your experience with previous delays? Can you apply your knowledge in statistics to say how confident you are with your estimation?
For knowledge in statistics and machine learning, attend your lectures and, outside of your degree, try to get involved in as many data science projects as possible. Data science is half a science but also half an art, so you need to be knowledgeable, but you also need to be practical and hands-on to be successful. You can find data sets on data science communities such as Kaggle. Again, for programming, teach yourself languages such as Python and SQL and apply them to your data science projects.’
You can also look for more formal work experience in a data science role, which will help you to develop these skills. Mott MacDonald, for example, offers summer internships.
How to impress Quy in applications and interviews
Quy is involved in assessing applications for Mott MacDonald’s graduate schemes and internships – and there are a couple of things he’s looking for.
'There will be a lot you (and I) don’t know about data science so, for me, the most important thing is to have a good learning attitude. Are you open to learning new skills or doing things you’re uncomfortable with?’ says Quy. ‘When I’m looking at an application form, I would be impressed by any side projects or online courses you’ve completed in your spare time.’
He adds: ‘I also want to know if you’re curious. If there’s something you don’t know, do you want to learn more about it? Do you spend time thinking about what’s happening around you? You’re applying to a civil engineering firm, so an interest in the built environment is still important, but I focus more on curiosity and learning attitude.’
Head to Mott MacDonald's employer profile on TARGETjobs to find out more about its opportunities for students and graduates.