What is your academic background? 

I have a PhD from the University of Western Australia and Master’s and Bachelor of Science (Honours) from Melbourne University. The focus of my academic studies was on environmental science, GIS, statistics and remote sensing.

Have you completed any other training in data science?

Absolutely. I take every chance I can to attend courses, seminars, webinars, conferences and online training. The field we work in is evolving at such a rapid pace that ongoing up-skilling and training this is essential. Plus, it is exciting to see new analyses and applications. One good outcome from COVID 19 has been the exponential increase in online learning opportunities.

If you pivoted into a data science role from another area, how did you go about this and what advice would you give others looking to do the same?

Once completing my PhD the focus of my work was in the marine environment. I lead a number of large collaborative postdoctoral research programs both here in WA and in Hawaii. My research focussed on marine management, biodiversity mapping and predictive species distribution modelling. However, my interest in data science is less academic and more about providing actionable intelligence. I wanted to transition from academia into industry and specialise in terrestrial remote sensing and data science.

I spent two years up-skilling. My philosophy was to embrace every opportunity to learn and gain experience. This included some interesting and unexpected ones. I worked as a Remote Piloted Aircraft (drone) Instructor, an Observer on Helicopters at Heliwest and an Airborne Sensor Operator at Australian Air Affairs. I wanted first-hand experience collecting data using airborne platforms (i.e. drones, planes, helicopters) to really understand the opportunities and limitations. All these experiences provided new skills enabling me to transition to the data scientist role I have now.

What sparked your interest in working with data?

I enjoy synthesising data to understand and make sense of our world. The recent increase in accessibility of remotely sensed data (e.g. drones, manned aviation and satellites) and availability of powerful tools to process and analyse this data is too exciting an opportunity not to explore. With this data we can map active bushfires and keep people safe, model species distributions and protect critical habitat, track vegetation health through time, measure surface deformation and map landslide risk. It is an incredibly rewarding career.

How did you come to work in your current role?

I applied for a Geospatial Data Scientist role advertised by Fortescue on Seek. They were looking for a combination of data science experience and remote sensing expertise. I was surprised by the breath of remote sensing applications the business relies on and excited by the maturity of the software and hardware systems currently in use. More importantly, I was impressed by Fortescue’s active commitment to their business culture and values. The best part of my current role is the great team that I work with within Technology and Autonomy.

What sort of projects have you been working on?

Currently, my team and I are responsible for supporting and delivering a range remote sensing data and geospatial data analytics to support Fortescue’s operations. These include a range of change detection analyses, risk modelling and mapping, vegetation health monitoring, bushfire risk management, mineral alteration mapping, surface deformation mapping, soil moisture mapping and ground disturbance mapping.

My most rewarding project has been working with Fortescue Emergency Services to develop a bushfire risk management platform. We applied a remote sensing approach to develop a targeted and spatially explicit bushfire risk assessment framework. The web portal and mobile application enables users to view and interact with high resolution bushfire risk data across Pilbara operations. The data platform supports managers to prioritise control measures such as prescribed burning.

What tools/platforms do you use in your work?

Software and data platforms I regularly use include Google Earth Engine, ArcGIS Enterprise, ENVI, ERDAS Imagine, FME, Geocortex, Python, Pix4D, QGIS, R, SNAP. Hardware includes a wide range of sensors (e.g. multispectral, SAR, LiDAR) and remote sensing platforms (e.g. satellites, drones, trains, planes and automobiles).

What has been a highlight of your data science career so far?

During the 2020 bushfire emergency I worked as an Airborne Sensor Operator with Australian Air Affairs. Our division was responsible for mapping active bushfires for NSW RFS and QLD DFES using manned aircraft. The high resolution ‘heat’ maps (i.e. visible, thermal and near infrared), downloaded within 10 minutes of us flying over the bushfire, were used to direct emergency response operations on the ground. My highlight was going to Queensland DFES head office to meet the emergency coordinators and see the data being used in real time.

What challenges have you faced as a data scientist?

My biggest challenges are data quality, quantity and managing realistic expectations. The advent of automation and machine learning has raised expectations that a machine can be trained to answer all our questions. While we see many great examples where this is possible, successful applications are very dependent on the question being asked, the quality of the data and ensuring results are validated. Managing expectations in this space can be challenging.

What are some of the big areas of opportunity/questions you want to tackle in this space?

With the massive increase in availability of earth observation data, I am currently working on improved data analysis workflows that optimise data integration and validation. Having robust workflows that automate the integration and analysis of disparate sources of long-term earth observation data and environmental data is a big area of opportunity.

What excites you most about recent developments in Data Science?

I am excited about the exponential increase in high resolution, high quality, open-source data. Especially analysis ready satellite imagery. This in conjunction with cloud-based platforms are addressing many of the bottlenecks in data science. The maturation of data analysis platforms ingesting this data and converting it into actionable intelligence supports businesses to focus on making informed decisions not wrangling data.

What does the future of data science look like?

A combination of increased data availability, accessibility and processing power means soon the only limit is our imagination! Future data scientists equipped with a sound understanding of how data is captured, manipulated and analysed will have the power to tap into any data stream to provide valuable insights, intelligence and automation. From business intelligence to robotics, earth observation, cyber security, environmental monitoring and emergency response. Anything is possible.

For people considering a career in data science, what is one piece of advice you would give?

Always keep your mind open to new directions and new opportunities. Don’t be afraid to change directions in your career. If you see an opportunity, find ways to chase that opportunity. Never stop investing in your career. Attend conferences, courses, do online training, up-skill, network. Be driven by finding opportunities to solve problems that matter.